You are on page 1of 147

PROCEEDINGS

ICDER - 2014
INTERNATIONAL CONFERENCE ON DEVELOPMENTS
IN ENGINEERING RESEARCH

Sponsored By
INTERNATIONAL ASSOCIATION OF ENGINEERING &
TECHNOLOGY FOR SKILL DEVELOPMENT
((Registered Under Indian Trust Act, 1882)


Technical Program
8
TH
October, 2014
Hotel Ramee Guestline, Tirupati

Organized By
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL
DEVELOPMENT
www.iaetsd.in

Copyright 2014 by IAETSD
All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system, or transmitted, in any form or by any means, electronic,
mechanical, photocopying, recording, or otherwise, without the prior written
consent of the publisher.

ISBN: 378 - 26 - 138420 - 9

http://www.iaetsd.in











Proceedings preparation, editing and printing are sponsored by
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY
FOR SKILL DEVELOPMENT COMPANY
About IAETSD:
The International Association of Engineering and Technology for Skill
Development (IAETSD) is a Professional and non-profit conference organizing
company devoted to promoting social, economic, and technical advancements
around the world by conducting international academic conferences in various
Engineering fields around the world. IAETSD organizes multidisciplinary
conferences for academics and professionals in the fields of Engineering. In order
to strengthen the skill development of the students IAETSD has established.
IAETSD is a meeting place where Engineering students can share their views,
ideas, can improve their technical knowledge, can develop their skills and for
presenting and discussing recent trends in advanced technologies, new educational
environments and innovative technology learning ideas. The intention of IAETSD
is to expand the knowledge beyond the boundaries by joining the hands with
students, researchers, academics and industrialists etc, to explore the technical
knowledge all over the world, to publish proceedings. IAETSD offers
opportunities to learning professionals for the exploration of problems from many
disciplines of various Engineering fields to discover innovative solutions to
implement innovative ideas. IAETSD aimed to promote upcoming trends in
Engineering.










About ICDER:
The aim objective of ICCTER is to present the latest research and results of
scientists related to all engineering departments topics. This conference provides
opportunities for the different areas delegates to exchange new ideas and
application experiences face to face, to establish business or research relations and
to find global partners for future collaboration. We hope that the conference results
constituted significant contribution to the knowledge in these up to date scientific
field. The organizing committee of conference is pleased to invite prospective
authors to submit their original manuscripts to ICDER 2014.
All full paper submissions will be peer reviewed and evaluated based on
originality, technical and/or research content/depth, correctness, relevance to
conference, contributions, and readability. The conference will be held every year
to make it an ideal platform for people to share views and experiences in current
trending technologies in the related areas.














Conference Advisory Committee:

Dr. P Paramasivam, NUS, Singapore
Dr. Ganapathy Kumar, Nanometrics, USA
Mr. Vikram Subramanian, Oracle Public cloud
Dr. Michal Wozniak, Wroclaw University of Technology,
Dr. Saqib Saeed, Bahria University,
Mr. Elamurugan Vaiyapuri, tarkaSys, California
Mr. N M Bhaskar, Micron Asia, Singapore
Dr. Mohammed Yeasin, University of Memphis
Dr. Ahmed Zohaa, Brunel university
Kenneth Sundarraj, University of Malaysia
Dr. Heba Ahmed Hassan, Dhofar University,
Dr. Mohammed Atiquzzaman, University of Oklahoma,
Dr. Sattar Aboud, Middle East University,
Dr. S Lakshmi, Oman University






Conference Chairs and Review committee:

Dr. Shanti Swaroop, Professor IIT Madras
Dr. G Bhuvaneshwari, Professor, IIT, Delhi
Dr. Krishna Vasudevan, Professor, IIT Madras
Dr.G.V.Uma, Professor, Anna University
Dr. S Muttan, Professor, Anna University
Dr. R P Kumudini Devi, Professor, Anna University
Dr. M Ramalingam, Director (IRS)
Dr. N K Ambujam, Director (CWR), Anna University
Dr. Bhaskaran, Professor, NIT, Trichy
Dr. Pabitra Mohan Khilar, Associate Prof, NIT, Rourkela
Dr. V Ramalingam, Professor,
Dr.P.Mallikka, Professor, NITTTR, Taramani
Dr. E S M Suresh, Professor, NITTTR, Chennai
Dr. Gomathi Nayagam, Director CWET, Chennai
Prof. S Karthikeyan, VIT, Vellore
Dr. H C Nagaraj, Principal, NIMET, Bengaluru
Dr. K Sivakumar, Associate Director, CTS.
Dr. Tarun Chandroyadulu, Research Associate, NAS

ICDER - 2014 CONTENTS

1
A Layered Security Approach through FEMTO CELL using Onion
Routing in MANET
1
2
A Survey on Enroute Filtering Scheme in
Wireless Sensor Networks
7
3
Estimation of Damping Torque for Small-Signal Stability of Single
Machine Infinite Bus System
13
4 An Enhanced Feature Selection for High-Dimensional Knowledge 22
5
Improving the Location of Nodes in Wireless Ad Hoc and Sensor
Networks Using Improvised LAL Approach
32
6
Survey on Wireless Sensor Networks Routing Protocols Based on
Energy Efficiency
38
7
AN EFFECTIVE ALARMING MODEL FOR DANGER AND ACTIVITY
MONITORING USING WEARABLE SENSORS FOR CHILDREN
45
8
An Efficient Way of Classifying and Clustering Documents Based
on SMTP
49
9
An Efficient Way of Detecting a Numbers in Car License Plate
Using Genetic Algorithms
54
10 RF Controlled Sailing Robot for Oceanic Missions 58
11 Solar Power Satellites 60
12
Autonomous pick and place rover for long distance surveillance
using ultrasonic sensors
66
13 Intelligent Bus Alert System For Blind Passengers 71
14
Traffic Sign Recognition for Advanced Driver Assistance System
Using PCA
78
15 Ethernet Based Intelligent Security System 84
16
Modified Artificial Potential Fields Algorithm for Mobile Robot
Path Planning
93
17
Protecting Privacy Preserving For Cost Effective Adaptive Actions
Using Memitic Algorithm
98
18
Use of Mobile Communication in Data-Intensive Wireless
Networks
105
19
Confidential Multiparty Computation with Anonymous ID
Assignment using Central Authority
111
20
DESIGN OF RF BASED VOICE-CONTROLLED MULTI-TERRAIN
ROBOT TO TRAVEL ON UNEVEN SURFACES
117
21 Zigbee For Vehicular Communication Systems 122
22
An Efficient and Large data base using Subset Selection
Algorithm for Multidimensional Data Extraction
129
23
A Review on ECG Arrhythmia Detection based on DD-DW
Transformation
136





Abstract In earlier process, making a secure routing only
discussed and never discuss about how to transfer data in a
secured manner. Even though we performed routing in a
secured manner, there will be chances of data should be
dropped or revealed by an illegal persons. We use an onion
routing to make a highly secured routing, so this routing
includes the mechanism of layered by layered approach from
one node to another nodes. And we transfer the data in
secured manner by sending dummy packets from source to
destination, and these dummy packets are mold up by the
mechanism of node characterization technique. And in earlier
process they never looked for time and speed reduction. That
is, in existing system, they use the concept of secured routing
each and every time when data will be sent. So this will increase
the speed reduction process which is at every moment of
transaction we need to perform separate routing, to overcome
this we proposed onion routing and node characterization. And
we using femto cell device for strengthening the signal when
there is no sufficient signal to work on the process.

Keywords MANETs, Routing Protocols, Secure Onion Routing,
Layered Approach, Packet authentication, Femto cell, NS2.

I. INTRODUCTION
1.1 MANET
The term MANET refers to Mobile Ad-hoc NETwork.
MANET is a less infrastructure network; nodes are under
mobility. It should be moved here and there and it wont rely
stable. Mobile Ad-hoc networks are most familiar to security
issues due to the characteristics of such networks such as a
wireless medium and dynamic topology. It is very harder to
provide trusted and secure communications in enemy
environments such as battle fields. On one hand, the
malicious persons outside the network may have an idea to
reveal the information about the communicating nodes, even
when communications are encrypted. On the other hand
nodes involved inside network will always be trusted, since
a trusted node may be hold by illegal person and become
malicious.
In MANET, a set of interacting nodes should cooperatively
implement routing functions to enable end-to-end
communication along dynamic paths composed by multi-


hop wireless links. Several multi-hop routing protocols have
been proposed for MANET, and most popular ones include:
Dynamic Source Routing (DSR), Optimized Link-State
Routing (OLSR), Destination-Sequenced Distance-Vector
(DSDV) and Ad Hoc On-Demand Distance Vector (AODV).
Most of these protocols are do their function on assumption
of trusted manner. But sometimes it could not been trusted it
behaves like secured less, because when there is a presence
of malicious node at that time it emerges the weakness of
MANET to cause various kinds of attacks.

1.2 ADHOC
Ad-hoc is a Latin word which refers to all purpose. For
example, take two access points as access point 1 and access
point 2. There will be twenty kilometer distance in between
those two access point.



Access Point 1 20 k Access Point 2
If user A starts to download the data from access point 1,
immediately user A tends to travel from access point 1 to
access point 2. Now download is under half fulfilled. When
user A which comes under out of coverage from access point
1 due to travelling. They face a problem about downloading.
Due to these criteria, we go for intermediate nodes concept
to prevent interrupts which occurred in data transfer. So, that
time the concept of Ad-hoc is established i.e. intermediate
node is formed. Now data could be transformed very
successfully but there will be a problem occurred under
security concept. Because now the job is hold under
intermediate nodes, we does not known details about
intermediate nodes i.e. either it is a legal or illegal. When it
is illegal, data will not been under secured way. So the major
cause of this paper is about,
Done secure routing
Along with that the data transfer also will be in an
secured manner
The remainder of this paper is organized as follows. The
basic concepts and routing processes are analyzed in Section
II. The protocol design is presented in Section III. The
protocol evaluation is discussed in Section IV. Performance

A Layered Security Approach through FEMTO CELL using
Onion Routing in MANET
T. Sathish
1
, M.Senthil Kumar
2

1
Department of Computer Science and Engineering, M.E. Scholar, Anna University,
Sree Sowdambika College of Engineering, Aruppukottai, TamilNadu, India.
pr i nceof per si a123555@gmai l . com
2
Department of Computer Science and Engineering, Assistant Professor, Anna University,
Sree Sowdambika College of Engineering, Aruppukottai, TamilNadu,India.
r msent hi k@gmai l . com
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
1



analysis is evaluated in Section V and Section VII concludes
this paper. The future scope is conferred in Section VII.

II. BACKGROUND AND RELATED WORK
In existing system, they tell about how to perform a secure
routing. They choose the main concept of secure routing as
onion routing. Why they named as secured means, how an
onion makes different layers when we start to cut it into
small pieces, likewise here also they we forms different
layers for checking each and every nodes as an one by one.
It is like a layered i.e. what is the action handled by two
nodes as a one layered. So we can name its for our
understanding as onion routing is also known as layered
approach.




Source Intermediate Nodes Destination
Fig.2.1 Layered approach
The main concept done in existing system is they did how to
do secure routing for providing path. This will be discussed
as follows. To accommodate secure rating these cases
requires three kinds of techniques, they are:
Setting a trap
Routing by means of layered approach(onion
routing)
Signature process
2.1 SETTING A TRAP
The general meaning of trap is to find out an unidentified or
illegal person by without knowing an alert for the person
whom is involved. The main concept involved in this
technique is as follows. Let consider, there are six nodes
from source to destination.




Source Destination

To set a trap separate id is given to each node. The id
should also been included for source and destination node
too. For Source Node,







I1 identity id for Node 1

So, these processes are going on until data or message
reaches the destination node. By setting an identity for each
node I can able to find the illegal persons, because illegal
persons do not have their identity number. So step by step, it
travels from source node to destination node which comes
across through different intermediate nodes. The nature
cause of this mechanism is to set a route request message
from source to destination node.

2.2 ONION ROUTING
Once a route request message is perfectly reached by
destination node, the destination node again they start a
work about route reply. The ROUTE REPLY is a main
process for this mechanism, by the mechanism using in my
project is about onion routing. In setting a trap, the route
request message is starts from source node to destination
node. But here in route reply it starts from destination node
to source node.
WORKING PROCESS OF ONION ROUTING
It is a simple process here we are going to removing the
identity number of each intermediate nodes from destination
end points to source end points(reverse process).


Source Intermediate Nodes Destination



Fig.2.2 Process of Onion Routing

In the above diagram route reply request is send from
destination node and its moved to the intermediate node.
Here how they will choose intermediate node means that the
path which the route request came across. So its starts from
the intermediate node first its goes for last intermediate
node. In that it removes the top of the key which is present.

Intermediate Node



From this diagram, I2 will be removed. This is a top of the
key. Once it is finished then it will be moved for another
node. This process is known as layered approach and finally
its came to source node and delivered the route reply
message.

2.3 SIGNATURE PROCESS
In signature process, we are going to generate group
signature and along with that session key. The main process
key is to finish their process with in their time limits which
we set for that.




SID DID Encrypted Message
SID DID Encrypted Message I1 I2
SID DID Encrypted Message I1 I2
SIO DIO Encrypted Message Source Sign I1 Sign
Layer 1 Layer 2
Layer 3
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
2



For example, the session number=5 packets means, there
should be five packets could be sent at a time. If there is
exceeding of any packet count there will be some jobs did by
an illegal person
III. PROTOCOL DESIGN
We invent a much effective method to make an protection
under both packet dropper and modifiers, in this method
lets take upon routing tree which is at sink it will be
accomplished very first. At the time of sensor data are make
an survey along the tree structure towards the sink, every
packet sender adds upon small number of extra bits to it,
which is also known as packet markers to the packet. And
then runs up the node characterization algorithm for
identifying nodes which hold packet dropper and modifier.
This is the one of the way to identify the bad nodes.

3.1 NODE CHARACTERIZATION ALGORITHM
This algorithm includes the following methods. They are,
o Stepwise ranking method
o Global wise method
o Hybrid ranking method
Using this algorithm, we are going to send dummy packet
with sequence number, random number and padding.
Sequence number is for the packet which we are
going to sent.
Random number for the node, here why we choose
random number for node means, to find out the
hacker which is either it was good node or bad
node.
Padding shows the position of the node. If it is a
first node means, it shows the position as 0th node
(zeroth node).




Source Intermediate Nodes Destination
Fig.3.1 Parameters of Node Characterization Algorithm
For example, I need to transfer 50 kp packets
means, I split as 10 kp as first packet and 20 kp as
second packet and 20 kp as third packet and it is an
automated one.
From the above diagram, first packet of node
includes,
This information will be in the source node only
and this is a general concept.

PROCESS WHICH ARE GOING
Step 1:



Source Destination
When the first dummy packets are in Node A that is under
source node it holds the information of,

0
th
node
Here sequence number provides some identity number for
first packet as for example: 1. And random number also
has been provided. After that padding will be stored as 0
th

position because its a first position of the node.

Step 2:
Similarly the same first packet will be moved from Node A
to Node B. The Node B holds the information like,

1
st
Node
When same first packet will be moved from Node A to Node
B, it should contains same sequence number, and also been
with source random number id, along with that new
intermediate node random number, and finally intermediate
node padding position is 1
st
Node.


C
Source Destination
(Now, we are in Node B)
Step 3:
The same first packets move from Node B to Node C. For
example, we found that we have packet modification. So,
this packet modification shows that the presence of changed
packet information. That is an anonymous person could
modify the data. For example, how can we find our data will
be modified means?
1. There will be changed in sequence number example
we use 1 as an sequence number means these will
be changed as 1.4 or 1.7 which what we given it
wont present here.
2. The random number should not been generated for
illegal person.
Step 4:
If we are going to send second packet means, the sequence
number will be changed as 2 then its move from Node A ,
Node B, Node C.
General Rule:
o First packet 1 sequence number
o Second packet 2 sequence number
o Third packet 3 sequence number
IV. PROTOCOL EVALUATION
The proposed system has the following features:
Identifies both a kind of malicious activity like
packet dropping and modifying.
Achieving low communication and overheads.
SID DID Sequence Number Random Number Padding
SID DID Sequence Number Random Number Padding
SID DID Sequence Number Random Number Padding
SID DID Sequence
Number
Intermediate
Random Number
Intermediate
Padding
B C D A
B C D A
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
3



Much more suitable for existing false packet
filtering schemes that is this scheme will make
an end point for malicious packets.
We going to deployment large number of sensor nodes in an
two dimensional area. The nature job of each sensor nodes
generate the data regularly and intended to forward packets
towards an sink and the sink is located inside the network.

A. STEP WISE RANKING METHOD
Yet now we find the one will be safer node. So, now we
going to note which node will be good node or which will be
bad node. So for finding, I will just make all nodes under
tree formation. That is the nodes which are nearby.















Fig.4.1 Tree Structure of Nodes Formation

For example, in node A we have 10 packets, now we going
to check of these whether this remaining nodes are good
node or bad node. Node A sent 10 packets to every nodes
which are present in the tree (b,c,d,e,f,g,h,i,j,k,l,m,n,o).



20/10
10/10


9/10 7/10






Fig.4.2 Step Wise Ranking Method

Good Node 10/10
Bad Node 20/10, 8/10, 7/10
Unknown Node (Either Bad or Good) 9/10,11/10
By using this, we may find either which one will be good
node or which one will be bad node. It is a step wise ranking
method. We are going to see step by step which is from top
layer of the tree. When we seen, Node B as 10/10 means it
will be good node and it will be selected. Similarly, take
Node D, it has 9/10 so it is under the scenario of unknown
either bad or good. If we reject these nodes, their child node
also deleted and it is a drawback of this method.

B. GLOBAL WISE RANKING METHOD
To overcome the drawback which are comes under stepwise
ranking method we need to seen these node under over all
view that is which one will be best. By selecting that best
one, we are going to use the best one for our process that is
for packet transaction.

C. HYBRID RANKING METHOD
These methods are combination of stepwise ranking method
and global ranking method. Here the node which is having
highest priority will choose as first and it is most likely bad
node.

D. FEMTO CELL
Femto cell is an device which act as wireless access point ,
its used to strength the signal and to provide the same signal
range whatever its came from the core network and the
range of femto cell process about 10 metres. These femto
cell are particularly designed for small environment, so this
theme got correlated with my project which is after selecting
an particular route for my data transmission here Ill have
some less amount of nodes , so the data starts to travel
across the node which I framed. The main problem is node
coverage will be differs according to distance of each node
from the core networks and this leads to higher time
consumption in data transfer. So when we use femto cell
near to node which does not have fulfilled coverage means
we can able to successful in over-come these drawback.




V. PERFORMANCE ANALYSIS
5.1 NETWORK CONFIGURATIONS
TOPOLOGIES
In our research the network area is about 1500m x 600m
with 80 nodes normally and it should be equally distributed.
Here the function of distributed coordination IEEE 802.11 is
used as the MAC layer. And the capacity of channel is about
A
B C
D E F
G
H I J K L O M N
A
B C
D E F
G
H I J K
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
4



3Mbps, transmission range is occurred was 250m.The
mobility in node should be there by using random way point
model. The total of 20 UDP based CBR sessions are used to
create the network traffic.

ATTACKING MODELS
The general assumption about the intermediate nodes along
the route may become malicious If there is any malicious
node then the routing packets are randomly dropped. For
example in previous routing like ANODR and AODV will
suffer more packet loss. The AASR detect the malicious
node by using the signature method and find out the details
of attackers in routing table.

SIMULATION RESULTS
Effects of Mobility Scenario
To simulate the enemy environment, we are going to choose
twenty percentage of total nodes, which is ten nodes as
malicious nodes and then we can able to change the network
mobility from one to eight m=s and record the performance
results. From the results the average nodal speed increases,
similarly the throughput also varies because nodes have an
capability of move randomly. Due to performance variation
secure routing always achieve highest throughput. The
following figures a, b, c describe the performance of
different mobility settings.

a. Per Flow Throughput


c. Packet Loss Ratio


c. End-to-end Delay


Effects of Malicious Attacks
The configuration of mobile network with an average as
4m=s, therefore in general the number of malicious nodes
increases similarly the throughput decreases. The following
figures a, b, c show the performance in the presence of
different number of malicious nodes

d. Per Flow Throughput



e. Packet Loss Ratio

13
13.5
14
14.5
15
15.5
16
1 2 3 4 5
A
v
e
r
a
g
e

T
h
r
o
u
g
h
p
u
t

P
e
r

F
l
o
w

(
k
b
p
s
)
Mean Node Speed (m/s)
AODV
SR
ANODR
8
10
12
14
16
18
20
22
1 2 3 4 5
P
a
c
k
e
t

L
o
s
s

R
a
t
i
o

(
%
)
Mean Node Speed (m/s)
AODV
SR
ANODR
0
50
100
150
200
250
300
350
1 2 3 4 5
D
a
t
a

P
a
c
k
e
t

D
e
l
a
y

(
m
s
)
Mean Node Speed (m/s)
AODV
SR
ANODR
14
14.5
15
15.5
16
16.5
17
1 3 5 7 9
A
v
e
r
a
g
e

T
h
r
o
u
g
h
p
u
t

P
e
r

F
l
o
w

(
k
b
p
s
)
Number of Malicious Node
AODV
SR
ANODR
2
4
6
8
10
12
14
1 3 5 7 9
P
a
c
k
e
t

L
o
s
s

R
a
t
i
o

(
%
)
Number of Malicious Node
AODV
SR
ANODR
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
5





f. End-to-end Delay


VI. CONCLUSION
There will be lot of MANET security routing concept can be
obtained, but yet now there will be no intimation to secure
packet authentication. If there is any malicious activity
means those packet will be dropped so the information might
be destroyed. Therefore this is a correct solution for an
hackers or intruders they achieve their task which means
they finish their main job. But Ill find a solution by
introducing the dummy packet concept with some secured
mechanism, this will helpful for lightning a successful path
of conveying original information from source to
destination. And Ill used femto cell to make more
strengthened signal coverage for achieving more efficient
process.
VII. FUTURE SCOPE
The nature cause of anonymous is come due to a
confidential message or data which is sent through an
intermediate node. So I first focused about how to make an
malicious intermediate node into trustable intermediate
node. Ill going to accomplish this task by generating
survey about nodes by finding out good node, bad node and
unknown status about either good or bad node. It is an
attainable solution by using the ranking method concepts.
VIII. REFERENCES
1. (Nevin) Lianwen Zhang and David Poole, Stepwise-Decomposable
Influence Diagrams, Department of Computer Science, University of
British Columbia, Vancouver, B.C, V6T/Z2, Canada.
2. N.Vanitha, G.J enifa, Detection of Packet Droppers in Wireless Sensor
Networks Using Node Categorization Algorithm, International J ournal
of Advanced Research in Computer Science and Software Engineering,
Volume 3,Issues 3,pp.69-74, March 2013.


3. Hussein Al-Bahadili and Khalid Kaabnel, Analyzing the
Performanceof Probabilistic Algorithm in Noisy MANETs,
International J ournal of Wireless and Mobile Networks, pp.82-94,
Vol.2, No.3, August 2010.
4. Shio Kumar Singh, M.P Singh and D.K Singh, Routing Protocols
in Wireless Sensor Networks - A Survey, International J ournal
of Computer Science and Engineering Survey, Vol.1, No.2,
November 2010.
5. Hee Yong Youn, Chansu Yu, Ben Lee,Routing Algorithms for
Balanced Energy Consumption in Ad-hoc Networks.
6. Zaiba Ishrat, Pankaj Singh, An Enhanced DSR Protocol Using
Path Ranking Technique, International J ournal of Engineering
Research and Applications, Vol.3, Issue 3, pp.1252-1256, May-
J UNE 2013.
7. Wenjia Li, AnupamJ oshi and TimFinin, SMART: An SVM-
Based Misbehavior Detection and Trust Management Framework
for Mobile Ad-hoc Networks.
8. Ruchi Rani, Manisha Dawra, Performance Characterization of
AODV Protocol in MANET, International J ournal of Advanced
Research in Computer Engineering and Technology, Vol. 1, Issue
3, May 2012.
9. Kannan Govindan, Member IEEE and Prasant Mohapatra, Fellow
IEEE, Trust Computations and Trust Dynamics in Mobile Ad-
hoc Networks: A Survey.
10. Changbin Liu, Yun Mao, Mihai Opera, Prithwish Basu, Boon
Thau Loo, A Declarative Perspective on Adaptive MANET
Routing, August 2008.
11. Mamatha.T, Network Security for MANETs, International
J ournal of Soft Computing and Engineering, Vol.2, Issue 2, MAY
2012.
12. Muhammad Arshad Ali and Yasir Sarwar, Security Issues
Regarding MANET (Mobile Ad-hoc Networks): Challenges and
Solutions, Master Thesis, Computer Science, Thesis No: MCS-
2011-11, March 2011.
13. DSF for MANETs (Distributed Services Framework for Mobile
Ad-hoc Networks).
14. Gagandeep, Aashima, Pawan Kumar, Analysis of Different
Security Attacks in MANETs in Protocol Stack A-Review,
International J ournal of Engineering and Advanced Technology,
Vol.1, Issue 5, J une 2012.
15. WWRF/WG4/Ad-hoc Networking-Subgroup WhitePpaper,
Version 1.0, 17
th
J une 2002.
20
40
60
80
100
120
140
160
180
200
220
1 3 5 7 9
D
a
t
a

P
a
c
k
e
t

D
e
l
a
y

(
m
s
)
Number of Malicious Node
AODV
SR
ANODR
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
6


A Survey on Enroute Filtering Scheme in
Wireless Sensor Networks


P.Pritto Paul, K.Thejaswi,
Asst Professor, Velammal Engg. College, PG-Student, Velammal Engg. College,
Anna University, Chennai. Anna University, Chennai.
p.prittopaul@gmail.com tejaswi.kessamsetti@gmail.com


Abstract Wireless Sensor Networking is one of the
most prominent technology that is used in almost all real
time applications. WSN is used in the estimation of
temperature in Cyber Physical Network System(CPNS)
where the sensor nodes are deployed in hostile
environment. In this environment the sensor nodes sense
the data and forward the report to the base station. When
the report is being forwarded to the base station the
attacker may forge the data or may inject false data into
the report by compromising the sensor nodes in the
network this leads the base station to generate false
decision. The solution to overcome the False Data
Injection Attack is to implement the Enroute Filtering
Scheme in WSN. The Enroute Filtering is used to check
the correctness of the data before it is being forwarded to
the base station. In this paper some of the most efficient
Enroute Filtering Schemes for filtering false data have
been discussed with their advantages and disadvantages.
And also forwarding and filtering of data in Cluster based
environment which provides high security than other
filtering schemes have been discussed.


I. Introduction
WIRELESS sensor networks are expected to
interact with the physical world at an unprecedented
level to enable various new applications. However,
a large-scale sensor network may be deployed in a
potentially adverse or even hostile environment and
potential threats can range from accidental node
failures to intentional tampering. Due to their
relatively small sizes and unattended operations,
sensor nodes have a high risk of being captured and
compromised. False sensing reports can be injected
through compromised nodes, which can lead to not
only false alarms but also the depletion of limited
energy resource in a battery powered network.










The false data injection in a cyber physical
network system can be overcome by the formation
of clusters where the neighbor sensor node with
nearly similar properties will be organized into the
form of clusters. In the hierarchical network
structure each cluster has a leader, which is also
called the cluster head (CH). The sensor nodes
periodically transmit their data to the CH nodes.CH
nodes aggregate the data and transmit them to the
base station (BS) either directly or through the
intermediate communication with other CH nodes.
The BS is the data processing unit for the data
received from the sensor nodes.The Base Station is
fixed at a place in a stationary manner which is far
away from the all the sensor nodes .The function of
each CH,is to perform common functions for all the
nodes in the cluster, like aggregating the data before
sending it to the BS. In some way, the CH is the
sink for the cluster nodes, and the BS is the sink for
the CHs.
The advantages of cluster based environment
is: 1) supporting network scalability and decreasing
energy consumption through data aggregation
2) It can localize the route setup within the cluster
and thus reduce the size of the routing table stored
at the individual node.
. The main parameters included in clustering are:
Number of clusters, Nodes and CH mobility, Nodes
types and roles, Cluster formation methodology,
Cluster-head selection.











Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
7



In past different scheme have been proposed for
filtering false data in wireless sensor networks
where the data is transferred in a environment
where the sensor nodes are scattered. For example
inStatisticalEnroute Filtering Scheme[1],Interleaved
Hop-by-Hop Schemes[2] have the limitation of
node compromising where the false data can be
injected in order to generate the false reports.
In the paper we discuss about the
compromise resilient enroute filtering scheme
where the sensor nodes are organized into the form
of clusters. And the data is transferred to Base
station (sink) with the help of forwarding nodes
which act as an intermediate between the cluster
and the base station.
The rest of the paper is organized as follows:
1) a brief survey on existing filtering schemes.
overview of compromise resilient enroute filtering
scheme in cluster based environment in WSN. 2) a
survey on compromised resilient enroute filtering
scheme in WSN.3) Then, a detailed literature
survey on enroute filtering devised for WSNs is
provided along with comments on their prominent
and lacking feature.
II.RELATED WORKS
We discuss about existing filtering schemes,
which make use of MAC(message authentication
Codes) for transferring the data.
a) Statistical Enroute Filtering (SEF)[1] is the
most basic mechanism in which dense deployment
of large sensor networks takes place. To prevent
any single compromised node from breaking down
the entire system, SEF sends only limited of
amount of security information assigned to each
node, and depends on the collective decisions of
multiple sensors for false report detection. As a
report is forwarded through multiple hops toward
the sink, each intermediate node verifies the
correctness of the MACs carried in the report and
drops the report if an incorrect MAC is detected.
The disadvantage of SEF is probability of detecting
incorrect MACs increases with the number of hops
the report travels. SEF[1] and IHA[4] have the T-
threshold limitations. That is, if the adversary
compromises T nodes from different groups, they
could inject false data to generate the false report.







To overcome the threshold limitation and to
reduce the increasing number of compromised
nodes which we have seen in SEF. We come up
with LBRS
b) Location-Based Resilient Security (LBRS)[2]
approach which make use of two techniques:
location-binding keys and location-based key
assignment. In location based resilient scheme the
location of the sensors and sink is stationary by
which it can assign fixed key values for the sensor
in order to provide security. Based on its location, a
node stores one key for each of its local
neighboring cells and a few randomly chosen
remote cells. LBRS provides a solution to this
security problem, but it depends on the stationary of
the sink and the fixed routing model such that it
cannot work with mobile sinks and various routing
protocols. The disadvantage of LBRS is it relies on
special data dissemination protocol to confirm a
bean model.
c) Grouping-Based Resilient Statistical Enroute
Filtering (GRSEF)[4]scheme for filtering false
data. The GRSEF does not depend on sink
stationary. It improves the filtering efficiency by
dividing the sensor nodes into certain number of
groups(e.g.: T-groups) and assigns authentications
to the groups. GRSEF employees a multi-axis
division technique to overcome the threshold
limitation problem that we have seen in SEF[1] and
IHA[4]. In GRSEF, the Redundancy is increased to
achieve the robustness against this attacks but the
disadvantage of GRSEF is it has no resilience to the
selective forwarding attack and report disruption
attack .
All the early proposed scheme have the
disadvantage of T-threshold limitations, this
schemes does not adopt to dynamic topology, takes
long time for a network to become stable, sensor
nodes are scattered in a wireless networks were
there is no security for the transmission of data,
Requires node localization and takes a long time to
be stable. Inorder to overcome the disadvantages
our survey has come up with formation of clusters
in wireless sensor network where the sensor nodes
are grouped to form clusters.






Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
8


III. Preliminary
a) The Basics of Enroute Filtering:
The enroute filtering is technique used in
wireless networks with which the intermediate
nodes checks the correctness of the data that is
being travelled along the route from source to the
sink with the help of intermediate nodes present in
the network. The intermediate node not only checks
the correctness of the data but also can filter the
false data effectively. The intermediate nodes after
receiving the report checks whether it contain valid
T-MAC. The report with less number of T-MAC
will be dropped. If any false data which is not
filtered by the intermediate nodes will be detected
by the sink where it gets filtered. The sink acts as
the final defense that catches false reports not
filtered out by forwarding nodes.
b) System model of Enroute Filtering:











c) Security Model Of Enroute Filtering:

We consider a large sensor network field
where nodes are deployed. So after the network
initialization phase the sensor nodes forms into
groups and elect a cluster head based on different
parameters like remaining energy etc. Whenever
events of interest occurs in the terrain say if a tank
moves, all the cluster members near to the event
will sense the happening and report to their cluster
heads. On receiving the reports cluster head








aggregates them and sends a single copy of the
valid report to the base station through selected
report forwarding nodes. The selections of report
forwarding nodes are up to the underlying routing
protocols work . And also the selection parameters
are independent of the application. We assume that
there are attackers present within the terrain are
capable of monitoring the communication pattern
between the sensor members and the cluster head to
guess the message from the reports if intercepted.
We assume that each cluster contains at most t-1
compromised nodes, which may collaborate with
each other to generate false reports by sharing their
secret key information. The potential attacks which
we consider in our work DoS attacks. DoS attacks
include selective forwarding and report disrupt

d) Proposed System for Enroute Filtering:
The proposed system for enroute filtering is
based on cluster environment where the sensor
nodes are organized into groups(clusters).In cluster
based sensor nodes makes use of two keys
authentication key and check key instead of MAC
used in the existing systems. The sensor nodes
within the cluster are assigned with authentication
key and the forward nodes are assigned with check
key in order to provide additional security. The
security keys for the sensing node and the
forwarding node is assigned by the sink.
Different nodes present in different clusters are
assigned with different authentication keys. In this
way a compromised node present in one cluster will
not effect the nodes present in the other cluster.
Therefore this scheme achieves better resilience to
the increased number of compromised nodes. The
main advantage of this scheme is that it does not
depend on static routes and node localization.
This scheme mainly consists of two principles:
Management of authentication information: this is
used to assign the key values to the nodes present in
the network.
Management of data security: this is used to detect
and filter the false data.
The report that is being forwarded from one
cluster should contain: the encrypted measurement,
cluster key, sensor node ID, local ID of the node
from where it has been generated, authentication
information of the measurement generated by the
node.




Sensor
node
sink
Compromised
node(injection
of false data)
Intermediate
node should
check for correct
MAC
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
9



e) System model of proposed system:































FS1, FS2 - Forward Sensor Nodes
C1, C2, C3 - Clusters
CH - Cluster Head
S1, S2, S3 - Sensor Nodes

f) Algorithms used in Enroute Filtering:



Type Of Algorithm

Algorithm Usage

Advantages
1. Kar and Banerjees
algorithm
2. Greedy algorithm

Sensor-deployment in a network

1. Achieve Coverage
2. Achieve Connectivity

Distributed algorithm
Compute Support Weight(SW)
between the sensor nodes in a
network.
1. Construct Local Neighborhood Graph.
2. Construct Best Support Path

Veltri et al. algorithm

Distributed localized algorithm
1. To find an approximate minimal
exposure path.
2. Linear programming formulation for
minimal- and maximal-exposure paths is
obtained.

Kanan et al. algorithm

Polynomial time algorithms
1. compute the maximum vulnerability of a
sensor deployment to attack by intelligent
adversary .
2. To compute optimal deployments with
minimal vulnerability.
Clustering Algorithm 1. LEACH C Algorithm
2. Efficient Cluster Head Selection
Scheme For Data Aggregation
[EECHSSDA]
3. Hybrid Energy- Efficient
Distributed Clustering

Cluster Head Selection takes place based on
following clustering algorithm.

Greedy base-station
algorithm

1. For unidirectional antennas
2. For omnidirectional antennas

Forwarding of data to the base station
FS1
C2
CH
S2 S3
C1
CH
S1
S2

SINK
FS1
C3 CH
S1 S3
C2
CH
S2 S3
FS2
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
10
IV. Literature Survey

V. Conclusion
The clustering scheme achieves not only
high en-routing filtering probability but also
high reliability for filtering the injected false
data with multi-reports without depending on
static routes and node localization. Due to the
simplicity and effectiveness, the cluster based
scheme could be applied to other fast and
distributed authentication scenarios in wireless
network.

VI. References
1) F. Ye, H. Luo, S. Lu, and L. Zhang,
Statistical en-route filtering of injection false
data in sensor networks, IEEE Journal on

3) L. Yu and J. Li, Grouping-based resilient
statistical en-route filtering for sensor
networks, in Proc. of the 28th IEEE
International Conference on Computer
Communications (INFOCOM09), 2009, pp.
17821790.

4) S. Zhu, S. Setia, S. Jajodia, and P. Ning,
An interleaved hop-byhop authentication
scheme for filtering of injection false data in
sensor networks, ACM Transactions on
Sensor Networks (TOSN), vol. 3, no
data in sensor networks, IEEE Journal on
selected areas in communication,VOL.23, NO.
4, April 2005

2) Toward Resilient Security in Wireless
Sensor Networks Hao Yang, Fan Ye, Yuan
Yuan, Songwu Lu, William Arbaugh
5) N.Parashuram, Y.Sanjay sai raj, A.Sagar,
B.Uma An Active En-route Filtering Scheme
for Secured Data Dissemination in Wireless
Sensor Networks IJCSET |April 2012| Vol 2,
Issue 4,1102-1


Schemes for filtering
false data
Advantages Disadvantages
Statistical Enroute
Filtering

Dynamic topology
shared key mechanism
Threshold problems
No resilience to attacks.

Location-Based Resilient
Security

Avoid threshold limitations
Location-basedkey generation

Not applied to dynamic
topology
Require node localization

Grouping-Based
Resilient Statistical

Uses multi-axis division
technique
Avoid threshold limitations


Require node localization
Lower resilience to attacks.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
11
6) Clustering in Wireless Sensor Networks
textbook Basilis Mamalis, Damianos Gavalas,
Charalampos Konstantopoulos,
and Grammati Pantziou.

7)Algorithms For Wireless Sensor Networks
Sartaj Sahni and Xiaochun Xu
Department of Computer and Information
Science and Engineering, University of
Florida, Gainesville, FL 32611
{sahni,xxu}@cise.ufl.edu
September 7, 2004

8) A Random Perturbation-Based Scheme for
Pairwise Key Establishment in Sensor
Networks Wensheng Zhang and Minh Tran
Dept. of Computer Science, Iowa State
University, Ames, IA 50014, USA

9) Filtering Schemes for Injected False Data
in Wsn IOSR Journal of Computer
Engineering (IOSR-JCE) e-ISSN: 2278-0661,
p- ISSN: 2278-8727Volume 13, Issue 6 (Jul. -
Aug. 2013), PP 29-31.





Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
12
Estimation of Damping Torque for Small-Signal
Stability of Single Machine Infinite Bus System
M.Venkateswara Rao A.Krishna veni
Associate professor PG scholar
mvr.venki@gmail.com, krishnaveni203@gmail.com
Department of Electrical and Electronics Engineering, GMRIT, Rajam, A.P. INDIA.

Abstract This paper discusses the
Estimation of damping torque coefficient
for Small-signal stability of infinite bus
system. This damping torque coefficient is
used to identify the angle stability of a
system. Initially a mat lab coding was
utilized to generate the time domain
responses of rotor angle, rotor speed and
electromagnetic torque under various
loading conditions. The particle swarm
optimization (PSO) technique is then used
for accurate estimation of damping
torque coefficient. The mat lab coding
results using PSO, under various loading
conditions shows the effectiveness of the
proposed control strategy.
Index terms Damping torque
coefficient, Particle swarm optimization,
Small-signal stability, and Synchronizing
torque coefficient.
I. INTRODUCTION
The power system instability can be
demonstrated in many different ways
depending on the system configuration and
working mode. Since power system works
on synchronous generators, an essential
condition for system operation is that all
synchronous machines remain in
synchronism [1-3]. The small signal stability
is the ability of the power system to
maintain synchronism when subjected to a
small disturbance [1]. The operating
condition of the power system changes with
respect to time because of the dynamic
nature of the system. The rotor angle
stability can be analyzed from the
Synchronizing torque coefficient
S
K and
Damping torque coefficient
D
K . For stable
operation of the system, both synchronizing
and damping torque coefficients must be
positive.
The electromagnetic torque deviation is
split into Synchronizing torque and
Damping torques. The Synchronizing torque
is responsible for restoring the rotor angle
excursion and the Damping torque damps
out the speed deviations [4, 5]. In general
the synchronizing and damping torques are
expressed in terms of Synchronizing torque
coefficient
S
K and Damping torque
coefficient
D
K . These
S
K ,
D
K can be calculated
frequently for stability assessments.
Various computational techniques like
Simulation Annealing (SA) algorithm,
Evolutionary programming (EP), Genetic
Algorithm (GA) and Differential Evolution
(DE) are employed for optimization problem
[7, 8]. These techniques need more
parameters, high calculation time and not
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
13
easy to implement when compared to
particle swarm optimization. Particle swarm
optimization (PSO) was developed by
Kennedy and Eberhart. This has appeared as
a promising algorithm for handling the
optimization problems [20]. PSO is a robust,
non-linear and population based stochastic
optimal technique which can generate high
quality solutions within shorter calculation
time. The single-machine power system
modeling and small-signal stability studies
are carried out using Eigen analysis based
technique With PSO (particle swarm
optimization) optimal strategy. The
suggested control technique is based on
estimation of damping torque coefficient
D
K
of a synchronous machine from the time
responses of the rotor angle ) (t
r
A , rotor
speed ) (t A and electromagnetic torque
) (t T
e
A . Thus PSO has been chosen to
coordinate the operation in estimating
Damping torque coefficient
D
K for stability
analysis [17-19].
II. POWER SYSTEM MODEL
A simplified block diagram model of
the small signal performance is shown in
fig1 [1]. In this work, the proposed method
has been tested on a system comprising a
single machine connected to infinite bus
system through a transmission line.
Normally, for small signal stability study a
second-order model is considered for the
synchronous generator. The single machine
infinite bus system model is linearized at a
particular operating point to obtain the
linearized power system model. This model
is represented with some variables, such as
electrical torque, mechanical torque, and
rotor speed and rotor angle.

Figure1. Block diagram model of small
signal performance.
In the classical generator model, the
acceleration circuit dynamic equations are:
) (
2
1
r D e m
r
K T T
H t

A =
A
A
(1)
r
t

A =
A
A
0
(2)
Where
m
T ,
e
T are mechanical torque,
electromagnetic torque and
0
=
0
2 f .
From the block diagram, the following state-
space form is developed.
BU AX X + =





The elements of the system matrix A are
function of
D
K , H ,
T
X and the initial
operating conditions.
The perturbation matrix B depends on the
system parameters only. From the block
diagram of figure1, we have


(3)
m
r
s D
r
T
H H
K
H
K
dt
d
A
(
(

+
(

A
A
(
(

=
(

A
A
0
2
1
0
2 2
0

( )
(

A + A A = A
m r D S
T K K
HS S

2
1
0
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
14

And the characteristic equation is

(4)
Therefore, the damping ratio is

(5)
T
B
t
jX
j E E
I
) sin (cos
'

=
(6)
E d T
X X X + =
'
(7)

0
'
cos
T
B
S
X
E E
K =
(8)
III. SMALL SIGNAL STABILITY
ASSESSMENT USING MODAL
MATRICES
The power system experiences small
disturbances by small changes in loads.
Then the system will be driven to an infinite
state
0 0
) ( X t X = at time
0
t =0. The system
responds according to the state equations.
The linearized state equations can be used to
find the Eigen values
i
of the system matrix
A , where
i i i
j = are the distinct
eigenvalues corresponding to a set of right
and left eigenvectors. Here
i
is a damping
factor and
i
is Damped angular frequency.
The right and left Eigen vectors are
orthogonal and are usually scaled to be
orthogonal. Real eigenvalues indicates
modes which are aperiodic and complex
eigenvalues indicates modes which are
oscillatory. The uses of right and left
eigenvectors are for identifying the
relationshipbetween the states and the
modes is that the elements of the
eigenvectors are dependent on units and
scaling associated with the state variables.
To overcome this participation matrix which
combines the right and left eigenvectors are
used. The participation factor provides a
measure of association between the state
variables and the oscillatory modes.
IV. SMALL-SIGNALSTABILITY
ASSESSMENT USING
SYNCHRONISING AND DAMPING
TORQUES.
The electromagnetic torque (
e
T ) deviation
of a machine can be expressed as its speed (
) and angle ( ) deviations, which are
called damping and synchronizing torques.
The synchronizing and damping torques are
expressed in terms of its synchronizing
torque coefficient (
S
K ) and damping torque
coefficient (
D
K ). Then the electromagnetic
torque deviation will be expressed as:
) ( ) ( ) (
0
t K t K t T
S D e
A + A = A
(9)
Where ) (t
r
A =change in rotor speed
) (t A =change in rotor angle
V. OVER VIEW OF PARTICLE
SWARM OPTIMIZATION
PSO is one of the evolutionary based
optimization techniques [Fukuyama, 1999;
Kennedy and Eberhart, 1995]. This method
is introduced based on the research of bird
and fish flock movements behavior. Due to
its many advantages like its simplicity and
easy implementation, the algorithm can be
widely used in function optimization.
PSO ALGORITHM:
The particle swarm optimization consists
of n particles and the particles position
stands for the potential solution in D-
0
2 2
0 2
= + +
H
K
S
H
K
S
S D

0
2
2
1

H K
K
S
D
=
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
15
dimensional space. Each particle can be
shown by its current speed and position.
Particles change its position according to
its:
1. Current position
2. Current velocity
3. Distance between its current
position and pbest
4. Distance between its current
position and gbest
Velocity of each particle can be modified
based on the following equation.
) ( ) (
2 2 1 1
1 k
id
k
id
k k
id
k
id
k k
id
k
id
x gbest r c x pbest r c wv v + + =
+

By using velocity equation, a certain
velocity which gradually gets close local
best and global best can be calculated. The
current position of the particle can be
modified by the following equation.
1 1 + +
+ =
k
id
k
id
k
id
v x x

Where, pbest represents the D-dimension
quantity of the individual i at its most
optimist position at its k times and gbest
represents the D-dimension quantity of the
individual i at its most optimist position at
its k times. The speed of the particle at its
each direction is confined in between
vdmix and +vdmax. If vdmax is too big,
solution is far from the best and if vdmax is
too low, it means that the solution will be
local optimism. C1, C2 represents speeding
figures which lies between 0 to 2. r1, r2
represents random fiction, and 0-1 is a
random number.
VI. ESTIMATION OF LOWER AND
UPPER LIMITS OF DAMPING
TORQUE COEFFICIENT.
For the linearized system model presented in
figure1, the eigenvalues of the local system
can be evaluated. The proposed method is
aiming to search for the optimal damping
torque coefficient, such that the damping
ratio can be maximized.


Where =damping ratio.
For stable operating condition of the system
the Damping ratio must be in [0.4, 0.7].
Hence the Corresponding values of
D
K will
be [35.712, 63.125]. The control parameters
can be tuned through the optimization
algorithm. The proposed algorithm will be
as follows.
VII. IMPLEMENTATION OF PSO FOR
OPTIMAL ESTIMATION OF
DAMPING TORQUE
Step1. Read the system input data, PSO
parameters.
Step2. Initialize population of particle (
D
K )
with random velocities and positions.
Step3. Evaluate fitness values using the
objective function.


Step4. Each particle has its own best
position called local best and the best
position among all the particles is called
global best.
Step5. Update the velocity of particle using
) ( ) (
2 2 1 1
1 k
id
k
id
k k
id
k
id
k k
id
k
id
x gbest r c x pbest r c v v + + =
+

0
2
2
1

H K
K
S
D
=
0
2 2
1

H K
K
S
D
=
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
16
Check the updated velocity, within the limits
or not
max min
i i i
v v v s s

Step6. Update the position of particle using
1 1 + +
+ =
k
id
k
id
k
id
v x x

Step7. Evaluate fitness values of the new
particles. Update the local best values as
current fitness values if these values are
better than previous values, update it. Then
find new global best values.

Step8. Repeat the procedure until the
stopping criteria is reached.
VIII. RESULTS AND DISCUSSIONS
In this work the optimal values of
D
K are
obtained using PSO and the rotor speed,
rotor angle and electromagnetic torque
responses are generated and are compared
with those obtained in [1]. In which the
damping torque coefficient is chosen
randomly [-10, 0, 10]. In addition the same
responses are generated for different loading
conditions using mat lab coding. It is
observed that the steady state stability of the
system is improved. These responses are
shown in figure2-a to figure4-c. The rotor
responses are obtained for various
conditions:
1. Nominal operating condition (P=0.9,
Q=0.3).
2. Light operating condition (20% of
the nominal values).
3. Heavy operating condition (50%
higher than the nominal operating
condition).
The following responses show the rotor
characteristics under various loading
conditions.
The following responses show the rotor
characteristics under various loading
conditions.

Figure 2-a. Rotor speed response for
P=0.9,Q=0.3.

Figure 2-b. Rotor Angle response for P=0.9,
Q=0.3.


Figure 2-c. Torque response for P=0.9,
Q=0.3.


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
17


Figure 3-a. Rotor speed response for P=1.35,
Q=0.45.

.
Figure 3-b. Rotor Angle response for
P=1.35, Q=0.45

Figure 3-c. Torque response for P=1.35,
Q=0.45.

Figure 4-a. Rotor speed response for P=0.72,
Q=0.24


Figure 4-b. angle response for P=0.72,
Q=0.24.

Figure 4-c. Torque response for P=0.72,
Q=0.24.



Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
18

CONCLUSION
In this project the steady state performance
improvement is obtained by the accurate
estimation of damping torque coefficient
using particle swarm optimization
technique. The matlab programming results
using PSO under various loading conditions
shows the effectiveness of the proposed
technique. Compared to normal operating
conditions under heavy and light operating
conditions the peak over shoots are very
high which are reduced using PSO
technique. The effectiveness of the proposed
controller is to provide good damping of low
frequency oscillations. It can be concluded
that the proposed PSO controller extends the
power system stability limit by enhancing
the system damping.
APPENDIX:
Input data:
1. Generator parameters:

85 . 0 , 003 . 0 , 3 . 0
, 76 . 1 , 81 . 1 , 8 , 5 . 3
'
'
0
= = = =
= = = =
sq sd a d
q d d
K K R X
X X T H

2. Transmission line parameters:
. 15 . 0 , 65 . 0 , 0 = = =
L e e
X X R

Table1. (Lower, Upper limits of control
parameter):
PARAMETER
D
K
Lower limit 35.712
Upper limit 63.125




Table2.PSO parameters:

Table3 (Loading conditions):

Table4. (Optimal value of control
parameter)



REFERENCES
[1] Kundur, P.: power system stability and
control, McGraw-Hill, 1994.
[2] Hsu Y.Y., Chen, C.L.:Identification of
optimum location for stabilizer
Application using participation
factors. In: IEEE Proc. Gen., Trans. &
Distr., Pt. C. VOL. 134. No.3.1987. P.
238-244.
[3] Demello F.P., concordia. C,Concepts
of synchronostability as affected by
excitation control. In: IEEE Trans.
Population size 10
Maximum number of generations 50
Acceleration coefficients(C1,C2) 1.4
Inertia weight 1
S.no Cases P(p.u) Q(p.u)
1 Nominal operating
condition

0.9

0.3
2 Heavy operating
condition (50% higher
than the nominal load)

1.35

0.45
3 Light operating
condition(20% of the
nominal operating
condition)

0.72

0.24
D
K 57.8
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
19
Power Apparatus and Systems,
Vol.PAS-88, No.4.April 1969, p.316-
329.
[4] Feilat, E.A., Younan, N., Grzybowski,
S.: Estimating the synchronizing and
damping torque coefficients using
Kalman filterin. In: Electric power
Systems Research, Vol.52, No.2,
1992.p.145=149.
[5] Hassan Ghasemi, Claudio canizares.
On-line Damping Torque estimation
and Oscillatory Stability Margin
Prediction . In IEEE Transactions on
Power Systems Vol. 22, No.2, May
2007.
[6] Gurunath Gurrala, Indraneel Sen.:
Synchronizing and damping torque
analysis of nonlinear voltage
regulators. In: In IEEE Transactions on
Power Systems Vol. 26, No.3, and May
2011.
[7] Y.U, Y.N.: Electrical Power System
Dynamics, Academic Press, 1983.
[8] N.A.M Kamari, I. Musirin, M.M.
Othman, Improving Power System
Damping Using EP Based PI
Controller. PEOCO 2011, June 2011,
PP.121-126.
[9] T.K. Rahman, Z.M. Yasin, W.N.W.
Abdullah, Artificial-Immune-Based
for Solving Economic Dispatch in
Power System. PEcon 2004, NOV
2004, PP.31-35G.
[10] M.Hunjan, G.K. Venayagamoorthy,
Adaptive Power System Stabilizer
Using Artificial Immune System.
ALIFE07, April 2007, PP.440-447.
[11] F.P. Demello, C. Concordia, Concept
of Synchronous Machine as Affected
By Excitation Control. In: IEEE trans.
Power system apparatus, PSA-87. 1968.
PP.835-844.
[12] Y.L. Abdel-Magid, A.H. Mantawy
Robust Tuning Of Power System
Stabilizers in Multimachine Power
Systems. In: IEEE Trans. Power
System, Vol. 15, No. 2, May2000.
[13] Kaith, T.: Linear Systems. Prentica
Hall, NJ, 1980.
[14] T.C. Hsia, system identification: least
square methods. Lexington, MA:
Lexington Books, D.C. Health and
Company, 1977.
[15] M.J. Gibbard, Co-Ordinated Design of
Multimachine Power System Stabilizers
Based On Damping Torque Concepts.
In: IEEE Proceedings, Pt. C, Vol. 135,
No. 4, July 1988, and PP.1276-284.
[16] F. Glover, Artificial Intelligence
Heuristic Frameworks and Tabu
Search. Managerial and Decision
Economics, Vol. 11, PP.365-375, 1990.
[17] Shaltout, A., Feilat, E.A.: Damping
and Synchronizing Torque
Computation in Multimachine Power
Systems. In: IEEE Trans. On Power
Systems, Vol, PWRS-7, No.1, February
1992, P. 280-286.
[18] Hiroshi Suzuki, Soichi Takeda, Yoshizo
Obata: An Efficient Eigenvalue
Estimation Technique for
Multimachine Power System Dynamic
Stability Analysis. In: Electrical
Engineering in J apan, Vol. 100, No.5,
2007, PP.45-53.
[19] F.P. Demello and C. Concordia,
Concepts of Synchronous Machine
Stability as Affected By Excitation
Control. In: IEEE Trans. Power
Apparatus and Systems, Vol. PAS-88,
No. 4, PP. 316-329, April 1969.
[20] J. Kennedy and R.C. Eberhart,. Particle
Swarm Optimization. Proceeding of
IEEE international conference of
Neural Networks, Vol. 4, 1995, PP.
1942-1948.


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
20


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
21







Abstract
Irrelevant features, at the side of redundant features, strictly
have an effect on the correctness of the knowledge machines.
Thus, feature set selection have to be compelled to be able to
determine and take away the maximum as much of the unrelated
and redundant knowledge as feasible. With this intention
choosing a subset of features with relation to the target notions,
feature set selection is an efficient alternative way for reducing
spatial property or dimensionality (ex: subset), removing
unrelated data (Ex: irrelevant data), increasing learning
accuracy, and generating Qualitative result. Feature selection
involves classifying a set of the foremost relevant features that
generates appropriate outcome as the original entire set of
features. Several feature set selection techniques are planned and
studied for machine learning applications. By this criterion, an
Enhanced fast clustering-based feature selection algorithm,
EFAST, is employed during this paper.
The EFAST algorithmic rule works in 2 steps.
In the starting step, features are classified into clusters by
exploitation graph-theoretic clustering approaches.
In the second step, the foremost relevant representative feature
that is powerfully associated with target categories is chosen from
every cluster to form a set of features.
Features in dissimilar clusters are comparatively autonomous
and the clustering-based strategy of EFAST includes a high
chance of generating a set of valuable and autonomous features.

Keywords-EFAST Algorithmic rule, Correlations, Feature set
Selection, and Graph based clustering

1 INTRODUCTION
The use of feature selection can develop
accurateness, relevancy, applicability and be aware of
a learning method. For this reason, several ways of
automatic feature selection are developed. Some of
these ways are based on the search of the features that
enables the data set to be measured consistent. In an
exceedingly search problem we usually tend to
evaluate the search states, in the case of feature
selection we measure the promising feature sets.
Feature set selection is a very important subject when
preparing classiers in Machine Learning (ML)
issues. Selection of Feature set is an efficient system
for dimensionality reduction, elimination of
inappropriate knowledge, rising learning
accurateness, and improving result unambiguousness.
Based on the minimum spanning tree methodology,
we propose an EFAST algorithmic rule. The
algorithmic rule is a 2 step method, in that features are
separated into clusters by way of using graph
theoretic clustering means. Within the succeeding
step, the frequently used representative feature that is
robustly related to target categories is specific from
every cluster to structure the ultimate subset of
features. Features in distorted clusters are
comparatively autonomous. The clustering-based
theme of EFAST includes a high risk of designing a
set of constructive and autonomous features. In our
planned EFAST algorithmic rule, it needs the building
of the minimum spanning tree (MST) from a
subjective comprehensive graph. The separation of
the MST into a forest by means of each tree signifying
a cluster and the collection of representative features
from the clusters.

The planned feature set selection algorithmic rule
EFAST was tested and the investigational results
demonstrate that, evaluated with different varied
forms of feature set selection algorithms, the
projected algorithmic rule not solely decrease the
amount of features, but also advances the
performances of the famed varied forms of classifiers.

The results, on publically obtainable
real-world high dimensional image, microarray, and
text knowledge, established that EFAST not only
produces smaller sets of features however improves
the performances of the classifier.
In our study, we tend to apply graph theoretic
clustering schemes to features. In exacting, we tend to
accept the MST based clustering algorithms, since
they do not imagine that knowledge points are
classified around centres or separated by a normal
geometric curve and are widely used in training.
Based on the MST method, we tend to suggest an
Enhanced Fast clustering-bAsed feature Selection
algoriThm (EFAST).
Good feature set is one that contains features
extremely correlative with the target, so far
uncorrelated with one another. In the above planned
An Enhanced Feature Selection for
High-Dimensional Knowledge
L.Anantha Naga Prasad
1

K.Muralidhar
2
1
M.Tech, Computer Science and Engg, Anantha Lakshmi Institute of Technology&Sciences,JNTUA, Andhra Pradesh, India
2
Assistant Prof, Department of CSE, Anantha Lakshmi Institute of Technology&Sciences, JNTUA, Andhra Pradesh, India
1
mail id: naga4all16@gmail.com
2
mail id: muralidhar.kurni@gmail.com


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
22




terms this system is an efficient fast filter method
which may categorize relevant features as redundancy
among relevant features without pair wise correlation
study, and repeatedly chooses features which exploit
their mutual information with the category to expect
provisionally to the reply of any feature formerly
elected. In contrast to from these algorithms, our
projected EFAST algorithmic rule utilizes clustering
based methodology to select features.
2 RELATED WORK
2.1 EXISTING SYSTEM
In the past approach there are many algorithms that
illustrate a way to maintain the knowledge into the
database and how to retrieve it quicker, however the
difficulty here is no one cares about the database
maintenance with ease manner and safe methodology.
A Distortion algorithmic rule, that creates a personal
space for every and each word fromthe already
elected transactional database, those are put together
named as dataset, which is able to be acceptable for a
collection of exacting words, however it'll be
problematic for the set of records. An inference
algorithmic rule build propagation to the higher than
downside, and cut back the issues occurred within the
existing distortion algorithmic rule, however here
conjointly having the matter known as knowledge
overflow, once the user get confused then they will
never get the knowledge back. The embedded ways
incorporate feature choice as a locality of the training
method and are sometimes specific to given learning
algorithms and as a result could also be improved than
the opposite 3 teams. Typical machine learning
algorithms like decision trees or artificial neural
networks are samples of embedded ways. The
wrapper techniques use the analytical accuracy of a
planned algorithmic rule to decide the goodness of the
actual subsets, the accurateness of the learning
algorithms is usually high. But the simplification of
the chosen feature is restricted and the process
problem is high. The filter ways are autonomous of
learning algorithms, with fine generality. Their
process complexness is low, however the accuracy of
the learning algorithms is not assured. The hybrid
techniques are a mix of filter and wrapper ways by
employing a filter methodology to diminish search
space which will be measured by the succeeding
wrapper. They primarily target on grouping filter and
wrapper ways to attain the most effective potential
performance with a selected learning algorithmic rule
with similar time complexness of the filter ways.
Hierarchical clustering has been implemented
in word choice within the context of text
classification. And it is noise-tolerant and strong to
feature communications, additionally as being
relevant for binary or continuous knowledge solely.
However, it does not discriminate between redundant
features and low numbers of training instances of the
algorithmic rule.
Relief-F could be a feature choice strategy
that chooses cases randomly, and altered the weights
of the feature importance based on the closest
neighbor. By its qualities, Relief-F is one in all the
foremost prosperous methods in feature choices.
2.2 Disadvantages of Existing System
The simplification of the chosen features is
restricted and hence the complexness is large.
Accurateness is not guaranteed.
Ineffective at deleting redundant features
Performance associated problems
Security problems
So the attention of our new system is to boost the
outturn for any basis to eliminate the knowledge
security lacks there in and build a more recent system
outstanding handler for handling data in an
economical manner.
2.3 Proposed System
In this proposed system, The Enhanced fast
clustering-based feature selection algorithmic rule
(EFAST) works in 2 steps. In the first step, features
are classified into clusters by exploitation
graph-theoretic clustering ways.
In the second step, the foremost relevant
representative feature that is powerfully associated
with target categories is chosen from every cluster to
form a set of features.
Features in dissimilar clusters are comparatively
autonomous and therefore the clustering-based
strategy of EFAST includes a high probability of
generating a set of valuable and autonomous features.
Inside this paper we tend to generate correlations
for high dimensional knowledge supported EFAST
algorithmic rule in four steps.
1. Removal of unrelated features:
If we choose a Dataset 'D' with m features F={F
1
,
F
2
... F
n
} and class C, mechanically features are
obtainable with target relevant feature. The
simplification of the chosen features is restricted
and the process complexness is huge. If F
i
is
relevancy to the target C if there exists some s
i
, f
i
,
and c specified for probability p(S
i
=s
i
, F
i
=f
i
) >0,
p(C=c | S
i
=s
i
, F
i
=f
i
)p(C=c | Si = si) otherwise
feature F
i
is an unrelated feature.
2. T-Relevance, F-correlation calculation:
If (Fi F) then target notion C is treated as
T-Relevance. If (Fi, Fj F ^ij) is named
F-correlation. T-Relevance among a feature and
the target notion C, the correlation F-Correlation
between a combine of features, the feature
redundancy F-Redundancy and therefore the
representative feature R-Feature of a feature
cluster will be outlined.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
23




3. MST construction by fuzzy logic :
We adopt the minimum-spanning tree (MST)
clustering way in competence view. during
this method we calculate a neighborhood
graph of occurrences, then take away any
edge in the graph that is a lot shorter/longer
(by fuzzy logic) than its neighbors.
4. Relevant feature calculation:
When removing all the unnecessary spare
edges, a Forest will be obtained. In that every tree
represents a cluster. Finally it contains feature set and
then calculates the accurate/relevant feature.
2.4 Problem Definition
Many algorithms that illustrate a way to maintain the
knowledge into the database and the way to retrieve it
quicker, however the matter is no one cares about the
database maintenance with ease manner and safe
methodology. The systems like Distortion and
congestion algorithmic rule, which makes an
individual space for every and each word from the
already elected transactional database, those are put
together known as dataset, which is able to be
appropriate for a collection of exacting words,
however it will be troublesome for the cluster of
records, once the user get confused then they will
never get the data back. The wrapper ways use the
analytical accuracy of a predetermined learning
algorithmic rule to verify the goodness of the chosen
subsets, the correctness of the learning algorithms is
usually high. Their computational difficulty is low,
however the correctness of the learning algorithms is
not assured. An EFAST algorithmic rule analysis has
targeted on sorting out relevant features. A famed
example is Relief which weighs every feature in line
with its ability to discriminate instances below
completely different targets supported distance-based
criteria task. Though, Relief is unproductive at
removing redundant features as too analytical
however extremely correlative features are seemingly
each to be extremely weighted. Relief-F expands
Relief, permitting this methodology to work with
strident and incomplete data sets and to cope with
multiclass issues, however still unable to
acknowledge redundant features.
2.5 Literature Review
Literature survey is the most vital step in code
development procedure. Before finding the tool it is
essential to decide the time facet, financial systemand
company strength. Once these things are satisfied,
then the subsequent step is to decide which operating
system and language can be used for developing the
tool.
Once the programmers commence building the tool
the programmers require set of external support. This
maintenance will be obtained from programmers,
from books or from websites. Before building the
systemthe above thought are taken into account for
developing the projected system. The core part of the
developing projected sector considers and totally
survey all the necessary needs for creating the project.
For every project Literature survey is the most vital
sector in code development procedure. Preceding to
developing the tools and the associated planning it is
necessary to decide and survey the time facet,
resource constraint, man power, financial system, and
company strength. Once these items are fulfilled and
totally reviewed, then the subsequent step is to make
a decision concerning the code specifications within
the relevant system such as what kind of operating
systemthe project would require, and what are all the
essential code are required to proceed with the
subsequent step such as developing the tools and the
associated operations.
3. FUZZY BASED FEATURE SET SELECTION
ALGORITHMS
EFAST Algorithmic rule is a classic algorithm for
frequent item set mining and association rule learning
over transactional databases. This EFAST
algorithmic rule internally contains an algorithmic rule
known as Apriori, which progresss by discovering the
frequent individual things in the database and
enlarging themto larger and well-built item sets as
long as those item sets seem sufficiently frequently in
the database. The common itemsets confirmed by
Apriori will be accustomed to determine association
rules that highlight general trends in the database.
3.1 Feature set selection algorithmic rule
In machine learning, statistics feature selection called
as variable selection or attribute selection or variable
set selection. Is the procedure of choosing a set of
relevant features to be used in model creation. The
central hypothesis employing a feature selection
technique is that the data contains several redundant
or irrelevant features. Redundant features are those
which supply no more information than the presently
specific features, and irrelevant Features offer no
helpful information in any background. Feature
selection ways are a set of the additional general field
of feature extraction. Feature extraction creates new
features from functions of the novel features, whereas
feature selection returns a set of the features. Feature
selection techniques are usually employed in domains
wherever there are several features and relatively few
samples (or knowledge points).
Feature selection techniques supply 3 main profits
once constructing analytical models:
Improved model interpretability,
Shorter training times,
Improved generalisation by reducing over fitting.
Feature selection is also useful as a part of the data
analysis method, as shows which features are vital for
prediction, and the way these features are connected.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
24




3.2 Definitions
In learning Machines [11], [15] Suppose Fto be the
complete set of features, i F be a feature,Si
=F{Fi} and Si Si. Let si be a value-assignment
of all features in Si, i a value-assignment of
featureFi, and ca value-assignment of the target
concept C. The definition will be formalized as
follows.
Definition: (Relevant feature) Fi has relevancy to
the target concept C if and only if there exists some
si, i and c, such that, for probability (Si =si,Fi
=i)>0,p(C=c Si =si,Fi =i) p(C=c Si =si). or
else, featureFi is an irrelevant feature. There are 2
sorts of relevant features due to different Si:
(i) if Si =Si, we will recognize that Fi is directly
relevant to the target concept
(ii) if Si Si, we could get that p(CSi,Fi)=p(CSi).
Definition: (Markov blanket) The definitions of
Markov blanket and redundant feature are introduced
as follows, correspondingly. let Hi F(Fi Hi ),Hi
is assumed to be a Markov blanket for Fi if and only if
p(FHi{Fi},C Fi,Hi)=p(FHi{Fi},C Hi).
Definition: (Redundant feature) let S be a
collection of features, a feature in S is redundant if and
only if it has a Markov Blanket withinS. Relevant
features have strong correlation with target concept
so are always essential for a finest set, while
redundant features are not since their values are fully
correlative with one another.
Definition: symmetric uncertainty (Su) is derived
by normalizing the entropies of feature values or
target categories. SU is the evaluation of correlation
between either two features or a feature and the target
concept.
In existing system we tend to calculate SU as
SU(X,)=20oin(X) / E(X)+E()
Where E(X)= xX p(x)log2p(x)
0oin(X)=E(X)E(X)
0oin(YX)=E()E(X)
E(X)= y p(y) xX p(xy)log2p(xy).
In our paper we are proposing SU as
SU(X,)=20oinRatio(X) / E(X)+E()
Where E(X)= xX p(x)log2p(x)

Intrinsic Information is the entropy of distribution of
instances into branches.

E(X)= y p(y) xX p(xy)log2p(xy).
Definition: (T-Relevance) The relevance between
the featureFi Fand the target concept C is consigned
as The T-Relevance of Fi and C, and denoted by
SU(Fi,C). If SU (Fi,C) is >a predetermined threshold
0, we say that Fi is a strong T-Relevance feature.

Definition: (F-Correlation) The correlation
between any pair of featuresFi andF] (Fi,] Fi]) is
named as the F-Correlation of Fi andF], and denoted
by Su(Fi,F]).
Definition: (F-Redundancy) Let S ={F1, F2... Fi...
Fk<F} be a cluster of features. if F] S,
Su(F], C) Su(Fi, C) Su(Fi, F]) > Su(Fi, C) is
often corrected for every Fi S (i ]), then Fi are
Redundant features with reverence to the given F].

Definition: (R-Feature) A feature Fi S ={F1, F2,
..., Fk} (k < |F|) is a representative feature of the
cluster S ( i.e. Fi is a R-Feature ) if and only if,
Fi =argmaxF]S Su(F], C).
By this we will say
a) Irrelevant features have no/weak correlation with
Target concept
b) Redundant features are assembled in a cluster and a
representative feature will be taken out of the Cluster.

3.3 EFAST Algorithm by Gain Ratio

Algorithm: EFAST
Inputs: D(f1,f2,fm, C) given data set
Output: S Feature Selection
// irrelevant feature removal
For i=1 to mdo
T-Relevance=SU(F
i
, C)
//SU is calculated based on Gain Ratio
If T-Relevance > then
// here is the Threshold value
S=S U {F
i
}
// MST Construction
G=NULL
For each pair of features { f
i
, f
j
} S do
F-Correlation =SU(f
i
, f
j
)
Add f
i
and/or f
j
to G with the F-Correlation as its
weight.
Minspantree =Prim(G);
//using Prim algorithm to construct MST
// clustering and Enhanced Feature Selection
Forest =Minspantree
For each edge E
ij
Forest do
If SU(f
i
, f
j
) <SU(f
i
,C) ^SU(f
i
, f
j
) <SU(f
i
,C) then
Forest=Forest - E
ij

S=
For each tree T
i
Forest do
F
j
R
=argmax F
k
Ti SU(F
k
, C)
S=S U { F
j
R
}
Return s





Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
25




3.4 About Gain Ratio
Information is measured in terms of bits
Given a probability distribution, the
info requisite to predict an event is the
distributions entropy
Entropy offers the knowledge needed
in bits (this will involve fractions of
bits!)
We can calculate entropy as
n n n
p p p p p p p p p log log log ) , , , entropy(
2 2 1 1 2 1


Gain ratio: a modification of the information gain
that reduces its bias on high-branch attributes
Gain ratio ought to be
Large when knowledge is equally
spread
Small when all knowledge belong to 1
branch
Gain ratio takes range and size of branches
under consideration when selecting an
attribute
It corrects the information gain by
taking the intrinsic information of a
split under consideration (i.e. what
quantity data can we have to be
compelled to tell that branch and
instance belongs to)
We can calculate Gain Ratio as

Intrinsic information: entropy of distribution
of instances into branches
Gain ratio (Quinlan86) normalizes info gain
by:


Ex: gainratio (Attribute) =gain (Attribute)
Intrinsicinfo (Attribute)
Ex: gainratio (Id) =0.94/3.8=0.24

4 Frame Work
Our projected feature set selection framework
structure involves irrelevant feature removal and
redundant feature elimination by using fuzzy logic.
It offers internal logical schema to form clusters with
the assistance of EFAST Algorithmic rule. Smart
feature subsets contain features extremely correlative
with the class, yet unrelated with each other. [20]
Frame work Analysis, it involves (i) building the
minimum spanning tree (MST) from a weighted
complete graph (ii) the partitioning of the MST into a
forest with every tree representing a cluster and (iii)
the selection of representative features from the
clusters.



Fig. 1: Framework of the Fuzzy Based

The outline of Frame work is characterized as



























Fig. 2: EFAST Architecture

Data Sets
We Use Gain Ratio for Effectiveness
Irrelevant Feature Removal
MST Construction
Clustering of MST
Enhanced Feature
Selection
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
26




5 Algorithmic Rule Analyses
The planned EFAST algorithmic rule rationally
consists of 3 steps:
(i) Removing irrelevant features,
(ii) Constructing a MST from relative ones,
(iii) Separating the MST and selecting Representative
features.
For a data set with m features F ={F1, F2... Fm}
and class C, we tend to compute the T-Relevance
Su(Fi, C) value for every feature
Fi (1 i m).

In the first step. The features whose Su(Fi, C) values
are larger than a predefined threshold 0 contain the
target-relevant feature subset F ={F1, F2... Fk} (k
m).
In the second step, we 1st calculate the
F-CorrelationSu(Fi, F]) value for every pair of
features Fi and F](Fi, F] F i ]). Then,
viewing features Fi and F] as vertices and Su(Fi,
F]) (i ]) as the weight of the edge between vertices
Fi and F] , a weighted complete graph 0 =(I,E) is
built wherever I ={Fi | FiF i [1, k]} and E =
{(Fi, F]) (Fi, F] F i, ] [1, k] i ]}.

In the third step, we 1st eliminate the edges E ={(Fi,
F]) (Fi, F] F i, ] [1, k] i ]}, whose
weights are smaller than both of the T-Relevance
Su(Fi, C) and Su(F], C), fromthe MST. Each
removal ends up in 2 disconnected trees I1 and I2.

The complete graph 0 reflects the correlations among
all the target-relevant features. If graph 0 has k
vertices and k(k1)/2 edges. For this we build a
MST, which connects all vertices such that the sum of
the weights of the edges is the minimum, using the
famous Primalgorithmic rule [14].The weight of edge
(Fi, F]) is F-Correlation Su(Fi, F]).

If (Fi, F] I (I)), Su(Fi, F]) Su(Fi, C)
Su(Fi, F]) Su(F], C) this property assurances
the features in I (I) are redundant.

Suppose the MST shown in Fig.3 is generated from a
complete graph 0. So as to cluster the features, we
1st pass through all the six edges, and then decide to
remove the edge (F0, F4) as its weight Su(F0, F4) =
0.3 is smaller than both Su(F0, C) =0.5 and Su(F4,
C) =0.7. This makes the MST is clustered into 2
clusters denoted as I (I1) and I (T2).

From Fig.3 we identify that Su(F0, F1) > Su(F1, C),
Su(F1, F2) > Su(F1, C) Su(F1, F2) >Su(F2, C),
SU(F1, F3) > SU(F1, C) SU(F1, F3) >SU(F3, C).
We also recognized that there is no edge exists
between F0,F2 and F0,F3 and F2,F3.

Fig.3: Clustering Step


After removing all the redundant edges, a forest is
achieved. Every tree Tj Forest denotes a cluster that
is signified as V(Tj). As examined above, the features
in every cluster are redundant, thus for every cluster
V (Tj) we tend to like a representative feature F
j
R

whose T-Relevance SU (F
j
R,
C )is the high. All F
j
R
(j =
1...|Forest|) comprise the final feature subset F
j
R
.
5.1 Time complexity analysis.
The computation of SU values for T-Relevance and
F-Correlation, that has linear complexness in
provisos of the number of instances in a given data
set.

The first part of the algorithmic rule includes a linear
time complexness O (m) as a result of the number of
features m.

If k (1 k m) features are elected as relevant ones in
the 1st half, when k =1, only 1 feature is chosen.
Therefore, there is no need to continue the rest parts
of the algorithmic rule, and therefore complexness is
O(k).

If 1 < k m, the second part of the algorithm first of
all constructs a complete graph from relevant features
and therefore complexness is O(k
2
), and then MST
complexness is O(k
2
). The third part partitions the
MST and chooses the representative features with the
complexness of O(k).

Thus when 1 < k m, the complexity of the
algorithmic rule is O(m+k
2
). This means when k m,
EFAST has linear complexness O(m), where as the
worst complexness is O(m
2
) when k =m. However, k
is drastically set to be lower bound of m*lg m in the
implementation of EFAST. Therefore the
complexness is O (m * lg
2
m), which is usually less
than O (m
2
) since lg
2
m< m. it makes EFAST has an
enhanced runtime performance with high dimensional
knowledge.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
27




6 knowledge Source
To evaluate the performance and effectiveness of
EFAST algorithmic rule we are using publicly existing
data sets. The numbers of features of the data sets
vary from 35 to 49152 with a mean of 7874. The
dimensionality of the 53.3% data sets exceeds 5000,
of which 26.6% data sets have more than 10000
features. The data sets cover a collection of
application domains like images, text and microarray
data categorization.

Table 1: sample benchmark data sets
__________________________________________
Data ID Data Name F I T Domain
__________________________________________
1 chess 37 3196 2 Text
2 mfeat-fourier 77 2000 10 Image, Face
3 coil2000 86 9822 2 Text
4 elephant 232 1391 2 Microarray, Bio
5 tr12.wc 5805 313 8 Text
6 leukemia1 7130 34 2 Microarray, Bio
7 PIX10P 10001 100 10 Image, Face
8 ORL10P 10305 100 10 Image, Face
__________________________________________

7 Results and Analysis
7.1 Main Form


Now we have to upload any dataset i.e.
Here we are uploading chess data set.








7.2 Loading Data Set

7.3 Calculating Entropy, Gain And GainRatio


7.4 calculating T-Relevance and Relevant
Attributes



Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
28




7.5 calculating F correlation


7.6 Generating MST


7.7 Relevant Feature Calculation



7.8 MST using InformationGain

7.9 clustering Inf.Gain Based MST

7.10 MST using Gain Ratio

7.11 clustering Gain Ratio Based MST

By Observing 7.8 to 7.11 we can say that Gain Ratio
makes effective clustering than Information Gain as
proposed in the existing system.
We can represent the same thing in graphical form
also. In 7.12 we are showing the graphical chart
illustration of information gain vs. gain ratio.

From the above discussions and experimental
results we can conclude that Gain Ratio gives
Enhanced Feature Selection than Information Gain.









Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
29




7.12 Graphical Representation of
Inf.Gain Vs Gain Ratio
Analysis on Chess Data Set
Construction of MST using Inf.Gain


Construction of MST using Gain Ratio


Construction of MST Clustering using Inf.Gain


Construction of MST Clustering using Gain Ratio







Construction of MST Clustering using Inf.Gain


Construction of MST Clustering using Gain Ratio



8 Conclusions
In this paper, we have presented a completely unique
relevant clustering-based EFAST algorithmic rule for
high dimensional knowledge. The algorithmic rule
involves a) removing unrelated features, b) building a
MST from relative ones, and c) divisioning the MST
and selecting representative features. a cluster
consists of features. Each cluster is referred as a single
feature and thus dimensionality is severely reduced.
The proposed algorithm gets the most fractions of
chosen features, the enhanced runtime, and the finest
classification accuracy. For the future work, we plan
to find out different types of correlation calculations,
relevance measures and revise some formal properties
of feature space.
ACKNOWLEDGEMENTS
The authors would be fond of to the editors and the
anonymous commentators for their intuitive and
helpful observations and suggestions that resulted in
significant developments to the current work.






Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
30




REFERENCES
[1] Almuallim H. and Dietterich T.G., Algorithms for
Identifying Relevant Features, In Proceedings of the
9th Canadian Conference on AI, pp 38-45, 1992.
[2] Almuallim H. and Dietterich T.G., Learning
boolean concepts in the presence of many irrelevant
features, Artificial Intelligence, 69(1-2), pp 279-305,
1994.
[3] Arauzo-Azofra A., Benitez J.M. and Castro J .L.,
A feature set measure based on relief, In Proceedings
of the fifth international conference on Recent
Advances in Soft Computing, pp 104-109, 2004.
[4] Baker L.D. and McCallumA.K., Distributional
clustering of words for text classification, In
Proceedings of the 21st Annual international ACM
SIGIR Conference on Research and Development in
information Retrieval, pp 96-103, 1998.
[5] Battiti R., Using mutual information for selecting
features in supervised neural net learning, IEEE
Transactions on Neural Networks, 5(4), pp 537-550,
1994.
[6] Bell D.A. and Wang, H., A formalism for
relevance and its application in feature subset
selection, Machine Learning, 41(2), pp 175-195,
2000.
[7] Biesiada J . and Duch W., Features election for
high-dimensionaldataa Pear-son redundancy based
filter, AdvancesinSoftComputing, 45, pp 242C249,
008.
[8] Butterworth R., Piatetsky-Shapiro G. and
Simovici D.A., On Feature Se-lection through
Clustering, In Proceedings of the Fifth IEEE
international Conference on Data Mining, pp
581-584, 2005.
[9] Cardie, C., Using decision trees to improve
case-based learning, In Proceedings of Tenth
International Conference on Machine Learning, pp
25-32, 1993.
[10] Chanda P., Cho Y., Zhang A. and Ramanathan
M., Mining of Attribute Interactions Using
Information Theoretic Metrics, In Proceedings of
IEEE international Conference on Data Mining
Workshops, pp 350-355, 2009.
[11] Chikhi S. and Benhammada S., ReliefMSS: a
variation on a feature ranking Relief algorithm. Int. J .
Bus. Intell. Data Min. 4(3/4), pp 375-390, 2009.
[12] Cohen W., Fast Effective Rule Induction, In
Proc. 12th international Conf. Machine Learning
(ICML95), pp 115-123, 1995.
[13] Dash M. and Liu H., Feature Selection for
Classification, Intelligent Data Analysis, 1(3), pp
131-156, 1997.
[14] Dash M., Liu H. and Motoda H., Consistency
based feature Selection, In Proceedings of the Fourth
Pacific Asia Conference on Knowledge Discovery
and Data Mining, pp 98-109, 2000.
[15] Das S., Filters, wrappers and a boosting-based
hybrid for feature Selection, In Proceedings of the
Eighteenth International Conference on
MachineLearning, pp 74-81, 2001.
[16] Dash M. and Liu H., Consistency-based search
in feature selection. Artificial Intelligence, 151(1-2),
pp 155-176, 2003.
[17] Demsar J ., Statistical comparison of classifiers
over multiple data sets, J . Mach. Learn. Res., 7, pp
1-30, 2006.
[18] Dhillon I.S., Mallela S. and Kumar R., A divisive
information theoretic feature clustering algorithm for
text classification, J . Mach. Learn. Res., 3, pp
1265-1287, 2003.
[19] Dougherty, E. R., Small sample issues for
microarray-based classification. Comparative and
Functional Genomics, 2(1), pp 28-34, 2001.
[20] Fayyad U. and Irani K., Multi-interval
discretization of continuous-valued attributes for
classification learning, In Proceedings of the
Thirteenth International J oint Conference on
Artificial Intelligence, pp 1022-1027, 1993.

L.Anantha Naga Prasad M.Tech,
Computer Science and Engg, Anantha Lakshmi Institute of
Technology&Sciences, JNTUA, Andhra Pradesh, India.
naga4all16@gmail.com. His current research interests
include data mining/machine learning, information
retrieval, computer networks, and software
engineering.


K.Muralidhar Assistant Professor,
Department of CSE, Anantha Lakshmi Institute of
Technology&Sciences, JNTUA, Andhra Pradesh, India.
muralidhar.kurni@gmail.com. His current research
Interests include data mining/machine learning,
computer networks, and software engineering.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
31
Improving the Location of Nodes in Wireless Ad
Hoc and Sensor Networks Using Improvised LAL
Approach


Abstract: This Paper aims to present a localization of many nodes
in wireless networks. Localization is an enabling technique for
many sensor network applications. In the deployment of network
includes many nodes, due to hardware or deployment constrains,
the networks almost not entirely localizable, there is possible for
any of the nodes placed as a non-localizable nodes in the wide
range of network area. From that, the server doesnt know where
the destination nodes have been deployed. It is difficult to transfer
the data from source to sink whether the deployed nodes is not in
the range of particular network. So, I proposed a Improvised LAL
approach that triggers a single round adjustment and that carries
and beware of node localizability, that makes easy for us to make
the non-localizable nodes in the network into localizable nodes.
Index terms: Localization, localizability, Improvised LAL
approach, wireless and sensor networks.

I. INTRODUCTION
Beyond the established technologies such as mobile
phones and WLAN, new approaches to wireless
communication are emerging; one of them are so called ad
hoc and sensor networks. Ad hoc and sensor networks are
formed by autonomous nodes communicating via radio
without any additional backbone infrastructure.
Localization is the main problem in wireless ad-hoc and
sensors networks in which each and every node determines
its own location in network region.If you board the wrong
train, it is no use running along the corridor in the other
direction- Dietrich Bonhoeffer . As he says, wireless
technology has the capability to reach any location on the
earth. Defining the ad-hoc network in terms of network as an
autonomous system of mobile hosts(MHs),connected by
wireless links, using that, the mobile hosts that connect with
the base station.
To locate non-localizable nodes, the traditional approach
mainly focuses on how to tune network settings according to
these nodes. At first they attempt to deploy additional nodes
or beacons in application fields. Beacons are act as a
backbone for our network. Due to this increment in the
deployment of node density and creates abundant internodes
distance constraints thus, enhancing the localizability. But
this attempt lacks to provide feasibility, since the additional
nodes and beacons should be placed in the region of non-
localizable nodes, in the network. The controlled motion of
beacons provides thorough information for the localization of
nodes, but it has a limitation on adjustment delay and
controlling overheads.
One approach is to augment that is to make the greater in
transmitting power of nodes stage-by-stage until all nodes
become localizable, which causes multiple rounds of
configuration dissemination and data collection in a network.
A straight-forward single-round configuration solution is
maximizing the ranging capability of the network regions.
The drawback of power maximization is that it introduces
many unnecessary distance measurements, which are
obtained with costs. In this paper, I propose an Improvised
localizability-aided localization (LAL) that arrogates
sufficient condition of node localizability to identify
localizable and non-localizable nodes.

II. RELATED WORK
The localization methods can be categorised as a range-
free and range-based methods. Range-free localization
methods, merely use neighbourhood information (such as
node connectivity and hop count) to determine node
locations. Range-based approaches assume nodes are able to
measure internodes distances, from that we can derive the
accurate locations of the nodes. There are number of range-
based algorithms are used to find out the accurate location of
the nodes, these algorithm adopt distance ranging techniques,
such as radio signal strength (RSS) and time difference of
arrival (TDoA).The RSS maps received signal strength from
the distance between two nodes like a signal extinction
model, while TDoA measures the signal propagation time for
distance calculation. One can see that naive approaches for
localization are not adequate for all scenarios. While it may
be possible to manually configure each node with its position
P. Saravanan
1
, Er S. Muthukumar
2
, Dr.K.Rubasoundar
3

1
Department of Computer Science and Engineering, M.E. Scholar, Anna University,
Sree Sowdambika College of Engineering, Aruppukpttai, TamilNadu, India.
saravananselva007@gmail.com
2
Department of Computer Science and Engineering, Professor/Head of the Department, Anna University,
Sree Sowdambika College of Engineering, Aruppukottai, TamilNadu,India.
muthukumar_ssce@yahoo.com
3
Department of Computer Science and Engineering, Professor/Head of the Department, Anna University,
PSR College of Engineering, Sivakasi, TamilNadu,India.
rubasoundar@yahoo.com

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
32
in small and static networks, this approach is impracticable in
envisioned large-scale networks of thousands of
geographically distributed nodes.
From the past, a few works and techniques are done on
localization in non-localizable networks. In previous
technique such as range-free localization the cop count
between two nodes are proportional to their distance. In
range-based localization technique that measure the accurate
distance between two nodes using specific ranging hardware.
A technique is a GPS-equipped mobile node to localize fixed
nodes by measuring the distance between the mobile node
and fixed ones, which use a mobile node with known location
information. Wu- proposes a similar approach with the
assumption that each node can move around and measure the
distances to its neighbors and the relative distances between
successive positions along its route. However, the availability
of mobile nodes is much more costly and not scalable for this
approach. In contrast to this mobility based approaches,
Anderson-propose a graph manipulation method to assure the
network localizability. A distributed range free localization
scheme is used in recent years for localization in non-
localizable networks. In DRLS a node called anchors, get
their own location information via GPS or some other
mechanism. The other nodes called normal nodes that do not
have its own location information a few algorithm has been
used to estimate the nodes location information effectively.

III. FINE-GRAINED APPROACH
3.1 Formation of Network Topology
Here a network has been formed with the help of
dynamic topology due to a mobility of the nodes in the
network. Setting up and organizing such a virtual
infrastructure is an important challenge. The inherent trade-
off between energy-efficiency and rapidity of event
dissemination is characteristic for wireless sensor networks
consisting of battery driven devices.
Localization in wireless ad hoc and sensor networks is the
main problem for every node to determines its own location,
a wireless ad hoc or sensor network cannot be ridiculously
dense because the mechanism of topology control is usually
used to reduce the collision and interference, ignoring the
localizability of thenetwork.Thetopology of these networks
often plays a crucial in the speed with which certain tasks can
be accomplished using these networks.


Figure 1 Formation of network
3.2 Localization Method
In localization method, first attempt to deploy a
network with corresponding topology that described in
previous section how to deploy a network with multiple
nodes. Then we are going to measure the deployed nodes
location whether it is, within the range of network or not,
there is a main problem to find out the locations of each and
every node in the network. From the fig 1, the localizable
nodes are shown within the network region and non-
localizable nodes are indicated as red dots, are out of region
of our network. If there is an additional nodes, which is going
to be deployed in the network that increases in node density
and creates abundant internodes distances constraints, this
enhancing the node localizability.
3.3 Measurements on Network Using Distance Graph
The network is said to be a unique and localizable,
then it must have a unique set of rigidity and distance graph
and also the set of anchors such of beacon nodes. Whether a
formed network cannot be localizable given its distance
graph, then it is called as a localizability problem.
Some of the method used to deploy a distance graph using
mathematical formation. The distance graph is formed from a
collection of points in the Euclidean space. The set of
measurements for the network can be modelled by a distance
graph G, let G =(V,E),where V denotes the set of vertices and
E denotes the set of edges. For (i,j) V ;and (i,j) E if the
distance between i and j can be measured or both of them are
in known locations, are exploited by a graph theory

3.4 Construction of Localizable Graph
It is important to construct a localizable graph
through an incremental construction. A common
mathematical formation can also be used to construct the
localizable graph. From the graph theory, Define G
2
as (V,E
U E
2
), where (i;j)E
2
if ij and k2 V such that (i,k) and
(j,k)E. Similarly, define G
3
as (V,EUE
2
UE
3
), where (i,j)E
3
if ij and kV such that (i,k)E and (j,k)E
2
.

IV. DESIGNING OF PROPOSED SYSTEM
4.1 Improvised Lal Design And Implementation
Design phase of Improvised LAL has been begun
with being aware of node localizability, Improvised LAL can
effectively conduct the adjustment of network ranges, but in
traditional approach that is only an indistinctive network
adjustment can be made out and could only make
indistinctive augmentations. Basically, Improvised LAL
consists of three major modules and is workflows are shown
in Fig. 1.
In improvised LAL approach, first the network has been
deployed by using the network topology is described in
previous sections. After the network formation, the proposed
techniques are explored in the network and the nodes in the
network are act depend upon the described techniques. After
the explosive of all techniques has been finished the nodes
are try to send the packet and location information. There is a
large number of nodes has been deployed in the network, so
the density of the network much more increased and each
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
33
node doesnt know the location information of other nodes.
An Improvised LAL approach used here to know the location
information of all other nodes. Processing and work flows of
Improvised LAL approach are explained below,
Module 1: Localizability testing. In Localizability testing,
after the network is deployed in an application field, due to
some hardware or environmental factors that is for
unpredictable issues in the design phase, it may not ready for
localization. So, node localizability testing is conducted in an
Improvised LAL, which identifies localizable and non-
localizable node in a network for further adjustment.
Module 2: Analysis of Network structure. In an analysis of
the structure of the network must support fine-grained
approach, to measure the accurate location information about
the entire node. So, we have to decompose a constructed
distance graph into two-connected components. These
components are managed as a tree structure and the one of
these components containing beacons in the root of the
distance graph. Adjustments are made out along with the tree
edges from the root to the leaves. Beacons are mainly used to
find out the Localization and improve the accuracy of the
nodes in the network.

Figure 2 Decomposition of a Graph
Module 3: Distinctive adjustment. In Improvised LAL
treats nodes differently regarding to their localization and
places in the component tree. Through vertex augmentation,
Improvised LAL converts all non-localizable in a single
round. The network adjustments made out by Improvised
LAL are localizable and can be localized by the existing
localization and localizability techniques.


Figure 3 Workflows between traditional and Improvised LAL approach,
Module 1 can be done by applying Theorem 2(shown in
below). For a given specific node, its localizability that
depends upon the property of disjoint paths and redundant
rigidity, which is being tested in polynomial time by network
flow algorithms and the pebble game algorithm, respectively.
That is, a node localizability testing can be conducted. In
Module 2, a created distance graph is decomposed into two-
connected components using depth-first search. And these
components are managed as a tree structure and the
adjustments can be made in these tree edges. In module 3,
node adjustments are made out along the paths of the
component tree starting at the root. I present an Algorithm
(Basic _Localization _Algorithm) to explain the module 3.
From that, the node will know the location information of
all other nodes in the network. In the network topology, the
nodes are formed as a cluster and a cluster has a cluster head,
a cluster head have location information about all of the other
nodes in the cluster. If a node placed in cluster A want to send
a packet to a other node that placed in cluster B, then it will
send a request to cluster A, a cluster head A have the location
information about all of these node and where it is deployed
then redirect the request to cluster head B, the B send the
packet to the corresponding node that formed under that
particular cluster
4.2 Add_Heuristic Approach


Figure 4 to find out all non-localizable neighbors of localizable
In Add_heuristic approach, find out all non-localizable
neighbors of localizable vertices. From the above figure 4, the
edges have been added for each non-localizable node to
convert these nodes as a localizable node. The edges have
been added by heuristic algorithm to connect these non-
localizable nodes for the purpose of delivering the data from
localizable nodes from non-localized nodes in the deployed
network. Algorithm for Add_heuristic is described in a later
section.
4.3 Geographical Routing Approach
Geographical routing is an emerging technique to find out the
direction of any nodes in the certain region of any network. A
2D graph has been deployed and it has x-axis and y-axis
using that a nodes direction has been found out. Basically a
network has been deployed and measured by a distance
graph. If a sender node try to deliver a packet to a receiver
node but it does not know the location information of the
receiver nodes, so using this approach the location that is the
direction of the node has been found out. Here a beacon
signals are act as a backbone of our network The fig 5 shows
that, a sender collects the nodes location information based
on the direction of the nodes, and finally they select a shortest
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
34
path from the all selected paths, from that shortest path they
sent the data packet to the corresponding receiver.


Figure 5 Geographical Routing

4.4 Ease of Use
Description of an Algorithm
A Basic _Localization _Algorithm, applied here for the
purpose of localize all the nodes in the network without any
incompatibility. In that algorithm, edges are added by vertex
augmentation of all non-localizable node vertices in G
A
, that
is Graph A and G
A
is localizable from the Theorem 3(shown
in below). This Algorithm repeats all of the steps, until all
components get handled. After applying of an Algorithm, the
entire network is localizable; all of node in the network gets
localized. Popular localization algorithms can then be used
seamlessly to localize all nodes in the network.
The repeated edges has been reduced by analyze the graph
properties of these components.

Figure 6 proofs of theorem3, red dots and black holes
Algorithm 2.Add_Heurictic _Algorithm.
If G
A
is used as an input of Add_Heuristic, the algorithm first
finds all non-localizable neighbors of localizable vertices, and
adds edges for each non-localizable one according to the
following analysis made out from the distance graph. Then
add edges to all non-localizable vertices from the localizable
vertices. After that, the adjustments are made out
continuously to localize these non-localizable nodes by
adding edges from non-localizable vertices to localizable
vertices.
4.5 Analysis from the Graph
From the above graph GA some non-localizable vertices have
at least one localizable neighbour vertex. Adding two edges
between the vertices, which connect two neighbor vertices on
different vertex-disjoint paths to the non-localizable vertex, is
enough to make the vertex localizable according to Theorem
2. From the analysis from the above graph, If a non-
localizable vertex has more localizable neighbours within
two-hop distance, that is, the distance between these
neighbours are connected with only two edges, then, these
localizable vertices makes it non-localizable as a localizable.
Some of the decomposed components in a distance graph are
not localizable due to the lack of beacons. If these nodes are
adjusted to be localizable in a decomposed component, the
component is instantly localizable without extra
manipulation.

V. THEOREMS
Theorem 1. A graph with nvertices is globally rigid in two
dimensions if and only if it is three-connected and
redundantly rigid. A graph G=(V,E) is called k-connected
(for kIN) if || >k and G-X is connected for every set X
V with ||<k.
Theorem 2. In a distance graph G=(V,E) with a set BV of
k3 vertices at known locations, a vertex is localizable if it
is included in the redundantly rigid component inside
which there are three vertex-disjoint paths to three distinct
vertices in B.
Theorem 3. Suppose G=(V,E) is a two-connected graph with
a set B of k3 beacons and BV . Let V
N
denote the set of
non-localizable vertices, and E
N
denote the set of edges
(i,j),iV
N
and (i,j)E
2
. Then, G=(V ,E U E
N
) is localizable.

VI. DISCUSSION
The network topology is a by product of some basic
services in ad hoc and wireless sensor networks,

Figure 7 GreenOrbs system in a campus
Its collection induces none or little additional overhead.
Inspired by this fact, this paper naturally adopts the
centralized scheme, which mainly comes from the
localizability testing part. The fig 7 shows the wireless sensor
nodes that scattered the signals to get the location information
of other nodes, which is intended to send the data packets
from source to sink and vice versa.
The rigidity and beacons are used as backbone of our network
to transform the datas from source to destination. A rigidity,
they find the co-ordinates and transforms these relative co-
ordinates to a global co-ordinates,
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
35
Table 1 Experimental results on Greenorbs
LAL_Basic Add_heuristic Indistinctive
adjustment
No of edges 340 309 736
No of added
edges
55 24 451
No of power
1 nodes
61 73 0
No of power
2 nodes
39 27 100
No of power
3 nodes
0 0 0
Assume that noisy results (the outliers with large errors) are
sifted by these approaches, and only used accurate ranging
results in Improvised LAL design.
VII. EVALUATION

7.1 Experiment
To examine the correctness and effectiveness of
Improvised LAL approach, I implement it on the ongoing
wireless sensor network system, that is Green Orbs system
and the data trace collected from the system Green Orbs.
From the data collection the important factor is to reduce the
energy consumption of the deployed nodes. Using, cycling
theorem, the transmission power is also well controlled under
the highest level to provide the enough connectivity for data
collection or other services.


Figure 8a Number of nodes plotted at different power levels
From the fig.9, there are 285 edges are needed to connect the
edges between the nodes. A comparison can be made out
between the edges and nodes with different transmitting
power output from the resultant topology, I just denote the
nodes that have original communication range as power1
nodes, the ones with doubled and tripled ranges are denoted
as power2 and power3 nodes from doubled and tripled like
wise
From those following figures, the comparison made between
the improvised LAL and traditional approach, to find out the
location of the nodes.

7.2 Simulation
Simulation is the process or operation that has a limitation
among the real time implementation of original process. In
this paper, I have to stimulate and further examined the
scalability and efficiency of Improvised LAL under different
network instances and varied network parameters.The
transmission power requirements of LAL_Basic and Add-
Heuristic algorithms in deployed nodes are shown in fig.8.
Plots the number of nodes at different power levels. As
shown in fig.8a and 8b, results of Add_Heuristic and
LAL_Basic, have much more power 1 nodes than the
traditional approach except R denotes original
communication range less than 0.66 and much less than
power


Figure 8c High power levels at Improvised LAL intend to add more edges in
the graph
3 nodes than traditional approach except when R greater than
0.99 that shows in fig.8b.
The results in fig.8c also explain about the Improvised LAL
that can have a ability to serve much extra edges than
traditional approach because higher power levels tend to
induce more edges than a lower one.
Abbreviations used
Traditional approachIND (Indistinctive Adjustment
Approach)
Improvised Traditional approachImprovised
IND(Improved Indistinctive Adjustment Approach)
Because Traditional approach treats the entire network
indistinctly and obviously needs more adjustments than LAL,
an improved approach (Improved IND) is implemented
instead. From the above all description it suggests that
LAL_Basic and Add_Heuristic are more fine grained than
improved IND.
Figure 8b Number of nodes at different power levels in traditional approach
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
36

Figure 9 Testing Improvised LAL and Improvised traditional approach on network instances consists of 400 nodes
VIII. CONCLUSION
The analysis of the limitation and power
requirements of existing approach on localization in non-
localizable networks, and propose a Improvised
Localizability-aided Localization approach named
Improvised LAL. Improvised LAL treats the network as
whole and localize all the nodes in the network, while if it is
in non-localizable state. That makes the adjustment
corresponding to node localizability results in the first
module, other than traditional approach, that includes the
nodes in localizability testing and made a indistinctive
adjustment. From that Improvised LAL approach, a nodes
need to be augment with their ranging capability to connect
and added new edges are needed to be measured. It also has
some good characteristics for the purpose of implementation
aspects in the real world application and I, implement the
Improvised LAL and demonstrate its effectiveness through
working system experiment and extensive simulations.

IX. FUTURE SCOPE
I, proposed the Adaptive Position Update strategy
to address these problems. The APU scheme employs
two mutually exclusive rules. The MP rule uses
mobility prediction to estimate the accuracy of the location
estimate and adapts the beacon update interval accordingly,
instead of using periodic beaconing. The ODL rule allows
nodes along the data forwarding path to maintain an accurate
view of the local topology by exchanging beacons in
response to data packets that are overheard from new
neighbors, Then, We mathematically analyzed the beacon
overhead and local topology accuracy of APU and validated
the analytical model with the simulation results.

X. REFERENCES

[1] Vaidyanathan Ramadurai,Mihail L. Sichitiu, Simulation-based
Analysis of a Localization Algorithm for Wireless Ad-Hoc Sensor
Networks,Department of Electrical and Computer Engineering North
Carolina State University
Raleigh,NC27695,20/03,http://www4.ncsu.edu/~mlsichit/Research/Pub
lications/opnet.pdf.
[2] Vaidyanathan Ramadurai,Mihail L. Sichitiu, Localization in Wireless
Sensor Networks: A Probabilistic Approach ,Department of Electrical
and Computer Engineering North Carolina State University Raleigh,
NC 27695,2003
http://www4.ncsu.edu/~mlsichit/Research/Publications/probabilisticLo
calizationICWN.pdf.
[3] Dragos Niculescu and Badri Nath, Ad Hoc Positioning System(APS)
Using AOA, DATAMAN Lab Rutgers University, IEEE Infocom
2003, http://infocom2003.ieee-infocom.org/papers/42_04.pdf.
[4] Sunil J ardosh, Prabhat Ranjan, Intra-Cluster Topology Creation in
Wireless Sensor Networks, DA-IICT,Gandhinagar,India IEEE 2007,
http://delab.csd.auth.gr/~papajim/presentations/files/04475768.pdf
[5] J ang-Ping Sheu, Pei-Chun Chen, Chih-Shun Hsu, A Distributed
Localization Scheme for Wireless Sensor Networks with Improved
Grid-Scan and Vector-Based Refinement, Department of Computer
Science, National Tsing Hua University, Department of Computer
Science and Information Engineering, National Central University,
IEEE transactions on mobile computing, September 2008,
http:www.researchgate.net/...A_Distributed_Localization_Scheme_for_
Wireless sensors networks.
[6] Wint Yi Poe,J ens B. Schmitt, Node Deployment in Large Wireless
Sensor Networks:Coverage, Energy Consumption, and Worst-Case
Delay disco | Distributed Computer Systems Lab University of
Kaiserslautern, Germany,AINTEC09, November 1820, 2009,
Bangkok,
Thailand.https://disco.informatik.unikl.de/discofiles/publicationsfiles/AI
NTEC09-1.pdf.
[7] Zhisu Zhu, Yinyu Ye, Rigidity: Towards Accurate and Efficient
Localization of Wireless Networks Stanford University Stanford,
Stanford, IEEE publications, infocom,
2010proceedingsconference,http://ieeexplore.ieee.org/xpl/freeabs_all.js
p?arnumber=5462057&abstractAccess=no&userType=inst.
[8] Amitangshu Pal, Localization Algorithms in Wireless Sensor
Networks: Current Approaches and Future Challenges. Department of
Electrical and Computer Engineering, The University of North
Carolina, Network
protocolandalgorithms2010,http://www.macrothink.org/journal/index.p
hp/npa/article/viewFile/279/276.
[9] Zheng Yang and Yunhao Liu, Understanding Node Localizability of
Wireless Ad-hoc Networks, Hong Kong University of Science and
Technology, IEEE Infocom
proceedings2010,http://www.cse.ust.hk/~liu/localizability.pdf.
[10] Zheng Yang,Chenshu Wu,Tao Chen,Yiyang Zhao,Wei Gong,Yunhao
Liu, Detecting Outlier Measurements Based on Graph Rigidity for
Wireless Sensor Network Localization School of Software and
TNList, Tsinghua University, Beijing 100084, ChinaNational
University of Defense Technology, Changsha 410073, China,IEEE
transactions on vehicular technology, vol. 62, no. 1, january 2013.
[11] Kuei-Li HuangLi-Hsing YenJ ui-Tang WangChao-Nan WuChien-
Chao Tseng , A Backbone-Aware Topology Formation (BATF)
Scheme for ZigBee Wireless Sensor Networks, National Chiao-Tung
University, Hsin-Chu, Taiwan, Industrial Technology Research
Institute, Hsin-Chu, Taiwan, ROC, Wireless Pers Commun (2013),
http://download.springer.com/static/pdf/185/pdf.
[12] J onathan Bachrach and Christopher Taylor, Localization in Sensor
Networks, Computer Science and Artificial Intelligence Laboratory,
Massachusetts Institute of Technology, Cambridge, MA 02139
[13] Quanjun Chen, Salil S. Kanhere, Mahbub Hassan, Adaptive Position
Update for Geographic Routing in Mobile Ad-hoc Networks, The
University of New South
Wales,sydney,Australia.http://www.cse.unsw.edu.au/~mahbub/PDF_Pu
blications/TMC_adaptive.pdf
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
37
Survey on Wireless Sensor Networks Routing
Protocols Based on Energy Efficiency
Ms. S.NIVEDITHA
1
, Ms.M.USHA
2
1
M.E., Department of Computer Science and Engineering, Velammal Engineering College, Chennai.
2
Assistant Professor, Department of Computer Science and Engineering, Velammal Engineering College,
Chennai.
uni que. ni ve@gmai l . com, umahal i ngam@gmai l . com

Abstract-Energy consumption and balancing has always been a
hot research topic in wireless sensor networks, in this survey,
some of the hierarchical routing protocols(LEACH,
HEED,PEGASIS, TBC, TREEPSI) with high energy efficient
were examined based on some strategies such as data aggregation
and clustering, routing and dynamic node allocation methods in
the clustering environment. The routing protocols are compared
with each other based on these strategies and based on the
network life time definitions. Simulation results show which
protocol has better prolonging lifetime than other traditional
protocols by reducing the energy consumption and load
balancing.
Key words: Load balancing, Data aggregation, clustering routing
protocol.

I. INTRODUCTION
Wireless sensor networks periodically collects the information
of the interested unmanned monitoring area independently
through deploying a large number of wireless sensor nodes
and send the data to the remote base
station.Wirelesssensornodesare deployedrandomlyand
denselyina
targetregion,especiallywherethephysicalenvironmentissohars
hthatthesensor counterpartscannotbe
deployed.Afterdeployment,thenetworkcannotworkproperlyun
lessthereissufficientbatterypower.
Wireless sensor nodes have so many important applications,
including military application, disaster prediction and
environment monitoring.
The military application is performed in the rough
environments where the sensor parts cannot deploy due to the
energy consumption.
Clusteringis very effectivetechnique.Itcan greatly contribute
to overall system scalability, lifetime, and energy
efficiencyinwirelesssensornetworks(WSNs).Grouping sensor
nodes into cluster to satisfy the scalability and achieve high
energy efficiency and prolong network lifetime in large scale
WSN environment. The hierarchical routing protocol imply
cluster based organization of the sensor node in order to
performdata fusion and aggregation which leads to energy
saving. In hierarchical network structure each cluster has a
leader, which is called the cluster head (CH) which perform
the data aggregation and data transmission to the base station
by direct or intermediate communication with other CH
nodes.

Energyconsumption ofanodeisduetouseful,
wastefuloperations.The
usefuloperationsincludestransmittingorreceivingdatamessag
es,andprocessingrequests.Thewastefulconsumption
includesconstructingroutingtree,overhearing,
retransmittingbecauseofharshenvironment,
dealingwithredundantover- head messages,andidlelisteningto
the media.Each node sends information directly to the base
station to full fill the basic task of the sensor nodes. If the
base station is located far away fromthe target area, the
sensor nodes will die quickly due to much energy
consumption. The distance is also very important factor and
the direct transmission leads to the unbalanced energy
consumption.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
38
In general WSN produce large amount of data so if data
fusion could be used, the through put could be reduced. Data
fusion reduces the redundant data by combining the redundant
data to reduce the transmission by suppression, min, maxand
average functions. Substantial energy can be saved through
the data aggregation this technique has used to achieve energy
efficiency and traffic optimization in a number of routing
protocols. The data fusion or data aggregation assume that the
length of message transmitted by each relay node should be
constant. That is each transmits the same volume of data.
The distance between the nodes to BS leads to two extreme
cases
1. The length of message transmitted by the parent
node not only depends on the own length also
depends on the length of the message received from
its child nodes.
2. The relay node should transmit the length of the
message which is the sum of its own sensed data
and the received data
Network lifetime has two different definitions: [1]
1. The time from the start of the operation to death of
the first node.
2. The time from the start of the operation to death of
the last node
Energy aware protocols and data gathering protocols offering
high scalability. A common solution for the balancing energy
consumption among all the network nodes, is to periodically
re-elect new CH in each cluster.
The remainder of the paper is presented as follow:Section
IIdiscuss the literature survey. Finally, Section III concludes
the paper.
II. LITERATURE SURVEY
A routingprotocol specifies the routers communicate with
each other, and selects the routes between any two nodes on
a computer network. Routing algorithms determine the
specific choice of route. Each router has knowledge of
networks attached to it directly. A routing protocol shares this
information first among immediate neighbors, and then
throughout the network. Routing strategy is generally based
on the following criteria
1. Achieve minimumcost forwarding, while design of
the most data forwarding protocol is based on the
optimality.
2. Reducing the minimum number of performed
operations
3. Scale to large network size with some constrains
WSN is classified into three categories based on the structure:
Flat, hierarchical, and location based routing.



2.1Hierarchical protocols:
Hierarchical routing is a method of routing in networks that is
based on hierarchical addressing. It imposes the structure on
the network to achieve energy efficiency, stability and
scalability. In hierarchical protocols nodes are organized in
clusters inwhich a node has some responsible takes the role of
cluster head which is responsible for coordinating and
message transmitting activities. Clustering reduces the energy
consumption to prolongs the life time, can balance the load
and increase the scalability.
2.1.1. LEACH: [2][3]
LEACHisthe most energy efficientclusteringalgorithm
forWSNs that forms nodes in clusterbased environment.
Based onthe receivedsignal strengthand uses these
localclusterheadsasrouters tothe BS.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
39

Data transferto theBSconsumesmoreenergy,all thesensor
nodeswithina clustertaketurns withthe transmission by
rotatingthe cluster heads. This
leadstobalancedenergyconsumption ofallnodes,and hence a
longerlifetime ofthe network
The basic operations of LEACH was performed in two
phases:
Setup phase:
Cluster formation.
Cluster head was selected and informed to
all other nodes.
Preparation of transmission schedule.
Steady phase:
Data fusion.
Transmission to destination node

Inthe setup phase, each node n chooses a randomnumber
between 0 and 1. If the number is less than a threshold T (n),
the node becomes a cluster head for the current round. The
threshold is set as:
T (n) = ___P______ if n belongs to G
1-P *(r mod1/P)
Otherwise 0

P =percentage of cluster heads,
r =current round
G =set of nodes
G is not cluster head in the last 1/P rounds. After the cluster
heads are selected, the other nodes organize the local clusters
by choosing the most appropriate cluster head (closest cluster
head).
In the steady-state phase the cluster heads receive data from
the nodes, and transfer the fused data to the base station. And
the communication is performed via the TDMA
Advantages of LEACH:
It performs most of the communication inside
the clusters, and provides scalability.
The data fusion leads to a limit on the traffic
generated in the network.
Single-hop routing from node to cluster head,
hence saving energy.
It increases network lifetime in three ways:
Distributing the role of CH to the other
nodes.
Aggregating the data by the CHs.
TDMA, which, assigned by the CH to
its members, puts most of the sensor in
sleep mode, in event-based
applications. Itincreases the network
lifetime and achieve a more than 7-fold
reduction in energy dissipation
compared to direct communication
It is dynamic clustering and well-suited
for applications where constant
monitoring is needed and data
collection occurs periodically to a
centralized location.All thesensor
nodeswithina clustertaketurns withthe
transmission by rotatingthe cluster
heads. This leadstobalancedenergy
consumption ofallnodes,and hence a
longerlifetime ofthe network.
Disadvantages of LEACH:
Due to the selection of cluster heads. Because
itdont consider the energy of the nodes for CH
selection.
It significantly relies on cluster heads and face
robustness issues.
Additional overheads due to cluster head
changes leads to the energy inefficiency for
dynamic clustering in large networks.
It has no inter cluster communication, and this
needs high transmission power.
It selects the CH in randommanner it results that
the low power node can become a cluster head, it
will deplete its energy quickly. And the local
cluster head only by distance between
themselves and it results the formation of
minimumor maximumcluster.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
40
Single hop method.

2.1.2. HEED: [4]
HEED (Hybrid energy efficient distributed clustering)
periodically selects cluster heads according based on the
node residual energy and other parameter like proximity
to node degree. HEED terminates in O (1) iterations, it
incurs low message overhead and achieve uniformcluster
head distribution across the network.
HEED was designed to select different cluster heads in a
field according to the amount of energy that is distributed
in relation to a neighboring node
The basic operations of HEED protocol was performed
by the following steps:
Defining parameters used in the cluster
formation.
Protocol operation.

Cluster formation:
Cluster head selection is based on the residual energy of
each node. The parameters are measuring the residual
energy and estimating the intracluster communication
cost. The first parameter select initial set of cluster heads,
the second parameter performing the cluster power level
which increase initial reuse and reserve power for
intercluster communication.

Protocol operation:
At each node, the clustering process requires a number
of iterations, which was referred as Niter. Every step
takes time tc to receive messages from any neighbor
within the cluster range. An initial percentage was
assigned to cluster heads among all n nodes, Cprob.
Cprob is only used to limit the initial cluster head
announcements, and has no direct impact on the final
clusters.
Before a node starts executing HEED, itsets its
probability of becoming a cluster head, CHprob, as
follows:
CHprob = Cprob * Eresidual / Emax
Eresidual =estimated current residual energy in the node.
Emax =reference maximum energy which is typically
identical for all nodes. The CHprob value of a node,
however, is not allowed to fall below a certain threshold
pmin. Every node then doubles its CHprob and moved to
next step.
In HEED energy consumption it extends the lifetime of
all the nodes, which adds to the stability of the neighbor
set. Nodes automatically update their neighbor sets by
periodically sending and receiving heartbeat messages.
If a node elected as a cluster head, it sends an
announcement message cluster_head_msg(Node ID,
selection status, cost), where the selection status is set to
tentative_CH if CHprob<1, or final_CH, if its
CHprob = 1. A node considers itself covered if it is
heard fromeither a tentative_CH.
If a node completes HEED execution without selecting a
cluster head that it considers itself uncovered, and
announces itself to be a cluster head.
Advantages:
HEED distribution of energy extends the lifetime of
the nodes within the network thus stabilizing the
neighboring node.
Does not require special node capabilities, such as
location-awareness
Does not make assumptions about node distribution
Operates correctly even when nodes are not
synchronized.
The advantages of HEED are that nodes only require
local (neighborhood) information to formthe clusters
The algorithmguarantees that every sensors is part of
just one cluster, and the cluster heads are well-
distributed.
Disadvantages:
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
41
The random selection of the cluster heads, may cause
higher communication overhead for:
The ordinary member nodes in
communicating with their corresponding
cluster head
Cluster heads in establishing the
communication among them, orBetween a
cluster head and a base station
The periodic cluster head rotation or election needs
extra energy to rebuild clusters
In HEED Multi level hierarchy cannot be achieved.
Recursive applications cannot be applied.
2.1.3. PEGASIS: [5]
PEGASIS (Power-Efficient Gathering in Sensor
Information Systems), is an optimal. In PEGASIS, each
nodes communicates only with the close neighbor and
transmits the information to the base station, it reduces
the amount of energy spent per round.
It distributes the energy load evenly among the sensor
nodes in the network. The sensor nodes are randomly
placed in the play field. The nodes are organized to form
a chain, which was accomplished by the sensor nodes
using a greedy algorithm starting from some node. On the
other hand the BS can compute this chain and broadcast it
to all the sensor nodes.
Construction of the chain starts with the farthest node
fromthe BS. The nodes farther fromthe BS have close
neighbors, as in the greedy algorithm the neighbor
distances will increase gradually. When a node dies, the
chain will be reconstructed and the threshold can be
changed to determine to elect leaders. In it (i mod N)
th

node is chosen to be a leader.

Data gathering in each round each node receives data
fromneighbor, fuses with its own data, and transmits to
the other node in chain.The leader was selected in each
round of communication in a random manner and random
position on the chain, because of the death of nodes at
randomlocations. Token passing approach initiated by
the leader to start the data transmission from the ends of
the chain.
It performs data fusion at every node except the end
nodes in the chain. Each node will fuse its neighbors data
with its own to generate a single packet of the same
length and then transmit that to its other neighbor (if it
has two neighbors).
Advantages of PEGASIS are
1. Local gathering distance is less than LEACH
2. The amount of data for the leader to receive
3. Limiting the number of transmissions and receives
among all nodes.
4. It eliminates the overhead of dynamic cluster
formation, minimizing the distance non leader-nodes
must transmit.
Disadvantages:
Whenevera node dies the whole chain has to be
reconstructed and the scalability is also an important issue
in PEGASIS.
2.1.4. Tree Based Clustering: [6]
Tree-Based Clustering (TBC) is also have cluster based
environment and each cluster has a cluster-head (CH).
Tree conguration:The cluster construct a routing tree
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
42
where the cluster-head acts as a root. All sensor in a
routing tree sends data to a single collection point know
as root.
All the nodes arelocation-aware. The distance between
the node and the root is for determining its level in the
cluster. The nearest node is elected asitsparentnode.
Data transmission: Data transmission performed
simultaneously. Data is gathered and each node fuses the
received data and transmits it to its parent. InTBC each
node holds the information of its neighbors. Fast data
collection is main advantage in TBC
2.1.4.1. TREEPSI:[7]
In TREEPSI sensor nodes are deployed randomly node i
(i
th
node) is placed at a random location. Compute the
path by BS and broadcast the path information to other
nodes. After the construction of the tree it collects
information fromthe field.
TREEPSI is a tree based multi-hop routing protocol to
construct a hierarchical path of the nodes. In this nodes
will fuse the received data with their own data and
forward the resultant data to their parent. It will repeat
this process till all the data are received by the root node.
The data are collected at the root and at last root takes
responsibility to transmit the
datatotheBS.Afterthedeathofnode,
anewtreepathisconstructed. Sothe overhead per
communication round is less as compared to the energy
spent in the data collection phase.
Two schemes:
1. After root initiate data gathering process by sending a
small control packet to the children nodes using
standard tree traversal algorithm.it will consume
some negligible amount of energy but with little
more delay.
2. All the leaf nodes send sensed information to their
parent. The parent nodes fuse the received data with
their own data and forward the resultant data to their
parent until data are received by the root node. It has
less delay, but needs to introduce some multiplexing
scheme to avoid collision.
This path is used for several rounds until node i dies.
After the death of node i, a new tree path is constructed
with node i+1 as the root.
The simulation results shows that TREEPSI outperforms
all the existing protocols in terms of energy efficiency.
2.1.5. GSTEB: [1]
In GSTEB (General self-organized tree based energy
balancing routing protocol) the basic tree construction
and communication operation is similar to the TCB.
In each round BS assigns a root node and broadcasts its
ID and its coordinates to all sensor nodes. And data
aggregation takes place.GSTEB changes the root and
reconstruct the routing tree with short delay and low
energy consumption by this betterbalancedloadis
achieved.
The operation of GSTEB is divided into Initial Phase
1. Tree ConstructingPhase
2. Self-OrganizedDataCollecting
3. Transmitting Phase
4. Information Exchanging Phase.
EL(i)=[residual energy(i)/constant reflect the
minimum energy unit]
GSTEBoutperformsLEACH, PE- GASIS,
TREEPSI [9]andTBC.Because GSTEBisaself-
organizedprotocol,itonlyconsumesasmallamountof
energy ineach round
tochangethetopographyforthepurposeofbalancingt
heenergyconsumption.Alltheleafnodescantransmit
datainthesameTDMAtimeslotsothatthetransmitting
delay
isshort.Whenlifetimeisdefinedasthetimefromthestar
tofthe networkoperationto thedeath of thefirstnode
in the network.

Advantages of GSTEB:
Load balancing
Self-organizing
More efficient than other routing protocols.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
43
GSTEB prolongs the life time by balancing
the load.


III. Conclusions:
In this survey paper, routing protocols are compared based on
the network life time and data aggregation and the load
balancing criteria. Then the new GSTEB protocols
simulations show that GSTEB outperforms LEACH,
PEGASIS, TREEPSI and TBC. Because GSTEB is a self-
organized protocol, it only consumes a small energy and it
balancing the energy consumption and it have short
transmission. GSTEB prolongs the lifetime by 100% to 300%
compared with PEGASIS [1].
GSTEBprolongsthelifetimeofthenetworkbymorethan100%
compared with HEED.
TABLE 1:
NETWORK LIFE TI ME
PROTOCOL ROUND TAKEN TO
A NODE DIE
ROUND TAKEN TO
ALL NODES DEAD
LEACH 118 248
PEGASIS 246 568
TREEPSI 267 611
TBC 328 629
GSTEB 389 677
REFERENCES:
[1] Zhao Han, J ie Wu, Liefeng Liu, J ie Zhang and kaiyunTian A General
self-organized tree based energy balance routing protocol for wireless sensor
networks., vol.61 , 0018-9499, 2014
[2] Ankit Solanki,prof.niten b. patel LEACH-SCH: An innovative routing
protocol for wireless sensor network 4
th
icccnt-2013
[3] An Integrative Comparison of Energy Efficient Routing Protocols in
Wireless Sensor Network ,Ali Norouzi1, Abdul HalimZaim2 1Department
of Computer Engineering, Istanbul University (Avcilar), Istanbul, Turkey
2Department of Computer Engineering, Istanbul Commerce University
(Eminonu), Istanbul, Turkey Email: norouzi@cscrs.itu.edu.tr,
azaim@iticu.edu.tr Received December 8, 2011; revised J anuary 11, 2012;
accepted J anuary 30, 2012
[4] HEED: A Hybrid, Energy-Efficient, Distributed Clustering Approach for
Ad Hoc Sensor Networks, OssamaYounis, Student Member, IEEE, and
Sonia Fahmy, Member, IEEE, IEEE TRANSACTIONS ON MOBILE
COMPUTING, VOL. 3, NO. 4, OCTOBER-DECEMBER 2004
[5] PEGASIS: Power efficient gathering in sensor information system
Stephanie Lindsey and Cauligi S. Raghavendra,IEEE AC paper #242 ,
updated Sept. 29,2001. 0-7803-7231- 2002 IEEE.
[6]O.durmazincel, a.ghosh,b.krishnamachari and k.chintalapudi Fast data
collection in tree based wireless sensor networks ieeecs,comsoc.
[7] TREEPSI: TRee based Energy Efficient Protocol for Sensor
InformationSiddharthaSankarSatapathy, NityanandaSarmaTezpur
University, Napaam-784028, 2006 IEEE

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
44
AN EFFECTIVE ALARMING MODEL FOR DANGER AND ACTIVITY
MONITORING USING WEARABLE SENSORS FOR CHILDREN

P.Vidyadhar
1
, PG Student, Department of ECE, ASCET, Gudur, Andhra Pradesh, India.
K. Dhanunjaya
2,
Head of the Department, Department of ECE, ASCET, Gudur, Andhra Pradesh, India.

AbstractThis paper presents a child activity recognition
approach using a single 3-axis accelerometer and a ultra-sonic
sensor which is belted around the waist of the baby so as to
prevent him from dangers and getting injured. The accelerometer
data is collected and monitored in a computer using IEEE 802.15.4
protocol. In addition to the activity recognition child body
temperature can also be monitored at regular time intervals. A
fire sensor is also embedded in the proposal so as prevent the baby
from fire accidents and a SMS using GSM will be sent to their
parents if they are going to be involved in any fire accidents. Child
activities are classified into 8 daily activities according to our
consideration which are moving left, right, front, back, standing
still, climbing up, climbing down, and stopping. The accuracy
obtained for every activity is around 90% in any situation using
single MEMS ADC sensor and ULTRA-SONIC sensor.
Index TermsAccelerometer, activity classification, activity
recognition, baby care, child care.

I. INTRODUCTION
At presentare getting into some accidents due to lack of
personal monitoring in this busy world, usually child start
walking between 9 and 16 months, there will bechance of
falling fromhigher heights or stairs. As the child learns to
climb, they will be at risk of falling from stairs, chairs and
beds. Children frequent come across injury due to these
accidents unknowingly.Medical research show that these
accidents are one of the most common cause of injuries that
require medical care, and in some situations non fatal injuries
also leads to hospitalization. The main areas these accidents
occur are at homes because of lack of parental care. Thus, a
new effective alarming model for danger and activity
monitoring using triaxial accelerometerfor children is required
to prevent child from accidents at homes. These accidents have
great effect on growth and development of the child.Accident
prevention measures are to be taken effectively. The
challenging thing is the classification activities in terms of
safety and damage occurring to the child. There are many
proposals for the recognition of activity but the challenging
task lies with the accurately recognition of activity the child is.
A smart sensor network is used in this proposal for rescuing the
baby from injuries and some small cracks on the body.
Multisensor link has been provided for elderly people and
children at home. This approach will give the activity data in a
simple and recognizable manner. According to theproposal the
human activity is recognized by fusing two highly accurate
sensors one which is attached to one of the foot and another
sensor to the waist of a baby subject, respectively. Due to the
use of multiple sensors robustness of the classification systems
has been improved drastically and increases thepartialness of
high-level decision making. On the other hand, the ultra-sonic
sensor could fail to detect the activities like head motion, body
tilt, and hand motion. In addition to that and for the purpose of
minimizing the number of sensors worn, it is important to know
the capability of a certain position to classify a set of activities.
Recently, Attallaet al. Investigated the effects of sensor
position and feature selection on activity classification
tasksusing accelerometers. Accelerometers are the mostbroadly
used sensors to recognizeambulation activities such as walking
and running the advantage of the accelerometer is inexpensive,
require relatively low power, and are greatest applications in
most of mobile phones. The study of the optimal sensor
concluded that positions depend on the activities being
performed by the subject and these activity recognition can be
made optimized using accelerometers.The dominant role in the
designing the system is that it must be designed in a small
space and low weight which can be bearable by the baby. As
we are optimizing the sensor activity so we have to
compromise in the accuracy. It is difficult to configure an
optimal system. It depends on mainly on two important factors
one is sensor activity and the positions of the baby. In our
study, to decrease the incomfortness the waist belt is kept
diaper for the children below three years of age, during
physical activity and to measure body motions such as moving
left, right forward backward, climbing up and climbing
down.In our proposal mainly we have designed a wearable
sensor device and a monitoring application to collect
information about the activities and to recognize baby activities
baby is doing. We classified baby activities into 8 daily
activities which are moving left, right, front, back, standing
still, climbing up, climbing down, and stopping. As multiple
sensors are embedded in a wearable devicewhich are more
accurate for collecting differenttypes of sensing informationbut
would be very inconvenient for users.

TABLE I
SENSOR AND TYPES OFTHECOMPONANTS USED

TYPE SENSOR VALUE FEATURE
Space

RFID Roomidentification
Location
(kitchen, dining
room, bed)
Object

RFID ID
Object Name
(electric socket)
Activity

3-axis
accelerometer
[-2g, +2g]
Activity
(moving left,
right)
Height

Ultra-sonic
sensor

[30kPa,129kPa]

Height fromthe
ground
Temperature LM35 [20C,100C]
Ambient
temperature
Fire Fire sensor -
Detection of fire
near the baby
SMS GSM -
Alarming the
baby is in danger

For this reason, we present only one single unit of sensor
nodes, which collects multiple types of information. The nature
of information interaction involved in sensor fusion can be
classified as competitive, complementary, and cooperative
fusion. In competitive fusion, each sensor provides equivalent
information about the process being monitored. In
complementary fusion, sensors do not depend on each other
directly, as each sensor captures different aspects of the
physical process. The measured information is merged to form
a more complete picture of the phenomenon. Cooperative
fusion of the two sensors enables recognition of the activity that
could not be detected by each single sensor. Due to the
compounding effect, the accuracy and reliability of cooperative
fusion is sensitive to inaccuracies in all simple sensor
components used. In this paper, we select the cooperative
fusion model to combine information from sensors to capture
data with improved reliability, precision, fault tolerance, and
reasoning power to a degree that is beyond the capacity of each
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
45
sensor. The main contributions of this paper over the earlier
previous work are
1) To extend the method to work with arbitrary every day
activities not just walking by improving the feature selection
and recognition procedure.
2) To performevaluation on a large (50 h) dataset recorded
fromreal life activities.
3) To have studied ten divers subjects: 16, 17, 20, 25, 27
months-old baby boys and 21, 23, 24, 26, 29 months-old girls.
4) To employ a barometric pressure sensor for improving
upon the previous algorithms.
The proposed method classified daily physical activity of
children by a diaper worn device consisting of a single triaxial
accelerometer and a ultra-sonic sensor. We demonstrate our
improvements in comparison to the accuracy results of only a
single-wearable device and multiple feature sets to find an
optimized classification method.
II. METHODS
A. Sensor Device

In order to recognize daily activities, we adopt multiple
sensors, as shown in Table I, as follows:
1) A 3-axis accelerometer measures the movement;
2) An absolute ultra-sonic sensormeasures absolute pressure
enabling a measurement of a distance between the ground and
the wearable sensor device;
3) A radio-frequency identification (RFID) (Skye Module M1-
mini) is selected to read/write tags and smart labels, which has
compatibility with most industry standard 13.56 MHzs.
We used the MEMS ADC that is a 3-axis accelerometer for
applications requiring high performance with low power
consumption. It consists of three signal-processing channels
where it is low-pass filtered and communicates with the
processing layer based on SPI bus that is a full duplex
synchronous 4-wire serial interface. We also used the as a ultra-
sonic sensor that measures absolute distance to measure
distance between the ground and the sensor. The distance and
temperature output data are calibrated and compensated
internally. The sensor Communicates with the processing layer
through an SPI bus. The Skye Module M1-mini has a
read/write distance that is typically greater than or equal to two
inches for an ISO15693 RFID inlay. The sensor allows us to
recognize objects and space that may cause dangerous
situations. Finally, we developed the prototype wearable sensor
device (size of 65mm 25 mm) including the dual-core
processor and sensors as shown in Fig. 1.

III. BLOCK DIAGRAM

In the baby side we will be having LPC2148, LM35, Fire
sensor, RFID reader, MEMS sensor and Ultra-sonic sensor a 3-
axis accelerometer measures the movement. An absolute ultra-
sonic sensormeasures absolute pressure enabling a
measurement of a distance between the ground and the
wearable sensor device. An absolute ultra-sonic
sensormeasures absolute pressure enabling a measurement of a
distance between the ground and the wearable sensor device.
We also used the as a ultra-sonic sensor that measures absolute
distance to measure distance between the ground and the
sensor. The temperature sensor which measure the body
temperature at regular intervals of time. The fire sensor which
helps baby to be involved in any fire accidents. When the baby
was in danger a sms will sent to their parents so they can come
for his rescue. These all devices can be embedded on single
chip and they can be attached on a belt so that it will not be
uncomfortable to the baby.




Fig.1. Prototype of the wearable sensor device and the RFID reader.
.
At the receiver section for monitoring purpose we will be
having a computer attached with a zigbee module so as soon as
the data is transmitted fromthe baby that will be displayed in
the hyper terminal. For every change in the activity the
corresponding activity name will be displayed on the screen.

Fig.2. Prototype of the receiver end for monitoring.
.
Here in this proposal RFID tags are provided for each roomand
any dangerous objects like power sockets e.t.c, so when the
active cards are detected by the RFID reader in the baby unit
then an alarming message will be displayed on the computer
hyper terminal so after seeing this a man can go and avoid the
baby from that danger preventing injuries and some fedial
injuries on the body. A message will also be sent to the chosen
person by using the gsm module present with the baby. Here in
this proposal we are giving a RFID tag to each and every room
in the house so as to monitor the baby position in the house
which makes it easy for the person to go for the rescue.



Fig.3. Prototype of the waist belt


The above figure shows the prototype of the waist belt we are
using in the proposal, this belt is properly placed on to the baby
so that it will not be uncomfortable for the baby.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
46
D. Activity Recognition

After the preparation of the all the mechanical possibilities are
ready and all the circuitry is complete then we will proceed to
the working of accelerometer and ultra-sonic sensor to
recognize the activities. A classification procedure separates the
childs activities from all other primitive features. The activity
recognition algorithm should be able to recognize the
accelerometer signal pattern corresponding to every activity.
Standing still: If the baby is standing still the position of the
the plane of the mems sensor will be perpendicular to x-axis
respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has no
task to perform. Then the data is transmitted i.e. nothing but
some milli volts voltage will be generated and it is compared
according to the program. Then the data will be transmitted to
the receiver section and displayed in the hyper terminal for
monitoring purpose with the help of zigbee modulus.
Moving left:If the baby is moving let the position of the the
plane of the mems sensor will be perpendicular to negative y-
axis respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has no
task to perform. Then the data is transmitted i.e. nothing but
some milli volts voltage will be generated and it is compared
according to the program. Then the data will be transmitted to
the receiver section and displayed in the hyper terminal for
monitoring purpose with the help of zigbee modulus.
Moving right:If the baby is moving right the position of the the
plane of the mems sensor will be perpendicular to y-axis
respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has no
task to perform. Then the data is transmitted i.e. nothing but
some milli volts voltage will be generated and it is compared
according to the program. Then the data will be transmitted to
the receiver section and displayed in the hyper terminal for
monitoring purpose with the help of zigbee modulus.
Moving forward:If the baby is moving forward the position of
the the plane of the mems sensor will be perpendicular to
negative z-axis respectively. The mems sensor will sense all the
positions accurately. Here in this activity the ultra-sonic sensor
has no task to perform. Then the data is transmitted i.e. nothing
but some milli volts voltage will be generated and it is
compared according to the program. Then the data will be
transmitted to the receiver section and displayed in the hyper
terminal for monitoring purpose with the help of zigbee
modulus.
Moving backward:If the baby is moving backward the position
of the the plane of the mems sensor will be perpendicular to z-
axis respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has no
task to perform. Then the data is transmitted i.e. nothing but
some milli volts voltage will be generated and it is compared
according to the program. Then the data will be transmitted to
the receiver section and displayed in the hyper terminal for
monitoring purpose with the help of zigbee modulus.
Climbing up:If the baby is climbing up the position of the the
plane of the mems sensor will be perpendicular to x-axis
respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has task
to perform. The ultra-sonic sensor will move up for a fixed
distance indicating that the baby is climbing up.Then the data is
transmitted i.e. nothing but some milli volts voltage will be
generated and it is compared according to the program. Then
the data will be transmitted to the receiver section and
displayed in the hyper terminal for monitoring purpose with the
help of zigbee modulus.Then the data is transmitted i.e. nothing
but some milli volts voltage will be generated and it is
compared according to the program. Then the data will be
transmitted to the receiver section and displayed in the hyper
terminal for monitoring purpose with the help of zigbee
modulus.
Climbing down:If the baby is climbing down the position of
the the plane of the mems sensor will be perpendicular to x-axis
respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has a
task a to perform. The ultra-sonic sensor will move up down for
a fixed distance indicating that the baby is climbing down.
Then the data is transmitted i.e. nothing but some milli volts
voltage will be generated and it is compared according to the
program. Then the data will be transmitted to the receiver
section and displayed in the hyper terminal for monitoring
purpose with the help of zigbee modulus.
Sitting down:If the baby is Sitting down the position of the the
plane of the mems sensor will be perpendicular to negative x-
axis respectively. The mems sensor will sense all the positions
accurately. Here in this activity the ultra-sonic sensor has task a
to perform. Then the data is transmitted i.e. nothing but some
milli volts voltage will be generated and it is compared
according to the program. Then the data will be transmitted to
the receiver section and displayed in the hyper terminal for
monitoring purpose with the help of zigbee modulus.

TABLE II
SENSOR POSSITION ACCOURDING TO ACTI VITIES

ACTIVITY
ACCELEROMETER
POSITION
ULTRA-
SONIC
POSITION
Standing still Perpendicular to x-axis -
Moving left
Perpendicular to
negative y-axis
-
Moving right Perpendicular to y-axis -
Moving
forward
Perpendicular to
negative z-axis
-
Moving
backward
Perpendicular to z-axis -
Climbing up Perpendicular to x-axis
Will move
upwards
Climbing
down
Perpendicular to x-axis
Will move
downwards

The above sensors data will stored at regular intervals of time
so as to analyze the baby behavior. So as we can take
preventive measures an avoiding those things which greatly
effecting the baby. Here temperature sensor plays an important
role in sensing the temperature of the baby, the abnormal
condition of the temperature is programmed between 28F to
100 F respectively. So these are the abnormal which are taken
into consideration. Here a fire sensor is provided so as rescue
the baby if he is involved in any fire accident or if he is near to
fire. Each and every device which is harmful to the baby in the
house is given with an rfid card so as to prevent the baby from
injuries. If in any situation the baby is near to these devices the
alarm will be blown and baby will rescued from all the
dangerous articles in the home. For smaller devices a RFID pil
can be provided as to facilitate comfortable working. So from
the proposal we are able to provide a secure home for the
babies below 2 years of age. This systemis more accurate
compared with all the present existing systems in the market.
This proposal has a greatest advent in present day world which
makes a great sense.
III. RESULTS
A. Experimental Setup
We have volunteered two families to conduct the experiment in
40 mbed room and 20 mgardens. In a controlled environment
setting, all data were collected fromten babies whoare 16, 17,
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
47
20, 25, 27 months-old baby boys and 21, 23, 24, 26,and 29
months-old baby girls. A supporter who observed the
experiments annotates raw data by clicking buttons or typing
name of the activities through the monitoring application. All
experiments were performed in a real home environment
consisting of one wearable sensor device for the child and the
monitoring application operated on a laptop computer. Baby
boys wiggled more than baby girls. They squirmed more and
get restless on the floor, and crawl over longer distances. While
the average boy did not move around much more than the
typical girl, the most active children were almost always boys,
and the least active children were girls.

The following are the results of different activities of a child in
different conditions.
i) This is the result when the baby is moving forward.


Fig.4. baby in forward motion

ii) This is the result when the baby is moving backward.


Fig.5. baby in backward motion

iii) This is the result when the baby is climbing down.


Fig.5. baby isclimbing down






iv) This is the result when the baby is climbing up.


Fig.6. baby is climbing up.

v) This is the result when the baby is moving right.


Fig.7. baby is moving right

vi) This is the result when the baby is moving left.


Fig.8. baby is moving left

vii) This is the result when the baby is baby is danger, sms will
be sent to the parents.



Fig.9. sms alert to the parents






vii) This is the result at the receiver section
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
48


Fig.9. Result at the receiver section

IV. CONCLUSION
This paper has presented the activity recognition method for
Children using only a triaxial accelerometer and anultra sonic
sensor. Results showed that using anultra-sonic sensor could
reduce the incidence of false alarms. The early warning system
will give the parents enough time to save their babies, and,
thus, minimize any instances of falling accidents or sudden
infant death syndrome.

ACKNOWLEDGMENT
The authors would like to thank all the volunteers MATHAJ I
REHABITATION CHILD CENTRE who have supported
greatly in our experiments. They would also like to thankK.
Dhanunjayafor previewing our manuscript.

REFERENCES
[1] A. M. Khan, Y.-K. Lee, S. Y. Lee, and T.-S. Kim. (2010,
Sep.). A triaxial
Accelerometer-based physical-activity recognition via
augmentedsignal features and a hierarchical recognizer. Trans.
Info. Tech. Biomed., [Online]. 14(5), pp. 11661172, Available:
http://dx.doi.org/10.1109/ TITB.2010.2051955 426 IEEE
JOURNAL OF BIOMEDICAL AND HEALTH
INFORMATICS, VOL. 17, NO. 2, MARCH 2013
[2] N. Li, Y. Hou, and Z. Huang. (2011). A real-time algorithm
based on triaxial accelerometer for the detection of human
activity state in Proc. 6th Int. Conf. Body Area Netw., ser.
BodyNets 11. ICST, Brussels, Belgium, Belgium: ICST
(Institute for Computer Sciences, Social-Informatics and
Telecommunications Engineering), [Online]. pp. 103106,
Available: http://dl.acm.org/citation.cfm?id=2318776.2318801
[3] A. G. Bonomi. (2011). Physical activity recognition using a
wearableaccelerometer. in Proc. Sens. Emot., ser. Philips
Research Book Series, J. Westerink, M. Krans, and M.
Ouwerkerk, Eds., Springer Netherlands,[Online]. vol. 12, pp.
4151. Available: http://dx.doi.org/10.1007/978-90-481-3258-
4_3
[4] S. Boughorbel, J. Breebaart, F. Bruekers, I. Flinsenberg,
and W.ten Kate. (2010). Child-activity recognition from multi-
sensor data.in Proc. 7th Int. Conf. Methods Tech. Behav. Res.,
ser. MB 10.New York, NY, USA: ACM, [Online].pp.38:1
38:3, Available:http://doi.acm.org/10.1145/1931344.1931382
[5] A. Fleury, M. Vacher, and N. Noury, Svm-based
multimodal classification
of activities of daily living in health smart homes: Sensors,
algorithms,and first experimental results, IEEE Trans. Inf.
Technol. Biomed., vol. 14,no. 2, pp. 274283, Mar. 2010.
[6] C. Zhu andW. Sheng. (2009).Multi-sensor fusion for human
daily activityrecognition in robot-assisted living. in Proc. 4th
ACM/IEEE Int. Conf.Human Robot Int., ser. HRI 09. New
York, NY, USA: ACM, [Online].pp. 303304, Available:
http://doi.acm.org/10.1145/1514095.1514187
[7] L. Atallah, B. Lo, R. King, and G.-Z. Yang, Sensor
positioning for activity recognition using wearable
accelerometers, IEEE Trans. Biomed.Circuits Syst., vol. 5, no.
4, pp. 320329, Aug. 2011.
[8] L. Bao and S. S. Intille. (2004). Activity recognition from
user-annotatedacceleration data. Pervas. Comput., [Online]. pp.
117, Available:
http://www.springerlink.com/content/9aqflyk4f47khyjd
[9] R. Luo and M. Kay, Multisensor integration and fusion in
intelligentsystems, IEEE Trans. Syst., Man, Cybern., vol. 19,
no. 5, pp. 901931,Sep./Oct. 1989.
[10] R. R. Brooks and S. S. Iyengar, Multi-sensor Fusion:
Fundamentals andApplications with Software. Englewood
Cliffs, NJ: Prentice-Hall, 1998.
[11] R. Luo, C.-C. Yih, and K. L. Su, Multisensor fusion and
integration:Approaches, applications, and future research
directions, IEEE SensorsJ., vol. 2, no. 2, pp. 107119, Apr.
2002.
[12] N. Ravi, N. Dandekar, P. Mysore, and M. L. Littman.
(2005). Activityrecognition from accelerometer data. in Proc.
17th Conf. Innovated. Appl.Artif. Intel., AAAI Press, [Online].
vol. 3, pp. 15411546,
Available:http://dl.acm.org/citation.cfm?id=1620092.1620107
[13] B. Baas and S. U. D. of Electrical Engineering. (1999). An
Approachto Low-Power, High-Performance, Fast Fourier
Transform ProcessorDesign. Stanford, CA: Stanford
University. [Online].
Available:http://books.google.com/books?id=61KFHwAACAA
J
[14] R. Kohavi. (1995). A study of cross-validation and
bootstrap foraccuracy estimation and model selection. in Proc.
14th Int. JointConf. Artif. Intel. Francisco, CA, USA: Morgan
Kaufmann PublishersInc., [Online]. vol. 2, pp. 11371143,
Available: http://dl.acm.org/citation.cfm?id=1643031.1643047
[15] J. R. Quinlan, C4.5: Programs for Machine Learning
(Morgan Kaufmann
Series in Machine Learning), 1st ed. San Mateo, CA: Morgan
Kaufmann,Oct. 1992.
[16] J. C. Platt. (1999). Fast training of support vector machines
usingsequential minimal optimization, Advances in kernel
methods.,B. Scholkopf, C. J. C. Burges, and A. J. Smola, Eds.,
Cambridge, MA: MIT Press. [Online]. pp. 185208, Available:
http://dl.acm.org/citation.cfm?id=299094.299105
[17] S. S. Keerthi, S. K. Shevade, C. Bhattacharyya, and K. R.
K. Murthy.(2001, Mar.). Improvements to platts SMO
algorithmfor SVM classifierdesign. Neur. Comput., [Online].
13(3), pp. 637649,
Available:http://dx.doi.org/10.1162/089976601300014493
.

.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
49
An Efficient Way of Classifying and
Clustering Documents Based on SMTP

U.Umamaheswari
[1]
, Mr.G.Shivaji Rao M.E.
[2]
,
M.E scholar, Assistant Professor,
[1]
Department of Computer Science and Engineering,
[2]
Department of Computer Science and Engineering,
Sree Sowdambika College of Engineering, Sree Sowdambika College of Engineering,
Aruppukottai-626 134, Aruppukottai-626 134,
Tamil Nadu, Tamil Nadu,
India. India.
13umamaheswari@gmail.com shivajirao88@gmail.com

Abstract
In text processing, the similarity measurement is the important process. It measures the similarities between
the two documents. In this project we proposed the new similarity measurement. The computation of similarity
measurement is based on the feature of two documents. Our proposed system contains three case to compute the
similarity. The three cases are, both two documents contains features, only one document contains feature, there is no
feature into the two documents. In first case, the similarity is increased when the differences of feature value is
decreased between the two documents. Then the given differences are scaled. In second case, fixed value is given to
the similarity. In third cases there is no contribution to the similarity. Finally our proposed measure method achieves
the better performance compared than other measurement methods.

Index TermsDocument classification, document clustering, entropy, accuracy, classifiers, clustering algorithms

I. INTRODUCTION
Text processing plays an important role
in information retrieval, data mining, and web
search. A document is any content drawn up or
received by the Foundation concerning a matter
relating to the policies, activities and decisions
falling within its competence and in the
framework of its official tasks, in whatever
medium (written on paper or stored in electronic
form , including e-mail, or as a sound, visual or
audio-visual recording). The term classification
means the allocation of an appropriate level of
security (as confidential or restricted) to a
document the unauthorised disclosure of which
might prejudice the interests of the Foundation,
the EU or third parties. Documents are
confidential when their unauthorised disclosure
could harm the essential interests of an
individual, the Foundation or the EU.
Documents are restricted when their
unauthorised disclosure could be
Disadvantageous to the Foundation, the EU or a
third party. Documents are restricted when their
unauthorised disclosure could be
disadvantageous to the Foundation, the EU or a
third party. The termoriginator means the duly
authorised author of a classified document. The
term downgrading means a reduction in the level
of classification. The term declassification
means the removal of any classification.

RULES FOR CLASSIFICATION
Foundation documents that are not
public shall be classified in one of the following
categories: confidential or restricted. Criteria and
guidance for classification are set out in Annex 1
to this decision. The classification of a
document shall be decided by the originator
based on these rules. All classified documents
shall be recorded in a register of classified
documents. Applications for access to classified
documents shall be examined by the Director. If
a classified document is to be made available in
response to a request from a member of the
public, it shall be first declassified by a decision
of the Director. Documents shall be classified
only when necessary. The classification shall be
clearly indicated and shall be maintained only as
long as the document requires protection. The
classification of a document shall be determined
by the level of sensitivity of its contents.
Classification of documents shall be
periodically reviewed. By request of the
Document Management Officer (DMO), the
originator of a document shall indicate if that
document or information may be downgraded
and declassified. Where a document or
information is declassified, details shall be
recorded in the register and the document shall
be archived appropriately. Where classification
is retained, details of the review shall be entered
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
50
in the register. All classified documents shall be
retained in a manner that ensures they are not
disclosed to unauthorised individuals. Security
and access levels to classified documents shall
be recorded in the register by the DMO. Either
the originator or the DMO may retain classified
documents. Where the originator retains
documents, they shall be physically safeguarded
as specified by the DMO.
Individual pages, paragraphs, sections,
annexes, appendices, attachments and enclosures
of a given document may require different
classifications and shall be classified
accordingly. The classification of the document
as a whole shall be that of its most highly
classified part. The originator shall indicate
clearly at which level a document should be
classified when detached from its enclosures.
The classification shall appear at the top and
bottom centre of each page, and each page shall
be numbered. Each classified document shall
bear a reference number and a date. All annexes
and enclosures shall be listed on the first page of
a document classified as confidential. The
classification shall be shown on restricted
documents by mechanical or electronic means.
Classification shall be shown on confidential
documents by mechanical means or by hand or
by printing on pre-stamped, registered paper.
DOCUMENT CLASSIFICATION
Document classification or document
categorization is a problemin library science,
information science and computer science. The
task is to assign a document to one or more
classes or categories. This may be done
"manually" (or "intellectually") or
algorithmically. The intellectual classification of
documents has mostly been the province of
library science, while the algorithmic
classification of documents is used mainly in
information science and computer science. The
problems are overlapping, however, and there is
therefore also interdisciplinary research on
document classification.
The documents to be classified may be
texts, images, music, etc. Each kind of document
possesses its special classification problems.
When not otherwise specified, text classification
is implied. Documents may be classified
according to their subjects or according to other
attributes (such as document type, author,
printing year etc.). In the rest of this article only
subject classification is considered. There are
two main philosophies of subject classification
of documents: The content based approach and
the request based approach.

II. RELATED WORK

The documents are represented as
vectors, each elements of vectors indicates the
corresponding feature values of the documents.
In existing systemthe feature values are termed
as frequency, relative term frequency and tf-idf
(term frequency and inverse document
frequency). The document size is large and most
of the vector value is zero in this system. In
existing system non symmetric measure is used
to measure the similarities. Canberra distance
metric is one of the existing methods which are
used to measure the similarity. This method is
applicable when the vector elements are always
non-zero. In existing system cosine similarity
measure is used to measure the similarities
between the two documents. It takes cosine of
angle between the two vectors. The phase based
similarity measure is used in one of the existing
system. So, we are our some problem in
following as the document is large in existing
system so it takes more memory to store the
documents. It contains zero vector values so it
makes severe challenges to measure the
similarity. It is not efficient and does not provide
accurate results. The performance is low. It takes
more time to produce the result.
In our proposed system new measure is used to
measure the similarities between the two
documents. This new measure is symmetric
measure. The similarities between the two
documents are measured with respect to the
features. Three feature cases are used to
measure the similarities. In first case the two
documents contains features value, in second
case, only one document contains the feature
value and in third case there is no feature value
in both documents. This measure is applied in
many applications such as single label and multi
label classification, clustering and so on. In our
proposed system we k-NN based single label
classification (SL-KNN) and k-NN base multi
label classification (ML-KNN) for classification
purpose. The documents are formed as cluster in
our concept for this purpose we used
Hierarchical Agglomerative Clustering (HAC).
It is the one type of k means clustering
algorithm. We three type of data sets such as
WebKB, Reuters 8 and RCV1. These data sets
are stored in the form of XML. So, In our project
improve following as It achieves better
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
51
performance compared than other measure, It
provides efficient results, It achieves accuracy in
similarity measurement, It take less time for
similarity measure.

III.PROPOSE SIMILARITY MEASURE

In our proposed system new
measure is used to measure the similarities
between the two documents. This new measure
is symmetric measure. The similarities between
the two documents are measured with respect to
the features. Three feature cases are used to
measure the similarities. In first case the two
documents contains features value, in second
case, only one document contains the feature
value and in third case there is no feature value
in both documents. This measure is applied in
many applications such as single label and multi
label classification, clustering and so on. In our
proposed system we k-NN based single label
classification (SL-KNN) and k-NN base multi
label classification (ML-KNN) for classification
purpose. The documents are formed as cluster in
our concept for this purpose we used
Hierarchical Agglomerative Clustering (HAC).
It is the one type of k means clustering
algorithm. We three type of data sets such as
WebKB, Reuters 8 and RCV1. These data sets
are stored in the formof XML.

Fig: System architecture
The following properties, among
other ones, are preferable for a similarity
measure between two documents:
1. The presence or absence of a feature is
more essential than the difference
between the two values associated with
a present feature.
Let as Consider two features wi
and wj and two documents d1 and d2.
wi =d2(some relationship)
wi d1(no relationship)
In this case, d1 and d2 are
dissimilar in terms of wi. If wj appears in both
d1 and d2. Then wj has some relationship with
d1 and d2 simultaneously. In this case, d1 and
d2 are similar to some degree in terms of wj.
For the above two cases, it is reasonable to say
that wi carries more weight than wj in
determining the similarity degree between d1
and d2.
2. The similarity degree should increase
when the difference between two non-
zero values of a specific feature
decreases.
3. The similarity degree should decrease
when the number of presence-absence
features increases.
4. Two documents are least similar to each
other if none of the features have non-
zero values in both documents. Let
d1 =< d11, d12, . . . , d1m > and
d2 =< d21, d22, . . . , d2m >.
If d1id2i =0,
d1i +d2i > 0
for 1 i m, then d1 and d2 are
least similar to each other. As mentioned
earlier, d1 and d2 are dissimilar in terms
of a presence-absence feature.
5. The similarity measure should be
symmetric. That is, the similarity degree
between d1 and d2 should be the same
as that between d2 and d1.
6. The value distribution of a feature is
considered, i.e., the standard deviation
of the feature is taken into account, for
its contribution to the similarity between
two documents. A feature with a larger
spread offers more contribution to the
similarity between d1 and d2.

IV. METHODOLOGY
1. DATA SET SELECTION
Here we used three types of data
sets. They are WebKB, Reuters 8 and RCV1.
These data sets are stored in the formof HTML
and text or SGM. Each data set is divided into
two types. They are training and testing data set.
WebKB contains web pages as the document
which is collected by the world wide knowledge
base. It does not contain predefined training set
and testing set. So we randomly divide these
data sets as training and testing set. Reuters data
set contains predesigned training set and testing
sets. If the resultant data set contains 71% of
training set and 29% of testing set then it is
called as Reuters 8. RCV1 also contains
predesigned training and testing data sets.
2. CLUSTERING
After the data set collection then we
perform the cluster formation. Cluster contains
the more than one data set. For cluster formation
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
52
we used k-means algorithm. In our proposed
system, we used HAC (hierarchical
agglomerative clustering). It is the one type k
means clustering algorithm and it used bottom
up approach for clustering. The HAC is mostly
used to measure the performance.
3. PERFORM CLASSIFICATION
After form the cluster then we performs
the classifications. We used SL-kNN and ML-
kNN method is used to perform the
classifications. SL-KNN is used for only one
category and ML-KNN is used for more than
two categories. These are used to measure the
similarity between the documents. It gives best
result for similarity. Classification is used to
retrieve process. It increases the fast of retrieve
the data. It easily identifies the similarity
between the documents. It gives related result
efficiently.

4. FIND SIMILARITIES
After select the classification
methods then we finds the similarities between
the data sets. The similarity measurement is
based on the features of documents. It finds the
word weight and distance to measure the
similarities. Similarity measure has three
properties they are, the presence and absence
feature is more important than the difference of
two values that are presented in the feature, the
similarity degree is increased when the
difference between the non zero values of
feature is decreased, the similarity degree is
decreased when the presence and absence of
feature is increased.
5. PERFORMANCE COMPARISON
After performall the process then
we compared the performance of process. In our
proposed system, the performance is high
compared than all existing system.
V.CONCLUSION
The main concept of our proposed system is
to find the similarity between the documents.
For similarity measure we used KNN based
single label classification and KNN based multi
label classification. Here we collect three data
sets such as WEBKB, reuters-8 and RCV1.
Classification is used for retrieve process and it
increases the fast of retrieve the data. It easily
identifies the similarity between the documents.
It gives related result efficiently. Here we used
HAC for clustering and it is the one type of k
means clustering algorithm. It also measures the
performance. In our proposed system, the data
sets are stored in the form the XML. Here we
used tf-idf (term frequency and inverse
document frequency) to retrieve the document
terms and It measures the similarity by
calculating the weight. Here we used
optimization and it identifies the no of cluster
that we formed in the document. It increases the
accuracy of similarity measurement.
VI.REFERENCES
1. Yung-Shen Lin, J ung-Yi J iang, and Shie-J ue Lee,
Member, IEEE, A Similarity Measure for Text
Classification and Clustering, in IEEE Transactions
On Knowledge And Data Engineering, Vol. 26, No. 7,
J uly 2014 1575
2. D. Cai, X. He, and J. Han, Document clustering using
locality preserving indexing, IEEE Trans. Knowl. Data
Eng., vol. 17, no. 12, pp. 16241637, Dec. 2005.
3. H. Chim and X. Deng, Efficient phrase-based
document similarity for clustering, IEEE Trans.
Knowl. Data Eng., vol. 20, no. 9, pp. 12171229, Sept.
2008.
4. S. Clinchant and E. Gaussier, Information-based
models for ad hoc IR, in Proc. 33rd SIGIR, Geneva,
Switzerland, 2010, pp. 234241.
5. K. M. Hammouda and M. S. Kamel, Hierarchically
distributed peer-to-peer document clustering and cluster
summarization, IEEE Trans. Knowl. Data Eng., vol.
21, no. 5, pp. 681698, May 2009.
6. Y. Zhao and G. Karypis, Comparison of agglomerative
and partitional document clustering algorithms, in
Proc. Workshop Clustering High Dimensional Data Its
Appl. 2nd SIAM ICDM, 2002, pp. 8393.
7. T. Zhang, Y. Y. Tang, B. Fang, and Y. Xiang,
Document clustering in correlation similarity measure
space, IEEE Trans. Knowl. Data Eng., vol. 24, no. 6,
pp. 10021013, J un. 2012.
8. M. L. Zhang and Z. H. Zhou, ML-kNN: A lazy
learning approach to multi-label learning, Pattern
Recognit., vol. 40, no. 7, pp. 20382048, 2007.
9. Strehl and J . Ghosh, Value-based customer grouping
fromlarge retail data-sets, in Proc. SPIE, vol. 4057.
Orlando, FL, USA, Apr. 2000, pp. 3342.
10. F. Sebastiani, Machine learning in automated text
categorization, ACM CSUR, vol. 34, no. 1, pp. 147,
2002.
11. T. W. Schoenharl and G. Madey, Evaluation of
measurement techniques for the validation of agent-
based simulations against streaming data, in Proc.
ICCS, Krakw, Poland, 2008.
12. D. D. Lewis, Y. Yang, T. Rose, and F. Li, RCV1: A
new benchmark collection for text categorization
research, J . Mach. Learn. Res., vol. 5, pp. 361397,
Apr. 2004.
13. V. Lertnattee and T. Theeramunkong,
Multidimensional text classification for drug
information, IEEE Trans. Inform. Technol. Biomed.,
vol. 8, no. 3 pp. 306312, Sept. 2004.
14. S.-J . Lee and C.-S. Ouyang, A neuro-fuzzy system
modeling with self-constructing rule generation and
hybrid SVD-based learning, IEEE Trans. Fuzzy Syst.,
vol. 11, no. 3, pp. 341353, J un. 2003.
15. G. Amati and C. J . V. Rijsbergen, Probabilistic
models of information retrieval based on
measuring the divergence from randomness,
ACM Trans. Inform. Syst., vol. 20, no. 4, pp. 357
389, 2002.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
53

An Efficient Way of Detecting a Numbers in Car
License Plate Using Genetic Algorithms













Abstract - To detect the numbers and characters inside the
license plate using image processing and genetic algorithm
(GA). For this Number plate detection many algorithms are
used. But in my project mainly focusing on the genetic
algorithm for provide perfect accuracy compare to any other
systems. This paper describes a detection method in which
the vehicle plate image is captured by the cameras and the
image is processed to get the plates numbers and
characters. The system is implemented using MATLAB and
various images are processed with to verify the distinction
of the proposed system.

Index Terms - Genetic algorithm (GA), Image processing,
License plate (LP), Number plate localization, Perfect
accuracy.

I.INTRODUCTION
Nowadays number of automobiles grows
quickly, the traffic problems arise as well, for example car
robbery, over speeding and moving on the red light. To
avoid these problems an efficient real time working
vehicle identification system is needed. Most usually
suitable technique is license plate (LP) detection based on
image processing by capturing license plates using
cameras. All the implemented techniques can be
classified according to the selected features. Color
information based systems have been built to detect
specific plates having fixed colors. Shape- based
techniques were developed to detect the plate based on its
rectangular shape. Edge-based techniques were also
implemented to detect the plate based on the high density
of vertical edges inside it. GAs has been used
infrequently because of their large computational needs.
Variety of research has been tried at different levels
under some constraints to minimize the search space of
genetic algorithms (GAs). Researchers in based their GA
on pixel color features to segment the image depending on
stable colors










followed by shape dependent policy to identify the plates
area. In, GA was used to search for the best fixed
rectangular area having the same texture features. GA
was used in to identify the LP symbols not to detect the
LP.
Detecting license character and at the same
time differentiating it from similar patterns based on the
geometrical relationship between the symbols
constituting the license numbers are selected approach
in this research. Consequently, a new approach genetic
algorithm is initiate in this paper that detects LP symbols
without using any information linked with the plates
external shape or interior colors to allow for the detection
of the license numbers in case of shape or color
distortion either physically or due to capturing
conditions. Further processes are explained in the next
sections.
II.PROPOSED TECHNIQUE
The proposed system is comprised of two
phases: image processing phase and GA phase. Each
phase is composed of many steps. The Fig. 1 depicts the
various image processing steps that finally produce
image objects to the GA portion. GA selects the best LP
symbol locations depending on the input geometric
relationship matrix (GRM).

III.IMAGE PROCESSI NG PHASE
In this phase, an input color image is used to a
sequence of processes to extract the relevant 2-D objects
that may represent the symbols. It has different stages, as
depicted in Fig. 1.
A. Color image to Grayscale conversion
he input image is used as a color image to
bring other information relevant to the concerned
vehicle. Color (RGB) to grayscale (gs) conversion is
S.Sudha M.E.,
Assistant Professor
Department of Computer Science and Engineering
Sree sowdambika College of Engineering Aruppukottai,
Tamilnadu St, India
sudhajiin@yahoo.co.in
C.Subha
M.E Computer Science
Sree Sowdambika College Of Engineering
Aruppukottai Tamilnadu St, India
mirthula31@gmail.com

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
54



Fig. 1. Overall systemFlowchart for localization of LP symbols.

performed using the standard NTSC method by removing the
hue and saturation information while holding the luminance
as follows:

gs=0.299*R+0.587*G+0.114*B (1)




Fig. 2. Converted grayscale image.

B. Grayscale to Binary Using Dynamic Adaptive Threshold
Converting the input image into a binary image is
one of the most important stages in localizing LPs to
overcome the illumination problems. In my system, a local
adaptive threshold technique has been implemented to
determine the threshold at each pixel depending on the
average gray level. This process as shown in Fig.3
C. Morphological Operations
Morphological operations, like dilation and
erosion, are important processes needed for pattern
recognition systems to eliminate noisy object.



Fig. 3. (a) Converted binary image for image in Fig. 2, using Otsus
method. (b) Car image with variable illumination. (c) Output when using
Otsus method for image in (b). (d) Output when applying local adaptive
threshold method for same image in (b).

In LP detection, closing and opening operations are applied
to fill noisy holes and remove objects.
Dilation:
This is the b asic operators in the part of
morphology. It is usually applied to binary image, but there
are versions run on grayscale image. the basic effect of the
operator on a binary image is to progressively extend the
boundaries of regions of foreground pixels (ie, white pixels).
Applications of dilation for bridging gaps in an image. It can
remove unwanted information. Opening of an image is
erosion followed by a dilation using the same structuring
element. Shown in fig 4.
Erosion:
This is also very important operator for the
morphological operation. The basic effect of the operator on
a binary image is to erode away the boundaries of regions of
foreground pixels (ie, white pixels). Shown in fig 5.



Fig 4. Effect of dilation using 3x3 square structuring element


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
55


Fig 5. Effect of erosion using a 33 square structuring element Strip away a
layer of pixels froman object, shrinking it in the process.
D. Connected Component Analysis
CCA is one of the technique in image processing that
scans an image and groups pixels in components depends on
pixel connectivity. The result of this stage is an array of N
objects.

E. Size Filtering

The output of the CCA stage are filtered on the
basis of their widths W
obj
and heights H
obj
lie between their
respective thresholds as follows:

W
min
W
obj
W
max
and H
min
H
obj
H
max
(2)
H
min and
W
min are
the value below which a symbol
cannot be recognized (for example 8 pixels) and W
max
can be
set to the image width divided by the number of symbols.
H
max
is estimated as W
max
divided by the aspect ratio of the
used font. The result of this stage is an array of M objects.
The output of this stage is given in Fig. 6.



Fig. 6. M objects (64) output after size filtering of N objects in Fig. 5(2).

IV.GENETIC ALGORITHM
In this phase M objects are given to the input. This
phase is used to resolve the 2D compound object detection
problem. It contains many steps.

A. Chromosome Encoding
In chromosome encoding an integer encoding scheme
is selected and each gene assigned to an integer. Seven genes
are forming a chromosome as shown in fig. 7. An output is
extracted as a M objects.



Fig 7. Chromosome of seven genes for representation of Saudi license plate.

B. Fitness Function
Simple function of the fitness measure is used by
some genetic algorithms to select individuals. In this proposed
system fitness is used as the inverse of the estimated objective
distance between the prototype chromosome and the current
chromosome.

C. Selection Method
In this selection method, the stochastic universal
sampling (SUS) method each individual is formed to a
continuous segment of a line. Depending on the percentage of
individuals to be selected by a number of pointers over the
line.
D. Mutation Operators
This mutation method is used to remove unfit
members in genetic iterations. It can eliminate some features
of genetic material. To maintain the mating pool variety by
Gas ensures that the new parts of the search space. They are
two kinds of mutation operators.
1) Substitution Operator
2) Swap Operator
E. Crossover Operator
In genetic algorithm crossover operator is used to
produce new chromosome (offspring) by groups two
chromosomes (parents). This new chromosome is better than
the both parents if it takes the best characteristics from each of
the parents. In my project, the two parents chromosomes are
combined into the array Carray as shown in fig.8. In my
project USPS crossover operator is used.
F. Replacement Strategy
A lot of alternate strategies are used to replacing only a
portion of the population between generations. The most
frequent strategy is to probabilistically replace the unfit
individuals in the earlier generation. In elitist strategy the
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
56
greatest fit individuals of the previous generation are
appended to the recent population. In my proposed
system, the best 10% of the parents are selected and
appended to the offspring(90%) to produce the new
generation (100%).




Fig. 8. Proposed crossover operator steps.


V.CONCLUSION
In this paper describes the localization of license plate in a
efficient manner. For this purpose i used genetic
algorithm (GA). The license plate contain many unwanted
details. These are first remove by the image processing
phase and then localized by the genetic algorithm phase.
The results were encouraging and a new approach for
solving the LP detection problem relying only on the
geometrical layout of the LP symbols. Also, a flexible
system was introduced that can be simply adapted for
any LP layout by constructing its GRM matrix. The
proposed system possessed high immunity to changes
in illumination either temporarily or spatially. A high
percentage success rate was achieved with the aid of
the adaptability aspect of the GAs. A very important
attainment is overcoming most of the problems arising in
techniques based on CCAT by allowing the GA.
Moreover, an enhancement in the performance of the
developed GA was achieved by applying the new
USPS crossover operators, which greatly improved the
convergence rate of the whole system.


REFERENCES

[1] A. Ahmadyfard and V. Abolghasemi, Detecting license plate
using texture and color information, in Proc. Int. Symp.
Telecommun., 2008, pp. 804808.
[2] G. Li, R. Yuan, Z. Yang, and X. Huang, A yellow license plate
location method based on RGB model of color image and texture
of plate, in Proc. 2nd Workshop Digit. Media Its Appl. Museum
Heritages, 2007, pp. 4246.
[3] X. Shi, W. Zhao, Y. Shen, and O. Gervasi, Automatic license
plate recognition system based on color image processing, in
Lecture Notes on Computer Science, Berlin, Germany: Springer-
Verlag, 2005, vol. 3483, pp. 11591168.
[4] M. Deriche, GCC license plates detection and recognition using
mor- phological filtering and neural networks, Int J. Comp. Sci.
Info Security, vol. 8, no. 8, pp. 263269, Dec. 2010.
[5] O. Villegas, D. Balderrama, H. Dom

inguez, and V. Sanchez,


License plate recognition using a novel fuzzy multilayer neural
network, Int. J. Comput., vol. 3, no. 1, pp. 3140, 2009.
[6] S. H. M. Kasaei, S. M. M. Kasaei, and S. A. Monadjemi, A novel
morphological method for detection and recognition of vehicle
license plate, Amer. J. Appl. Sci., vol. 6, no. 12, pp. 20662070,
2009.
[7] A. Theja, S. J ain, A. Aggarwal, and V. Kandanvli, License plate
extraction using adaptive threshold and line grouping, in Proc.
ICSPS, J ul. 2010, vol. 1, pp. 211214.
[8] P. Tarabek, Fast license plate detection based on edge density
and integral edge image, in Proc. Int. Conf. Appl. Mach.
Intell. Inform.,2012, pp. 3740.
[9] V. Abolghasemi and A. Ahmadyfard, A fast algorithm for license
plate detection, in Proc. Int. Conf. Visual Inform. Syst., 2007,
vol. 4781, pp. 468477.
[10] S. Roomi, M. Anitha, and R. Bhargavi, Accurate license plate
local- ization, in Proc. Int. Conf. Comput. Commun. Electr.
Technol., 2011, pp. 9299.
[11] S. K. Kim, D. W. Kim, and H. J . Kim, A recognition of vehicle
license plate using a genetic algorithm based segmentation, in
Proc. Int. Conf. Image Process., 1996, vol. 1, pp. 661664.
[12] J . Xiong, S. Du, D. Gao, and Q. Shen, Locating car license plate
under various illumination conditions using genetic algorithm, in
Proc. ICSP,2004, vol. 3, pp. 25022505.
[13] Z. J i-yin, Z. Rui-rui, L. Min, and L. Yinin, License plate
recognition based on genetic algorithm, in Proc. Int. Conf.
Comput. Sci. Software Eng., Dec. 2008, vol. 1, pp. 965968.
[14] V. P. de Arau jo, R. D. Maia, M. F. S. V. DAngelo, and
G. N.R. DAngelo, Automatic plate detection using genetic
algorithm, in Proc. 6th WSEAS Int. Conf. Signal Speech Image
Process., Sep. 2006, pp. 4348.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
57
RF Controlled Sailing Robot for Oceanic Missions
R. Srikanth Kumar
1
, K. Dhanunjaya
2

1
PG Student, Department of ECE, ASCET, Gudur, Andhra Pradesh, India
2
Head of the Department, Department of ECE, ASCET, Gudur, Andhra Pradesh, India

Abstract:
Ocean exploration and navigational
research by supporting expeditions with computer
vision techniques have shown potential for sailing
robots development in order to make
measurements at the surface. Sailing robots
explores the science and technologies for the
identification of underwater features. Key
applications of sailing robot are measuring depth
and sensing metals under the water. An idea
presented is the robot that can sail on water,
controlled by laptop keypad. Other modules are
gps and camera.
Keywords: sailing robot, laptop, gps, camera
I. INTRODUCTION
For oceanic missions such as finding metals,
spy applications, measuring depth, etc., one has to
travel on the boat. But it is risky, because climate can
change suddenly on the ocean. For example cyclone
may occur suddenly. Also while spying, the opposite
one may fire and can result in death of driver
controlling the boat.
To overcome this, the boat has to be
designed in such a way that it can be operated
without a driver. So it has to use any means of
wireless communications. In this paper RF pro
wireless communication is presented.
II. RF WIRELESS COMMUNICATION
The commands for respective movements of
sailing robot are send from laptop keypad through RF
PRO wireless communication. The module used for
this purpose is RFSv4.3. It provides easy and flexible
wireless data transmission between devices. RFSv4.3
uses 2.4 GHz carrier frequency. On pressing keys on
laptop keypad respective action takes place. For
example on pressing 1 the sailing robot moves
forward. For this project two RFSv4.3 modules are
necessary. The data received from sensors is also
transmitted to laptop through these modules.


III. SENSORS
A. Depth sensor:

One of the applications of sailing
robot is to measure depth under water. For this an
ultrasonic sensor is used. The module used is
Ultrasonic ranging module HC-SR04. The basic
principle depends on echo reception. A short 10s
pulse is supplied to the trigger input. The module will
send out an 8 cycle burst of ultrasound at 40 KHz.
After hitting ground the pulse signal is reflected back.
Now the range can be calculated through the time
interval between sending trigger signal and receiving
echo signal. The formula is,
Distance =(time taken velocity of sound)/2
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
58
B. Obstacle Sensor:
Although the sailing robot is
monitored through camera, there may be chances of
collision with obstacles because of controller
inactive. So to avoid this an ultrasonic sensor is used
to detect the obstacles. When an obstacle is detected
the sailboat automatically stops.
C. Metal sensor:


To detect metals under the water, a
metal sensor is used. The metal sensor used is ID18
3008NA. It is water proof. This module is useful in
finding missing ships.
IV. GPS

As the climate on ocean is
unpredictable, there is a chance of missing of
sailboat. This may occur because of cyclone or some
other reasons. So a GPS module is used to detect the
sailboat when missed. The GPS module used is
Royaltek REB-1315S4. GPS module can also be used
for navigation purpose.
V. CAMERA
This module sends continuous
video to the laptop for continuous monitoring. The
camera used in this project can send video from 50 to
100 meters distance if it in line of sight with receiver.
This can also be used for spy applications.
VI. X-CTU TERMINAL
All the values received from
sensors can be displayed on laptop using X-CTU
terminal.


VII. REFERENCES
1. www.google.com
2. www.keil.com
3. www.nxp.com
4. www.robosoftsystems.co.in
5. www.royaltek.com
6. www.ideal.net.in







Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
59
Solar Power Satellites


K. Deepak K. Poojitha
IV B.Tech EIE IV B.Tech EIE
Sree Vidyanikethan Engineering College Sree Vidyanikethan Engineering College

Abstract
Solar Power Satellites (SPS)
converts solar energy in to micro waves
and sends that microwaves in to a beam to
a receiving antenna on the Earth for
conversion to ordinary Electricity. SPS is
a clean, large-scale, stable electric power
source. For SPS Wireless power
transmission is essential. WPT contains
microwave beam, which can be directed to
any desired location on Earth surface. This
beam collects Solar Energy and converts it
into Electrical Energy. This concept is
more advantageous than conventional
methods. The SPS will be a central
attraction of space and energy technology
in coming decades. It is not a pollutant but
more aptly, a man made extension of the
naturally generated electromagnetic
spectrum that provides heat and light for
our sustenance.
Keywords: Solar Power Satellites;
Microwaves;

1. Introduction
The new millennium has introduced
increased pressure for finding new renewable
energy sources. The exponential increase in
population has led to the global crisis such as
global warming, environmental pollution and
change and rapid decrease of fossil
reservoirs. Also the demand of electric power
increases at a much higher pace than other
energy demands as the world is industrialized
and computerized. Under these
circumstances, research has been carried out
to look in to the possibility of building a
power station in space to transmit electricity
to Earth by way of radio waves-the Solar


Power Satellites. Solar Power Satellites
(SPS) converts solar energy in to micro
waves and sends that microwaves in to a
beam to a receiving antenna on the Earth for
conversion to Ordinary Electricity. SPS is a
clean, large-scale, stable electric power
source. Solar Power Satellites is known by a
variety of other names such as Satellite
Power System, Space Power Station, Space
Power System, Solar Power Station, Space
Solar Power Station etc. One of the key
Technologies needed to enable the future
feasibility of SPS is that of Microwave
Wireless Power Transmission.
WPT is based on the energy transfer
capacity of microwave beam I e, energy can
be transmitted by a well-focused microwave
beam. Advances in Phased array antennas
and rectennas have provided the building
blocks for a realizable WPT system

1.1 Why SPS
Increasing global energy demand is
likely to continue for many decades.
Renewable energy is a compelling approach
both philosophically and in engineering
terms. However, many renewable energy
sources are limited in their ability to
affordably provide the base load power
required for global industrial development
and prosperity, because of inherent land and
water requirements. The burning of fossil
fuels resulted in an abrupt decrease in their
availability. It also led to the green-house
effect and many other environmental
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
60
problems. Nuclear power seems to be an
answer for global warming, but concerns
about terrorist attacks on Earth bound nuclear
power plants have intensified
environmentalist opposition to nuclear
power. Earth based solar panels receives only
a part of the solar energy. So it is desirable to
place the solar panel in the space itself,
where, the solar energy is collected and
converted in to electricity which is then
converted to a highly directed microwave
beam for transmission. This microwave
beam, which can be directed to any desired
location on Earth surface, can be collected
and then converted back to electricity. This
concept is more advantageous than
conventional methods. Also the microwave
energy, chosen for transmission, can pass
unimpeded through clouds and
precipitations.

1.2 SPS The Background
The concept of a large SPS that would
be placed in geostationary orbit was invented
by Peter Glaser in 1968.The SPS concept was
examined extensively during the late 1970s
by the U.S Department of Energy (DOE) and
the National Aeronautics and Space
Administration (NASA). The DOE-NASA
put forward the SPS Reference System
Concept in 1979. The central feature of this
concept was the creation of a large scale
power infrastructure in space, consisting of
about 60 SPS, delivering a total of about
300GW.But, as a result of the huge price tag,
lack of evolutionary concept and the
subsiding energy crisis in 1980-1981, all U.S
SPS efforts were terminated with a view tore-
asses the concept after about ten years.
During this time international interest in SPS
emerged which led to WPT experiments in
Japan.

1.3 SPS-A General Idea
Solar Power Satellites would be
located in the geo-synchronous orbit. The
difference between existing satellites and
SPS is that an SPS would generate more
power-much more power than it requires for
its own operation. The solar energy collected
by an SPS would be converted into
electricity, then into microwaves. The
microwaves would be beamed to the Earths
surface, where they would be received and
converted back into electricity by a large
array of devices known as rectifying antenna
or rectennas.
Each SPS would have been massive;
measuring 10.5 km long and 5.3km wide or
with an average area of 56 sq. km. The
surface of each satellite would have been
covered with 400 million solar cells. The
transmitting antenna on the satellite would
have been about 1 km in diameter and the
receiving antenna on the Earths surface
would have been about 10 km in diameter. In
order to obtain a sufficiently concentrated
beam; a great deal of power must be collected
and fed into a large transmitter array. The
power would be beamed to the Earth in the
form of microwave at a frequency of 2.45
GHz. Microwaves have other features such as
larger band width, smaller antenna size, sharp
radiated beams and they propagate along
straight lines.
Microwave frequency in the range of
2-3 GHz are considered optimal for the
transmission of power from SPS to the
ground rectennas site.The amount of power
available to the consumers from one SPS is
5GW. The peak intensity of microwave beam
would be 23 m W/cm. SPS has all the
advantage of ground solar, it generates power
during cloudy weather and at night. In other
words
SPS receiver operates just like a solar
array. It receives power from space and
converts it into electricity. This reduces the
size and complexity of satellite.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
61

Fig1: Basic Design Of SPS

2. Wireless Power Transmission
Transmission or distribution of 50 or
60 Hz electrical energy from the generation
point to the consumer end without any
physical wire has yet to mature as a familiar
and viable technology. The 50 Hz ac power
tapped from the grid lines is stepped down to
a suitable voltage level for rectification into
dc. This is supplied to an oscillator fed
magnetron.
The microwave power output of the
magnetron is channeled into an array of
parabolic reflector antennas for transmission
to the receiving end antennas. To compensate
for the large loss in free space propagation
and boost at the receiving end the signal
strength as well as the conversion Efficiency,
the antennas are connected in arrays. A
simple radio control feedback system
operating in FM band provides an
appropriate control signal to the magnetron
for adjusting its output level with fluctuation
in the consumers demand at the receiving
side. The overall efficiency of the WPT
system can be improved by-Increasing
directivity of the antenna array. Using dc to
ac inverters with higher conversion
efficiency .Using schottky-diode with higher
ratings.

2.1 Microwave Power Transmission In
SPS
The microwave transmission system
have had three aspects:
1. The conversion of direct power from the
photovoltaic cells, to microwave power on
the satellites on geosynchronous orbit above
the Earth.
2. The formation and control of microwave
beam aimed precisely at fixed locations on
the Earths surface.
3. The collection of the microwave energy
and its conversion into electrical energy at the
earths surface.

The key microwave components in a
WPT system are the transmitter, beam
control and the receiving antenna called
rectennas. At the transmitting antenna,
microwave power tubes such as magnetrons
and klystrons are used as RF power sources.
Rectenna is a component unique to WPT
systems. The following section describes
each of these components in detail.

2.2 Transmitter
The key requirement of a transmitter
is its ability to convert dc power to RF power
efficiently and radiate the power to a
controlled manner with low loss. The
transmitters efficiency drives the end-to-end
efficiency as well as thermal management
system. The main components of a
transmitter include dc-to-RF converter and
transmitting antenna. Power distribution at
the transmitting antenna = (1-r), where r is
the radius of antenna [7].There are mainly
three dc-to-RF power converters:
magnetrons, klystrons and solid state
amplifiers.


Fig. 2: Klystron amplifier
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
62
Klystron
Fig. 2 shows the klystron amplifier [15]. Here
a high velocity electron beam is formed,
focused and send down a glass tube to a
collector electrode which is at high positive
potential with respect to the cathode. As the
electron beam having constant velocity
approaches gap A, they are velocity
modulated by the RF voltage existing across
this gap. Thus as the beam progress further
down the drift tube, bunching of electrons
takes place. This variation in current enables
the klystron to have significant gain. Thus the
catcher cavity is excited into oscillations at its
resonant frequency and a large output is
obtained.
The tube body and solenoid operate at
300C and the collector operates at 500C.
The overall efficiency is 83%. The
microwave power density at the transmitting
array will be 1 kW/m for a typical 1 GW SPS
with a transmitting antenna aperture of 1 km
diameter. If we use 2.45 GHz for MPT, the
number of antenna elements per square meter
is on the order of 100.

2.3 Rectenna
Brown was the pioneer in developing the first
2.45GHz rectenna. Rectenna is the
microwave to dc converting device and is
mainly composed of a receiving antenna and
a rectifying circuit. Fig .4 shows the
schematic of rectenna circuit. It consists of a
receiving antenna, an input low pass filter, a
rectifying circuit and an output smoothing
filter. The input filter is needed to suppress
radiation of high harmonics that are
generated by the non-linear characteristics of
rectifying circuit. Because it is a highly non-
linear circuit, harmonic power levels must be
suppressed.


Fig. 3: Schematic of rectenna circuit.

For rectifying Schottky barrier diodes
utilizing silicon and gallium arsenide are
employed. Diode selection is dependent on
the input power levels. The breakdown
voltage limits the power handling capacity
and is directly related to series resistance and
junction capacitance through the intrinsic
properties of diode junction and material. For
efficient rectification the diode cut off
frequency should be approximately ten times
the operating frequency.

Diode cut off frequency is given by =1/
[2_RsCj], where is the cut off frequency, Rs
is the diode series resistance, Cj is the zero-
bias junction capacitance.

2.5 Recently Developed MPT Systems
The Kyoto University developed a
system called Space Power Radio
Transmission System (SPORTS).The
SPORTS is composed of solar panels, a
microwave transmitter subsystem, a near
field scanner, a microwave receiver. The
solar panels provide 8.4 kW dc power to the
microwave transmitter subsystem composed
of an active phased array. It is developed to
simulate the whole power conversion process
for the SPS, including solar cells,
transmitting antennas and rectenna system.
Another MPT system recently developed by
a team of Kyoto University, NASDA and
industrial companies of Japan, is an
integrated unit called the Solar Power Radio
Integrated Transmitter (SPRITZ), developed
in 2000. This unit is composed of a solar cell
panel, microwave generators, transmitting
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
63
array antennas and a receiving array in one
package.

3. Construction of SPS from Non
Terrestrial Materials: Feasibility and
Economics
SPS, as mentioned before is massive
and because of their size they should have
been constructed in space. The aluminum and
silicon can be refined to produce solar arrays.
Among them are the shallow gravity wells of
the Moon and asteroids; the presence of an
abundance of glass, metals and oxygen in the
Apollo lunar samples; the low cost transport
of those materials to a higher earth orbit by
means of a solar powered electric motor; the
availability of continuous solar energy for
transport, processing and living. One major
new development for transportation is
required: the mass driver. The mass driver is
a long and narrow machine which converts
electrical energy into kinetic energy by
accelerating 0.001 to 10 kg slugs to higher
velocities. The mass driver conversion
efficiency from electrical to kinetic energy is
close to 100 percent.

3.1 Microwaves-Environmental Issues
The price of implementing a SPS
includes the acceptance of microwave beams
as the link of that energy between space and
earth. Because of their large size, SPS would
appear as a very bright star in the relatively
dark night sky. SPS possess many
environmental questions such as microwave
exposure, optical pollution that could hinder
astronomers, the health and safety of space
workers in a heavy-radiation (ionizing)
environment, the potential disturbance of the
ionosphere etc. The atmospheric studies
indicate that these problems are not
significant, at least for the chosen microwave
frequency.
On the earth, each rectenna for a full-
power SPS would be about 10 km in
diameter. This significant area possesses
classical environmental issues. These could
be overcome by sitting rectenna in
environmentally insensitive locations, such
as in the desert, over water etc. However, the
issues related to microwaves continue to be
the most pressing environmental issues. On
comparing with the use of radar, microwave
ovens, police radars, cellular phones and
wireless base stations, laser pointers etc.
public exposures from SPS would be similar
or even less. Based on well-developed
antenna theory, the environmental levels of
microwave power beam drop down to
0.1W/cm.
Serious discussions and education are
required before most of mankind accepts this
technology with global dimensions.
Microwaves, however is not a pollutant but,
more aptly, a man made extension of the
naturally generated electromagnetic
spectrum that provides heat and light for our
substance.

4. Advantages and Disadvantages
The idea collecting solar energy in
space and returning it to earth using
microwave beam has many attractions. The
full solar irradiation would be available at all
times expect when the sun is eclipsed by the
earth. Thus about five times energy could be
collected, compared with the best terrestrial
sites. The power could be directed to any
point on the earths surface. The zero gravity
and high vacuum condition in space would
allow much lighter, low maintenance
structures and collectors. The power density
would be uninterrupted by darkness, clouds,
or precipitation, which are the problems
encountered with earth based solar arrays.
The realization of the SPS concept holds
great promises for solving energy crisis.
The concept of generating electricity
from solar energy in the space itself has its
inherent disadvantages also.


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
64
Some of the major disadvantages are:
The main drawback of solar energy
transfer from orbit is the storage of electricity
during off peak demand hours. The frequency
of beamed radiation is planned to be at 2.45
GHz and this frequency is used by
communication satellites also. The entire
structure is massive. High cost and require
much time for construction. Radiation
hazards associated with the system. Risks
involved with malfunction. High power
microwave source and high gain antenna can
be used to deliver an intense burst of energy
to a target and thus used as a weapon.

5. Conclusion and Future Scope
The SPS will be a central attraction of
space and energy technology in coming
decades. However, large scale retro directive
power transmission has not yet been proven
and needs further development. Another
important area of technological development
will be the reduction of the size and weight of
individual elements in the space section of
SPS. Large-scale transportation and robotics
for the construction of large-scale structures
in space include the other major fields of
technologies requiring further developments.
The electromagnetic energy is a tool to
improve the quality of life for mankind. It is
not a pollutant but more aptly, a man made
extension of the naturally generated
electromagnetic spectrum that provides heat
and light for our sustenance. From this view
point, the SPS is merely a down frequency
converter from the visible spectrum to
microwaves.

References-
[1] Hiroshi Matsumoto, Research on solar
power satellites and microwave power
transmission in Japan, IEEE microwave
magazine, pp.36-45, Dec 2002.
[2] James O. Mcspadden & John C.
Mankins,Space solar power programs and
Microwave wireless power transmission
technology, IEEE microwave magazine,
pp.46-57, Dec 2002.
[3] J.C. Mankins,A fresh look at space solar
power: new architectures, concepts and
technologies in 38th Astronautical
Federation.
[4] Seth Potter, Solar power satellites: an
idea whose time has come [online] Available
on www.freemars.org/history/sps.html, last
updated on Dec.1998
[5] Consumer Energy Information: EREC
Reference Briefs [online] Available on
www.eere.gov/consumerinfo/rebriefs/123.ht
ml,last updated on Apr.03.
[6] Mc GrawHill Encyclopedia of Science
and Technology, vol.16, pp.41.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
65
Autonomous pick and place rover for long distance
surveillance using ultrasonic sensors
V.Sampath kumar
1
, Amarendra J adda
2

1
PG Student, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India.
2
Assoc. Professor, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India

1
ver appa1989@gmai l . com
2
amar endr a. j adda@gmai l . com
Abstract:
The technology advancements takes place in
the field of robotics. This paper presents the
autonomous rover where it can be used for long
distance surveillance. It is based on the ultrasonic
sensors,Temperature sensors and humidity sensors
meant for to detect the obstacles as well as to estimate
the parameters such as temperature, humidity
respectively. The wireless communication protocols
used in between PC and Robo are zigbee and RF
protocols.This robo is capable of pick and place of
objects using the special type of grippers. In this
paper, the usage of LPC2148 is another added
advantage which makes the robotic system highly
Reliable.
Index Index Terms---- Autonomous Robot, Sensors,
RF protocol, Zigbee protocol, DC motor, Camera,
Pick and place robot
I.Introduction:
Robots play a vital role because they will
work under any conditions without the tiredness.
They are used in Industries, IC manufacturing,
Pharmaceutical etc.,. for different purposes. They
may be static or dynamic robots operates by the
commands given manually or simply by automated
instructions. This robos can perform multiple tasks
by attaching devices like sensors, cameras inorder
to calculate the various parameters based on the
application required. These robos can be controlled
by either the microcontrollers or processors which
are dependent on the protocols used such as zigbee
and RF protocals. The dynamic robos are taking
directions under the guidelines of the operator.
II. Related Work:
This paper focuses on the dynamic
approach of calculating various parameters based
on the sensors used along with the pick and place
mechanism. In this approach the camera and sensor
modules are used which are communicating with
the PC based on the RF, Zigbee protocols
respectively. LPC 2148 is used for controlling of
the robo by giving the instructions. The various
types of sensors are used such as temperature
sensor, humidity sensor and ultrasonic sensor.
A.Pick and Place module:
The name itself indicates that the objects
are to be picked from one place and placed on
another place. The prime objective is to pick the
objects based on the environment which can
handles any shape of the object. The picking arm
can rotate upto 360 degrees by making the base of
the robot fixed.
B.Sensor module:
I. Temperature Sensor:
In General temperature sensors are used to
sense the temperature and thereby transducers are
converting these sensed temperature values to the
electrical values. Here the temperature sensor used
is LM35 which is analog in nature and its
functional range lies in between 0 degree Celsius to
100 degree Celsius. It has a sensitivity of 10mV per
degree Celsius.

Fig 1. Pin diagramof LM35
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
66

II. Humidity Sensor:
t is used to measure the moisture content
in the environment, designed to give the enhanced
parameters of sensitivity and response time.
Humidity sensor is highly stable and reliable. These
sensors are constructed by a planar capacitor and
another polymer layer to protect against the
environmental effects that may be hazards.
C. Ultrasonic Sensor:
These sensors are used to find the
obstacles if any that found in front of robot.
Ultrasonic sensors are based on the principal of the
Radar or Sonarwhere it gets the echo signal. It will
act as a transceiver in which it generates the high
frequency signal and analyse the reflected back
echo signal. It measures the distances up to 2.5
meters at an accuracy of 1 centi- meter.


Fig2. Block Diagramof Ultrasonic Sensor
III. Communication Protocols:
The different type of protocols are used
here in order to communicate between the PC and
the Robot. The protocols that are used are Zigbee
protocol and RF protocol respectively.
A. Zigbee protocol:
Its a wireless protocol based on the IEEE
802.15 standard that suits for low power devices
communication. Generally it is intended for short
distance communication but we may extend to long
distance using intermediate devices.

Fig 3. Zigbee Module
It provides low latency and duty cycle that
supports for multiple network topologies such as
static, mesh and star topologies. It will support upto
65,000 nodes on a network. It avoids the collision
of packets and provides the acknowledgment
status.
IV. Hardware Module:
It consists of the components such as DC
motor, LPC 2148, and an LC293D. DC motor is the
one which converts electrical energy into
mechanical energy used for various industrial
applications. The driver circuit is used for driving
these motors where in which here LC293D is used
as a driver circuit. The brief content of the these
modules as follows:
A. DC motors:
There are many types of DC motors are
available in which most of the cases two types of
motors most commonly used such as brushed and
brushless types. Here in this paper mainly focusing
on the brushed DC motor which will operate at
12V DC and 0.6A. The speed of the DC motor is
controlled by either changing the voltage applied to
armature or by varying the field current. Most of
the DC motors are driven by the Driver ICs at
present.
B. LC293D:
This IC acts as driver circuit in which it is
available in the 16-bit DIP. It will control the two
small motors simultaneously in either forward or
reverse direction, in which 4 pins is sufficient in
order to performthis tasks.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
67

Fig4. DC Motor
There exists certain features and
drawbacks of this IC that is used such as output
current capability is limited to 600mA per channel
in the sense this IC cannot drive the bigger motors.
Most probably this would be fit for the small
motors which suits for the Robot armmovements.
If the output current exceed the 1.2 A repeatedly
then the IC will be destroyed.
The supply voltage will maintain up to
36V maximum. If the output current exceed beyond
the 1.2A then the IC gets destroyed so here in order
to protect the IC fromover temperature an internal
sensor will be used for sensing the temperature. If
the temperature will exceeds beyond the set point
then automatically this sensor will send an alert and
stops the motors. The additional feature of this IC
is using clamp diode which protects the IC fromthe
voltage spikes.
In this IC if the voltage exceeds 1.5v then
only the pin will gets enabled that makes suitable
for high frequency applications like switching
applications. Other than driving the motors it also
be used to drive the solenoids, stepper motors etc.,
This is mainly based on the concept of H-bridge in
which the voltage will be flow in either of the
directions. By changing the voltage it will make the
Robot to rotate in clockwise or anticlockwise
direction. In this LC293D two H-bridge circuits are
present inside the IC which will be able to rotate
two dc motors independently.
Here two pins are used namely pin 1 and
pin9 respectively. For driving the motor with left
H-bridge we enable the pin 1 whereas driving right
H-bridge we make the pin 9 to be enable. If both
the pins are disable then the motor used will be
suspended in which it acts like switch.


Fig 5. IC L293D
C. Camera Module:
In this paper camera used is wireless
camera used for both the transmitter and receiver
section. Its range lies in between 100 and 150
metres based on the principal of the RF protocol.
The camera used here operates at the range of 1.2
GHZ. The transmitter and receiver section uses the
power supply voltage of 9v battery.

Fig 6. Wireless Camera
D. LPC 2148 Processor:
ARM is a family of instruction set
architectures based on RISC architecture developed
by ARM Holdings. This ARM holdings had
reported that shipments was around 6100 million
ARM-based processors to manufacturers of chips
based on ARM architectures where the maximum
percentile usage of smart phones and the remaining
of 35 per cent of digital TVs and set-top boxes. As
of 2014, ARM is the usage of 32-bit instructions is
the maximumfor major number of applications.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
68

Fig 7. H-Bridge
V. Experimental Results:
The robot used here is used to
pick the objects and is rotated in 360 degrees
rotation. The In this project is to implement a 32-bit
low-power ARM processor into the development of
a Mobile Robot using ARM Processor for Line
Following Application mobile robot equipped with
IR sensor capabilities. The scope of the project is
LPC2148 ArmProcessor DC brushless motor 100
RPM Motor is used. At the end of this project, the
ARM processor has successfully for motor control
and respond to the digital input from the interface
of IR sensor to the Microcontroller.
snapshots of the robot used in various views such
as front view, side view and top view respectively.

Fig 8. Top View of Robot

Fig 9. Side View of Robot
The Robot thus shows in various
directions of different views where we used the
sensors and camera based modules for surveillance.

Fig 10. Front View of Robot
VI. Conclusion:
The Robot in this paper used various
sensors which senses the temperature and humidity
respectively. The camera is used to estimate the
position and orientation of robot. This will be used
to work at hostile environments where not possible
for the humans to do that. In this paper this Robot
will handles over the distance of 100 metres
maximum. We may expect that this coverage of
surveillance can be increased to a long distances
where a major number of applications comes
without the human intervention.



Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
69
References:
[1] J acquier E, Polson NG, Rossi PE (1993) Priors and models
for multivariate stochastic volatility.Unpublished manuscript,
Graduate School of Business,University of Chicago.
[2] L. Righetti, A. Ijspeert, Design methodologies for central
pattern generators: an application to crawling humanoids, in:
G.S. Sukhatme, S. Schaal, W.Burgard, D. Fox (Eds.),
Proceedings of Robotics: Scienceand Systems,MIT Press, 2006.
[3] J . Pisokas, U. Nehmzow, Experiments in subsymbolic action
planning with mobile robots, in: Adaptive Agents and Multi-
Agent Systems II, in: Lecture Notes in Artificial Intelligence,
Springer, 2005, pp. 8087.
[4] Franklin Hanshar and Beatrice Ombuki-Berman. Dynamic
vehicle routing using genetic algorithms.Applied Intelligence,
27(1):5291, 2007.
[5] Thomas L. Dean and Michael P.Wellman. Planning and
control. Morgan Kaufmann Publishers Inc., 1991.
[6] G. Antonelli and S. Chiaverini, Linear estimation of the
physical odo- metric parameters for differential-drive mobile
robots, Auton. Robots, vol.23, no. 1, pp. 5968, J ul. 2007.
[7] A. Martinelli, N. Tomatis, and R. Siegwart, Simultaneous
localization and odometry self calibration for mobile robot,
Auton. Robots, vol. 22, no. 1, pp.
7585, J an. 2007.
Amarendra Jadda is working as
Assiciate Professor in ECE Dept,
ASCET, GUDUR. He has been
guiding UG & PG students since
three years in this institution. He
pursued his M.Tech FromJ NTUH,
Kukatpally Hyderabad. He
presented papers in four
international journals & four international conferences. His
intresting fields are Communications &Signal Processing,
Embedded Systems.

Sampath kumar veerappa did his
B.Tech in Electronics and
communication engineering from
Prathyusha Engineering College,
Chennai. He is currently pursuing
his M.Tech degree in Embedded
systems from audisankara college
of engineering, Gudur.







Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
70



Intelligent Bus Alert System For Blind Passengers
N.Surendra Yadav
1
, Tulasi Sanath Kumar
2

1
PG Student, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India.
2
Asst. Professor, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India

1
cupi d. sur i @gmai l . com
2
t ul asi sanat h@gmai l . com


Abstract:Talking signs, guide cane, echolocations
are all useful innavigating the visually challenged
people to reach theirdestination, but the main
objective is not reached that it fails tojoin them with
traffic. In this project we propose a bus systemusing
wireless sensor networks (WSNs).The blind people in
thebus station is provided with a Zig-Bee unit which
is recognized bythe Zig-Bee in the bus and the
indication is made in the bus thatthe blind people is
present in the station. So the bus stops at theparticular
station. The desired bus that the blind want to take
isnotified to him with the help of speech recognition
systemHM2007. The blind gives the input about the
place he has toreach using microphones and the voice
recognition systemrecognizes it .The input is then
analyzed by the microcontrollerwhich generates the
bus numbers corresponding to the locationprovided
by the blind. These bus numbers are converted
intoaudio output using the voice synthesizer APR
9600. TheZig-Beetransceiver in the bus sends the bus
number to the transceiverwith the blind and the bus
number is announced to the blindthrough the
headphones. The blind takes the right bus parked
infront of him and when the destination is reached it
is announcedby means of the GPS-634R which is
connected with the controllerand voice synthesizer
which produces the audio output. Thisproject is also
aimed at helping the elder people for
independentnavigation.

Index Terms:Wireless sensor networks, Speech
Recognition System,Voice Synthesizer, GPS, Zigbee.

I. INTRODUCTION

Out the 6.7 billion people that populate the world,161
million are visually impaired. Each visually
impairedindividual faces different challenges based
on their specificlevel of vision. With the rise of
various support-basedorganizations, more visually
impaired people have been giventhe opportunity to
education and many other means. But stillthe issues
of navigation for the blind are very complex
andtroublesome especially when they walked down
in street andalso navigate to distant places by public
transport system. Fora visually impaired person,
doing things such as reading trafficsignals and street
signs can be extremely challenging, if not itis
impossible to do.

In order to overcome these challenges, a visually
impairedperson might use walking cane, guide dog,
and sighted guide. These alternatives also called as
assistive devices can behelpful to the blind but not so
effective. The sighted guide canbe immensely
effective, as well provide social comfort, but
itrestricts the independence of the blind individual.
Guide dogsand walking canes allow for a more
independent means oftravelling, but they are limited
in unfamiliar environments.RFID is feasible and cost
effective but it is more suitable forindoor
communication only. Also it provides only one
waycommunication and a very short range of
identification. Asystem with an augmented walking
cane, a pair of augmentedglasses and identifiable
items tagged with semacode/datamatrix tags is used
for outdoor navigation of blind people. If aman has to
take the bus, he walks along the pavement and
hiswalking cane recognizes a tag. But the image
quality of theweb camera is fairly poor. Tag
recognition in darkness or inbad lightning conditions
might be a problem. Another issue isthat camera
needs a visual, so if a tag is hidden behind aperson or
another object, then the camera cannot detect it.Tags
on all environments will properly contaminate
theenvironment and meet resistance from many
citizens.

To overcome the drawbacks of currently
availableassistive devices, we propose a Wireless
sensor network systemwith Zig-Bee for blind
identification by the bus and embeddedsystem for
providing the bus number and [mally GPS
fordestination indication. Wireless sensor network
(WSNs)consists of sensors that continuously
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
71


monitors theenvironmental conditions and send their
data to the mainnetwork [51. Zig-Bee is an embedded
device for use in a WSNwhich is tiny in size. These
nodes have processing andcomputational capability
and generally consist of an RFtransceiver, memory,
on board sensors/actuators and a powersource [71.
Zig-Bee have CC2420 which is a true single-chip
2.4GHz IEEE 802.15.4 compliant RF (Radio
Frequency)transceiver designed for low-power and
low-voltage wirelessapplications so we can send or
receive useful informationthrough using this chip.
The number of the bus parked in frontof the blind is
send to the Zig-Bee in the blind system.
Anotherfunction of Zig-Bee is identification of blind
in the bus station.If both the numbers match the
buzzer in the bus unit alarmsand indicates the driver
that there is blind in the bus station.

The software part is Embedded C with MPLAB
IDE(Operating System) for programming the
controller. MPLABIDE runs as a 32-bit application
on MS Windows, is easy touse and includes a host of
free software components for fastapplication
development and super-charged debugging.MPLAB
IDE also serves as a single, unified graphical
userinterface for additional Microchip and third party
software andhardware development tools.

II. DESIGN PRINCIPLES OF ELECTRONIC
TRAVELLING AIDS

The most important travelling aid for a visually
impairedperson is still the white cane. It is after all
an excellentexample of a good travelling aid:
multifunctional, cheap andreliable. It also tells others
that the person is visuallyimpaired. Another
irreplaceable travelling aid is a guide dog.Among
other things the dog is also a friend and a companion.

In studies about visually impaired person navigation
it hasbeen noted that even a small amount of extra
informationabout the environment makes a
remarkable increase inperformance. Also it seems
that a good travelling aidshould produce only small
amounts of meaningfulinformation and the ETA
should not block hearing or othersenses so that the
visually impaired can still use theirtraditional
methods to acquire information about
theenvironment. If the user needs to concentrate
heavily on usingthe ETA, he or she has no capacity
left for normalenvironment perception.

Therefore, instead of trying to develop ETAs to
replaceprimary travelling aids, one should develop
complementarysystems.Navigation systems have
usually worked well in smallscaleimplementations,
but a large-scale implementation maybe extremely
expensive (especially with beacon basednavigation
systems). The amount of visually impaired personsof
the population is small (~ 1,6 %) and therefore
largeinvestments to special infrastructure are not
sensible.As an example, there have often been
suggestions aboutequipping buses with radio
transmitters to help the visuallyimpaired to know
when the bus is coming.

The visuallyimpaired would in turn carry a radio
receiver.In Prague there is a pilot system in
operation. However, forexample in Finland, where
we have about 80 000 visuallyimpaired and 10 000
buses, a similar system would cost atleast 10 M just
for the bus transmitters.

Other methods need to be found to ensure that the
visuallyimpaired persons have equal possibilities to
access sameservices than all the other citizens.

2.1. Visually Impaired Persons and Public
Transport

Generic travelling difficulties for the visually
impairedpersons are localization and environment
perception, selecting and maintaining the correct
heading, detecting obstaclesabove waist and
detecting unexpected roadworks.
If we examine problems a visually impaired person
meetswhen using public transport, we recognize the
following list(the list depends slightly of the
transportation):

planning a trip
finding a stop or station
finding an entrance to the station
navigating inside the station
finding the right platform and waiting place
knowing when the right vehicle arrives
finding a vehicle entrance
payment
finding a seat
receiving passenger information during the trip
depart on the right stop
finding the destination.

Most of these tasks are trivial for the sighted persons,
butvery difficult for the visually impaired. For
example therehave been cases when a blind has spent
several hours at a busstop, because he couldn't
recognize the arrival of the rightvehicle.

For true door-to-door navigation for a visually
impairedthere are requirements for continuous
positioning, continuous(Internet) access to real time
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
72


public transport information andavailability of
accurate map data together with
roadworkinformation.

Electronic maps are designed for car navigation and
notsuitable for pedestrian route planning purposes.
Informationabout pavements and pedestrian crossings
is collectedseparately and not included in typical map
data. Door-to-doorguidance requires map data
pointing entrances to houses andcontinuous guidance
would require indoor maps and indoorpositioning.
These are generally unavailable today.

Nevertheless, in our studies we did not find many
specificinformation needs for the visually impaired
group alone. Theinformation needed and sought for is
in most cases useful forall passengers or already
accessible to some user groups. Onlythe means for a
visually impaired to access the informationwould be
different. Therefore it is very important to
offeradditional interface, when new passenger
informationservices are designed.

III. NOPPA ARCHITECTURE AND
PROTOTYPE

Our approach is to improve public transport
accessibilityby creating access to passenger
information with a personalmobile device rather than
building physical infrastructure.NOPPA architecture
(fig.1) is based on public and/orcommercial
information services and databases available via the
Internet, a client-server approach with near
continuousTCP/IP connection (GPRS for practical
reasons) andprogrammable mobile devices with
capabilities for speechuser interface (mainly speech
synthesis) and satellitepositioning.



Fig.1. Architecture of the NOPPA system

The terminal devices in NOPPA user tests have by
far beenvarious PDAs with Microsoft's Pocket PC /
Pocket PC PhoneEdition operating system and there
is also an early version formobile phones with
Symbian Series 60 operating system.
Design goals for NOPPA were:

Easy and fast to use (preferably faster than any
traditionalmethod)

Affordable to the user

Access to public transportation and
passengerinformation systems

Applicable both indoors and outdoors

Integration of products and services for
personalnavigation

Modular, easy to update, easy to add functions

Speech user interface

Easy to customize for various user groups and
purposes.

The Information Server is an interpreter between the
userand Internet information systems. It collects
filters andintegrates information from different
sources and delivers theresults to the user.

The server handles speech recognition (e.g. from
13200street and destination names) and functions
requiring eitherheavy calculations or large data
transfer from the Internet.The data transfer between
the server and the client is kept inminimum. The
client terminal (fig.2) holds speechsynthesis, user
interface, positioning and most of routeguidance. The
user interface is menu-based and selections aredone
with hardware buttons and speech input.

As the mobile devices gain more memory and
fasterprocessors some of the speech recognition work
can be donein the user terminal which will further
reduce the need fordata transfer between the client
and the server. It will alsoenable menu selections
with speech user interface when thereis no server
connection. Nevertheless, the speech
recognitionrequires a very large vocabulary (street
names) which also hasto be updated from time to
time, so it may be unpractical tocompletely do the
processing in the terminal.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
73


NOPPA terminal software with speech synthesis
needs tobe installed on the device, completely
replacing theunderplaying operating system's user
interface. If the operatingsystem supports a screen
reader for example, more functions(such as phone
calls, SMS and MMS) can be left to originalsoftware.

The first prototype system has the
followingcharacteristics:

Speech recognition and synthesis
6 simultaneous users per single server computer (a 2
GHz
PC) for speech processing time limits

Access to three route planners (commuter and
intercitytraffic both bus and train, also a possibility to
calculatecar navigation type of routes)
Guidance and route following during a trip

Personal in-vehicle stops announcements

Roadwork information (connection to a city's
database)

Access to some bus, tram and train real time
informationsystems (only early development)

Flight departure information at the largest airport in
Finland, real time

Several news services, local weather

Watch

Memo

GSM phone and SMS services (basic
implementation)

Bluetooth and GPRS connectivity (also WLAN
possible)

GPS and GSM positioning, optional pedometer
andcompass unit

Indoor navigation features based on Bluetooth,
WLANpositioning or compass/pedometer

Own recorded walking routes, basic GPS functions

Search of current address

POI (Point of Interest) and AOI (Area of
Inters)databases


Fig.2. NOPPA Pocket PC terminal and a
Bluetooth GPS receiver
The prototype is now at user evaluation phase.
Usability,reliability and recovery after an error are
known to beimportant issues. For example a
continuous GPRS serverconnection is not possible
when moving in a train, elevator orbasement.

The system must be designed to handle connection
failures so that they don't break guidance or prevent
usingother functions (phone call, SMS, memo
etc).Commercial sensitive GPS receivers are able to
operateinside a bus and a tram, but still greatly
benefit from antennaplacement near window. Also
GPS receivers' slow time tofirst fix (TTFF, typically
30-60 seconds) can be a problemwhen turning a GPS
first on or leaving a building after beinga long time
inside with no update to receiver satellite data.

The speech output in guidance and in describing a
routemust be carefully planned to avoid
misunderstandings and tohelp create a mental image
of the route. The program shouldnot try to give more
accurate guidance than it safely can. Forexample
when standing near a bus stop, if the program
wouldadvice that "the bus stop is 10 meters
forwards", the usermight very well end up standing
on a driveway. Combinedinaccuracy of GPS
positioning and map data is very oftenover 10 meters
and the program should not really try to guidethat
short a distance (at least not require the user to
move),even though there would seem to be a clear
differencebetween GPS and target coordinates.

The difficulty is to tell the user without
misunderstandings,that the calculated target is maybe
20 meters forwards, but theuser has to find the exact
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
74


location himself and it may not beeven safe to move
the full 20 meters. Often there is someinformation
even in the short distances, so the user mightwant to
hear the target distance and direction after all,
insteadof just hearing "the target is near".In practice,
one must take into account that map data canhave
outdated information or inaccuracies, positioning can
beunavailable or inaccurate, or wireless data
transmission is notalways available. Therefore a lot
of responsibility is left forthe user and guidance is
complementary.

IV. RESULTS AND DISCUSSION

When the person reaches the bus station, he canfind
the buses that pass through a particular location with
thehelp of voice recognition system and voice
synthesizer.

When the bus approaches the bus station, there is
anindication in the bus by the beep sound of a buzzer
that thereis a blind person available in the bus
station.This is achieved with the help of Zig-Bee unit
both in the busunit and blind unit. Finally when the
bus reaches the stationthe bus number is announced
to the blind through headphones.

There are currently available systems for the
outdoornavigation but they will not assist in
travelling to unfamiliarareas. Some systems use PDA
which is not so economic andcannot be afforded by
all. In most of the systems RFID tagsare used which
are required in 1000s of numbers for trackingof route.
Also it provides only one way communication.
Thesystem we use is a mobile unit, weightless and
economicallyfeasible.

V. CONCLUSION

Primarily, the blind person in the bus station is
identifiedwith RF communication. The blind informs
the location heneeds through the microphone which
is given to the voicerecognition system which
produces the output of bus numbersin the voice
synthesizer unit which is heard in headset. Thenthis
location is transmitted to the transceiver in the bus. If
thenames in the transceiver in the bus matches with
that of thename send by the blind, then there is an
alarm in the bus unitalerting the presence of blind and
a voice to the user's headsetthat the particular bus has
arrived. With the help of GPStracker connected with
audio output the destination chosen bythe blind is
intimated when the bus reaches the correctlocation.
PDA's can be used for GPS tracking but it is not
costeffective.

VI. REFERENCES

[1] G.Lavanya ME, Preethy. W, Shameem.A,
Sushmitha.R, Passenger BUS Alert System for
EasyNavigation of Blind, 2013 International
Conference on Circuits, Power and Computing
Technologies.

[2] Baudoin,G., Sayah,J, Venard, O. and EI Hassan,
B.(2005), 'Simulation using OMNeT++ of the
RAMPE systemanInteractive Additive Machine
helping blinds in PublicTransports', EUROCON,
Belgrade,pp.1-5.

[3] Bolivar Torres, Qing Pang, (2010), 'Integration of
anRFID reader to a Wireless sensor network and its
use to identify an individual carrying RFID tag',
InternationalJournal of ad hoc. sensor& ubiquitous
computing ,voU,no.4,pp.1-15.

[4] Brendan D Perry, Sean Morris and Stephanie
Carcieri,(2009), 'RFID Technology to Aid in
Navigation andOrganization for the Blind and
Partially Sighted', pp. 1-52.

[5] HerveGuyennet, KamalBeydoun and Violeta
Felea,(2011), 'Wireless sensor network system
helping navigation ofthe visually impaired', IEEE
international conference onInformation and
Communication Technologies: from Theoryto
Applications, version 1, pp. 1-5.

[6] Hyn Kwan Lee, Ki Hwan Eom , Min Chul Kim
and TrungPham Quoc, (2010), 'Wireless Sensor
Network Apply for theBlind U-bus System',
International Journal of u- and eservice,Science and
Technology,VoI.3,No.3,pp.l3-24.

[7] Jack Loomis,M. and Roberta Klatzky,L.
(2008),'Navigation System for Blind', Massachusetts
Institute ofTechnology, Vol 7, No.2, pp.193-203.

[8] Jain.P.C ,Vijaygopalan.K.P. (2010), 'RFID and
WirelessSensor Networks', Proceedings of ASCNT,
CDAC, Noida,India, pp. 1 - 11.

[9] Loc Ho ,Melody Moh, Teng-Sheng Moh and
ZacharyWalker (2007) , 'A Prototype on RFID and
Sensor Networksfor Elder Health Care', Taylor &
Francis Group, LLC,pp.314-317.

[10] Oyarzun, C.A and Sanchez, J.H. (2008), 'Mobile
audioassistance in bus transportation for the blind',
Department of Computer Science, University of
Chile, pp.279-286.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
75


[11] Ravi Mishra and SornnathKoley, (2012),'Voice
OperatedOutdoor Navigation System For Visually
Impaired Persons',International Journal of
Engineering and Technology,Vol 3,Issue 2,pp.l53-
157.

[12] RiazAhamed, S.S (2009), 'The Role of Zig-
BeeTechnology In Future Data Communication
System', Journalof Theoretical and Applied
Information Technology,India, pp.129-134.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
76
7

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
77
Traffic Sign Recognition for Advanced Driver
Assistance System Using PCA
Kande Prasyam
1
, S. Himabindu
2

1
PG Student, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India.
2
Asst. Professor, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India

1
499pr asyam@gmai l . com
2
bi ndu437@gmai l . com
Abstract:
Traffic sign Recognition plays a vital role
for the drivers in order to avoid the hurdles like speed
breakers, narrow bridge or even accident zone etc.
This paper presents the effective recognition of traffic
signs using Principal component analysis. This could
be done by placing a camera which captures the road
sign images and it will be displayed as a video file in
the GUI. This video file is converted into frames
called array indexing. Here this technique uses
different methods of image processing such as image
segmentation, sign recognition and sign classification.
The Eigen values of these images calculated and given
to LPC 2148 processor where it will be interfaced
with the audio amplifier and shows the sign direction
in LCD.
Index Terms---- Road sign, Principal Component
analysis, Graphical user interface, Eigen values,
Eigen vectors, LPC 2148.
I. Introduction
Image itself a matrix, it will be arranged in the
forms of Rows and columns. For comparing of the
similar images Independent Component analysis is
sufficient but for comparing different images with
different Eigen values principal component
analysis came into picture. Here the Eigen values
are calculated and are compared with the database
values so that how close the value matches that
would be treated as image for sign recognition.
The primary objective of this paper is to
extract the details of type of sign that exists in sign
board and intimates to the driver through the voice
alert. The other advantage is the voice alert is in the
language of the driver.
II. RELATED WORKS
This paper focuses on the image processing
modules and hardware modules where for
interfacing between themRS232 cable is used.
The software module consists of three modules
such as image segmentation, sign recognition
and sign classification and hardware
components such as LPC 2148, audio amplifier
and LCD Display.
A. Image Processing module:
This module by name itself indicates that
processing of image using different techniques
such as RGB Colour segmentation,
Recognition of signs and classification of
them.
Image segmentation:
Initially sign board images are captured
using camera and can be segmented in order to
determine the exact boundaries of that image
used for effective analysis. This colour
segmentation is used to be converted to 2D
image and then calculate the Eigen values
easily
Sign Recognition:
After segmentation of image the sign is
recognised from the sign board used for
advanced driver assistance. These recognized
signs given as an input to the classification
module.
Sign Classification:
The classification stage includes
compares the signs that are recognised and
with the database images.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
78
Block Diagram:





Fig 1. Block Diagram of image processing
module
The sign board images gives as an input to
the personal computer where it processes the image
and compares the image with the database images
and finally selects the image that is closer to
database images. The extracted image after
comparing with the database images will send to
the RS232 cable.
This RS232 cable sends the index number
to LPC 2148 that is generated from Personal
computer using Keil software. ARM processor
interfaces with the audio amplifier as well as
displaying the type of sign in LCD.
Eigen Values and Eigen Vectors:
Eigen vectors are non-zero vectors of a
linear operator and result in a scalar multiple of
themwhen operated on by the operator. The scalar
then called the Eigen value where In association
with the Eigen vector. The property of matrix is in
which when a matrix acts on it only the vector
magnitude is changed but not the direction.
Consider the Eigen vector of X where A is a vector
function
AX=X (1)
By using equation 1 we get equation 2
(A-I)X=0 (2)
where I is the n x n Identity matrix.
The above mentioned is a homogenous systemof
equations and we know that a non trivial solution
exists only when
det (A-I) = 0 (3)
Where det() indicates the determinant and
this above equation is also known as characteristic
equation of A. If A is nxn , then there are n
solutions or n roots of the characteristic
polynomial. This characteristic polynomial is of
order n. Hence there are n Eigen values of A
satisfying the equation.
AXi=Xi (4)
Where i=1,2,3,.n
If all the Eigen values are distinct, there are n
associated linealy independent eigenvectors, whose
directions are unique, which meant for an n
dimensional Euclidean space.
Eigen Sign Approach:
The Eigen values of the input signs
captured is compared with the database signs. If
any matching occurs with the database image then
accordingly based on the index number it will
shows the turn right, turn left, turn curve etc., based
on the assignment of the index value to the
corresponding sign.
When the Sign image to be recognized
may be known or unknown that is captured by a
camera and we get the weights associated with the
Eigen signs, that linearly approximate the sign or
can be used to reconstruct the sign. Now these
weights are compared with the weights of the
known sign images that are available in database
so that it can be recognized as a known
sign.The Euclidean distance between the image
projection and known projections is calculated; the
sign image is then classified as one of the signs
with minimum Euclidean distance.
Mathematically calculations:
Let a sign image I(x,y) be a two
dimensional N by N array of (8-bit) intensity
values. An image may be considered as a vector
Sign
board
images
PC
RS232
LPC
2148
Audio amplifier
LCD
Display
Speaker Indications
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
79
of dimension N
2
, for a typical image it will be a
256 by 256 size and would be the vector of
dimension 65,536 or equivalently we will say that
65,536-dimensional space. An ensemble of images,
then, maps to a collection of points in this huge
space. Principal component analysis would find the
vectors that best account for the distribution of the
sign images within this entire space.
Let us consider a set of sign images be
T
1
,T
2
,T
3
,.T
M.
This sign images data set has to be
mean adjusted before calculating the covariance
matrix or Eigen vectors. The average sign is
calculated as = (1/M)
1
M
T
i,
Each image in the
data set differs from the average sign by the vector
= T
i
.This is actually mean adjusted data. The
covariance matrix is
C = (1/M)
1
M

i

i
T
(5)
= AA
T

where A = [
1,

2,
.
M
].
The covariance matrix considered here is a
N
2
by N
2
matrix and would generate N
2
eigenvectors and eigenvalues. It is impractical to
calculate with image sizes like 256 by 256, or even
lower than that.
An effective solution is needed to
calculate the Eigen vectors. Set of images that are
considered is less than the no of pixels in an image
(i.e M <N
2)
, then we can solve an M by M matrix
instead of solving a N
2
by N
2
matrix. Consider the
covariance matrix as A
T
A instead of AA
T

The eigenvector v
i
can calculated as follows,

A
T
Av
i
=
i
v
i
(6)
where
i
is the eigenvalue. Here the size of
covariance matrix would be M by M.Thus we can
have m eigenvectors instead of N
2
. Premultipying
equation 6 by A, we have
AA
T
Av
i =

i
Av
i
(7)
The right hand side gives us the M Eigen signs of
the order N
2
by 1.All such vectors would make the
image space of dimensionality M.
As the accurate reconstruction of the sign
is not required, we can now reduce the
dimensionality to M instead of M. This is done by
selecting the M Eigen signs which have the largest
associated Eigen values. These Eigen signs now
span a M-dimensional subspace instead of N
2
. A
new image T is transformed into its Eigen sign
components.
w
k =
u
k
T
(T - ) (8)
where k =1,2,.M.
The Euclidean distance of the weight
vector of the new image from the sign class weight
vector can be calculated as follows,

k
= ||
k
|| (9)
where
k
is a vector describing the kth sign
class.Euclidean distance. The sign is classified as
belonging to class k when the distance
k
is below
some threshold value . Otherwise the face is
classified as unknown. Also it can be found
whether an image is a sign image or not by simply
finding the squared distance between the mean
adjusted input image and its projection onto the
face space.

2
= || -
f
|| (10)
where
f
is the face space and = T
i
is the
mean adjusted input.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
80
Using these we can say whether the image as
known sign image, unknown sign image and not a
sign image.
B. Interfacing Module:
In generally most of the personal
computers are provided with two serial ports and
one parallel port. A parallel port sends and receives
data bits very faster nut the required number of
wires is more whereas for serial communication it
will send one bit at a time through the single wire
and hence slower but the required number of wires
are less.
RS232 is meant for serial communication
transmission of data in order to connect the DTE
(Data Terminal Equipment) and DCE (Data
Communications Equipment). For example,
connecting a computer terminal with the printers,
modems, UPS and other peripheral devices.
RS232C is the latest one where RS232 is
Recommend Standard number and C is the latest
revision of the standard. It specifies that 25-pin D
connector and most of the PCs are equipped with
the male type D-connectors consists of only 9 pins.
III. Hardware Module:
After the selection of image based on the
index value obtained from RS232 cable is given as
an input to the hardware such as LPC2148. This
LPC2148 will be interfaced with both the audio
amplifier and LCD display.
A. LPC 2148:
ARM7 is most widely used in embedded
system application such as ranging from mobile
phones to automotive braking systems. The number
of transistors used in ARM7 is fewer which reduce
the costs and power consumption. The ARM7 is
based on a 16bit/32 bit with real-time emulation
and embedded trace support. This also provided
with the 512 kilobytes of embedded high speed
flash memory.
This LPC2148 also consists of 128-bit
wide memory interface and unique accelerator
architecture which will enable 32-bit code
execution at maximumclock rate.

Block Diagram:








Fig 2. Block diagramof typical sign recognition system
B. Liquid Crystal Display:
It is very thin and flat panel used for electronically
displaying information such as text, images as well
as moving pictures. It has enormous applications
include monitors for personal computers,
televisions, instrument panels, and other devices
ranging from aircraft cockpit displays to every-day
consumer devices such as gaming devices, clocks,
watches, calculators, and telephones.
Hardware Circuitry:

Fig 3. Before displaying the type of sign

Image
acquisition
Pre-
processing
Feature
Extractor
Sign image
Normalized
sign image
Training sets
Classifier
Sign database
Feature vector
known or
unknown
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
81

Fig 4. Voice alert in accordance with the type of sign
displayed in LCD
C. Audio amplifier:
It is an electronic device that increases the
strength of audio signals that pass through it. Audio
amplifies up to the level that is suitable for driving
loudspeakers. The different amplifiers that exist are
car audio amplifier, PC audio amplifier, TV audio
amplifier etc., and can be chosen based on the
application. Finally this audio amplifier output is
given to the speakers.
IV. Experimental Results:
The input images are captured by camera
where it will be processed and compared with the
database images. If any matching occurs then the
corresponding sign image will be displayed and
simultaneously the type of sign is also displayed at
the bottomof the GUI.
This would be done for various sign
images and voice alert also provided with respect to
the sign image displayed. LCD screen is also
interfaced in order to display the type of sign that is
present.


Fig 5. Traffic sign recognition results
In the above figure shown that the sign board is
displayed on the GUI as well as it will show the
type of sign in the box at the bottom. Here it will
display the Right hand curve sign image and like
manner we will process and display the image
using PCA in an effective way.
V. Conclusion:
This paper provides an effective
recognition of traffic signs using PCA algorithm
and thereby providing the voice alert to the drivers
as well as display. After processing of the input
sign image in GUI, then the allotted index number
in accordance with the sign image is given to
RS232 cable. Based on the index number the
corresponding sign image direction will be audible
in the speakers and displayed on LCD module. In
the future we may expect the same feature of traffic
sign recognition without the interfacing of GUI
module with the LPC2148 using RS232 cable in
TMS320CXXX in which the performance also be
increased.
References:
[1] Prof. V.P. Kshirsagar, M.R.Baviskar, M.E.Gaikwad, Face
Recognition Using Eigen faces.
[2] Facial Recognition using Eigenfaces by PCA by Prof. Y.
Vijaya Lata, Chandra Kiran Bharadwaj Tungathurthi, H.
RamMohan Rao, Dr.A. Govardhan, Dr.L.P. Reddy,
International J ournal of Recent Trends in Engineering,
Vol. 1, No. 1, May 2009.
[3] FaceRecognition using Eigenface Approach by Vinaya
Hiremath, Ashwini Mayakar,.
[4] FaceRecognition using Eigenfaces and Neural Networks
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
82
by Mohd Rozailan Mamat, Mohamed Rizon, Muhammad
Firdaus Hasim, American J ournal of Applied Sciences 2
(6): 1872-1875, 2006 ISSN 1546-9239, 2006 Science
Publications.
[5] L.D. Lopez and O. Fuentes, "Color-based road sign detection
and tracking", Proc. Image Analysis and
Recognition(ICIAR), Montreal,.CA, Agust 2007.
[6] ARM Data Manual sheets
http://www.keil.com/dd/docs/datashts/philips/lpc2141_42_44_4
6_48.pdf
[7] A. D. L. Escalera, J . M. A. Armingol, and M. Mata, "Traffic
sign recognition and analysis for intelligent vehicles ",
Image and Vision Computing, vol. 21, pp. 247258, 2003.

































Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
83




Ethernet Based Intelligent Security System
Ch.J agadeesh
1
, Tulasi Sanath Kumar
2

1
PG Student, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India.
2
Asst. Professor, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India

1
j agadeeshchal l a1508@gmai l . com
2
t ul asi sanat h@gmai l . com

Abstract:In this paper we design and implement an
embeddedsurveillance system by use of ultrasonic
signal coding ofultrasonic sensors with multiple
pyroelectric infrared sensors(PIR) to detect an
intruder in a home or a storehouse. The PIRsensors
are placed on the ceiling, and the ultrasonic
sensormodule consists of a transmitter and a receiver
which are placedin a line direction; however,
ultrasonic sensors with the samefrequency are subject
to interference by crosstalk with each otherand have a
high miss rate. To overcome these disadvantages
ofthe ultrasonic sensor, our design reduces the miss
rate from theenvironmental interference by using an
ultrasonic coding signal.Both ultrasonic sensors and
PIR sensors are managed by themajority voting
mechanism (MVM).

Index Terms:Embedded Surveillance System;
Majority VotingMechanism; PIR Sensor; Ultrasonic
Sensor

I. INTRODUCTION

Recently surveillance systems have become more
importantfor everyones security. The embedded
surveillance system,frequently used in a home, an
office or a factory [2-4], uses asensor triggered to
turn on a camera [5-6]. Some designs usedifferent
types of sensors to achieve reliability by means of
thedifferent features of each sensor [7-8]. In this
paper we extendour previous design not only by
using both multiple PIR sensors and ultrasonic
sensors as a sensor group, but also byusing the
MVM. Ultrasonic receivers and transmitters
arelocated at opposite ends [9-10]. However, to
reduce theinterference from other frequencies in
ultrasonic signals, weuse a coding signal to enhance
the ability to distinguish therandom interference [11].
To enhance system reliability in theexperiment, we
focus on how to improve the shortcomings ofthe
ultrasonic sensor. Some research explores the
influence ofattenuation in air and crosstalk of
ultrasonic signals by using acoding signal [12], while
some provides improvement ofthe ultrasonic signal
by using different coding signal types. Other research
uses the application of a coding signalto increase
resolution and contrast of images. Yet
anotherapproach builds a 3D image with an
ultrasonic sensor in the PNcode that solves the
problem with time delay. To enhancethe reliability of
the ultrasonic sensors group, we proposeadding to the
number of bits with coding to reduce theprobability
of code breaking.

II. SYSTEM ARCHITECTURE

Fig. 1 shows the home embedded surveillance system
which has two groups of sensors, indoor and outdoor.
The outdoor sensor group contains a number of PIR
and pressure sensors placed near windows and doors
of a home. When the outdoor sensors sense an
intruder, the MCU is woken up and turns on the
power for the indoor PIR and ultrasonic sensors for
the Majority Voting Mechanism. When this is
completed, the decision signal passes to the
embedded board GPIO (General purpose input and
output). The software module of the power embedded
board turns on the Web camera to capture images and
user can view the images captured by the home
surveillance system.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
84



///

Fig.1.The home embedded surveillance system
with ultra-low alert & GSM modem

A. PIR Sensor

PIR is basically made of Pyroelectric sensors to
develop an electric signal in response to a change in
the incident thermal radiation. Every living body
emits some low level radiations and the hotter the
body, the more is emitted radiation. Commercial PIR
sensors typically include two IR-sensitive elements
with opposite polarization housed in a hermetically
sealed metal with a window made of IR-transmissive
material (typically coated silicon to protect the
sensing element). When the sensor is idle, both slots
detect the same amount of IR, the ambient amount
radiated from the room or walls or outdoors. When a
warm body like a human or an animal passes by, it
first intercepts one half of the PIR sensor which
causes a positive differential change between the two
halves. When the warm body leaves the sensing area,
the reverse happens, whereby the sensor generates a
negative differential change. These change pulses are
what is detected. In order to shape the FOV, i.e. Field
Of View of the sensor, the detector is equipped with
lenses in front of it. The lens used here is inexpensive
and lightweight plastic materials with transmission
characteristics suited for the desired wavelength
range. To cover much larger area, detection lens is
split up into multiple sections, each section of which
is a Fresnel lens. Fresnel lens condenses light.
Providing a larger range of IR to the sensor it can
span over several tens of degree width. Thus total
configuration improves immunity to changes in
background temperature, noise or humidity and
causes a shorter settling time of the output after a
body moved in or out the FOV. Along with
Pyroelectric sensor, a chip named Micro Power PIR
Motion Detector IC has been used. This chip takes
the output of the sensor and does some minor
processing on it to emit a digital output pulse from
the analog sensor. Schematic of PIR sensor output
waveform is shown in Fig. 2.


Fig.2.PIR sensor output waveform

B. Ultrasonic Sensor

Ultrasonic sensor is non-contact distance
measurement module, which is also compatible with
electronic brick. Its designed for easy modular
project usage with industrial performance. A short
ultrasonic pulse is transmitted at the time 0, reflected
by an object. The senor receives this signal and
converts it to an electric signal. The next pulse can be
transmitted when the echo is faded away. This time
period is called cycle period. The recommend cycle
period should be no less than 50ms. If a 10s width
trigger pulse is sent to the signal pin, the Ultrasonic
module will output eight 40 kHz ultrasonic signal and
detect the echo back. The measured distance is
proportional to the echo pulse width and can be
calculated by the formula above. If no obstacle is
detected, the output pin will give a 38ms high level
signal.

C. GSM Modem

Global System for Mobile communications (GSM:
originally from Group Special Mobile) is the most
popular standard for mobile phone in the world.
GSM/GPRS Smart Modem is a multi-functional,
ready to use, rugged and versatile modem that can be
embedded or plugged into any application. The Smart
Modem can be customized to various applications by
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
85



using the standard AT commands. The modem is
fully type-approved and can directly be integrated
into your projects with any or all the features of
Voice, Data, Fax, SMS, and Internet etc.

III. WORKING CIRCUIT
The total system can be divided into three segments

A. Sensor and signal processing segment:

This segment consists of five parts: - PIR sensor
module: The PIR sensor module is fed from the
output of fixed output voltage regulator IC LM7805.
PIR positive input terminal is fed with a +5V supply
and negative terminal is grounded. PIR sensor
module output pin is connected to MCU pin. For re-
triggering purpose, a jumper (JP) is attached on the
COMMON (C) pin and HIGH (H) pin. - LM7805:
LM7805 is a fixed output voltage regulator IC. 7805
takes +12V input and gives a fixed regulated output
voltage of +5V. - LM35: This is temperature sensor
IC rated for full -55 to +150C temperature range.
This is a transducer IC that takes voltage input and
gives a voltage output proportional to the ambient
temperature. +VS pin is connected to the output pin
of LM7805 and the VOUT pin is connected to one of
the analog input channels available on MCU.

- Switch: This is a mechanical switch which is of NO
(Normally Open) type.

One end of the switch is connected to the +5V supply
and the other end is connected to one of the MCU
input pins. For practical use, electronic remote
controlled switch is a better option to secure the
system operation.

- MCU: For this system, PIC 16f876A is used as the
MCU, i.e. Microcontroller unit. It has built-in
USART module which is necessary for passing AT
commands to the GSM modem.

The PIR sensor module output is tied to the pin RB I.
The output of temperature sensor IC and one end of
the mechanical switch is connected to pins RA3 and
RBO respectively.

A LED is connected to the pin RC3. The MCLRI
VPP is connected to +5 V supply. A 4 MHz crystal is
connected between OSCI an OSC2 pins. This crystal
determines the clock speed of the MCU operation.

B. Alarm segment:

This segment consists of three parts: - 74LS75: This
is a D-Latch IC. The input voltage level on D 1 is
kept unchanged on Ql and inverted on Q2. Inverted
output images the voltage level set by MCU keeping
the alarm on even when MCU is at SLEEP mode. -
Alarm: The alarm has two pins- VCC and GNO. The
power pin is connected to Ql output pin of the D-
Latch IC. The alarm can be set to ring by MCU. -
MCU: Pin RC4 of PIC 16f876A is connected to the 0
1 input of 74LS75.

C. GSM Modem interfacing segment:

As GSM modem uses serial communication to
interface with other peripherals, an interface is
needed between MCU and GSM modem. This
segment consists of four parts:

-DB9 male connector: The serial port used here is a 9
pin DB9 male connector as the GSM modem side
uses a female connector. Pin 14 and 13 of MAX232
are connected to pin 2 and 3 ofOB9 respectively. Pin
5 ofOB9 is grounded.

MAX232: This particular IC is necessary for
increasing the voltage swing at the outputs. It takes
OV and +5V inputs and makes it a +12V and - 12V
output voltages. This increased voltage swing is a
requirement for serial communications. Two 1 F
capacitors are connected between pins 4, 5 and 1, 3
of MAX232. V+and V- pins are fed from VCC and
GNO, i.e. G round through two 1 IF capacitors.
Between VCC and GNO pins, one 10 F capacitor is
placed.

- GSM modem: GSM modem is connected through a
DB9 female connector to the interfacing circuit.

- MCU: The VCC, i.e. power pin, TTL input and
TTL output pins of MAX232 are connected to the
pins RCO, RCI and RC2 of MCU respectively.

3.1CIRCUIT OPERATION

A. Sensor and signal processing segment:

As the jumper of PIR sensor module is placed
between C and H, the output will stay on the entire
time something is moving. The regulator IC serves
regulated +5V to the LM35 and PIR sensor module.
Prior to any operation, external interrupt is disabled
in software of MCU. When the mechanical switch is
closed, pin RBO gets an input voltage. This sets the
system to run. The analog voltage output fromLM35
is taken and converted to an equivalent binary value
which represents the ambient temperature. As PIR
sensor module does not perform satisfactorily below
15C temperature, MCU monitors the temperature
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
86



and light LED on pin RC3 when the temperature is
equal to or greater than the critical temperature. After
the LED is on, the MCU waits a pre-defined time for
the place to be fully evacuated. After that time is
over, the system is online. After activation of the
system, if there is any movement on that place within
the coverage region of the PIR sensor module, it
outputs a pulse which is taken as input by MCU.
MCU then waits a pre defined time and checks for
that signal again. This is done for avoiding false
triggering.

B. Alarm segment:

RC4 remains HIGH right from the beginning. Thus,
the output pin IQ of 74LS75 stays LOW and the
alarm does not ring. If the signal is still present
during the second check, MCU makes pin RC4
LOW. This makes a HIGH on IQ of 74LS75 and the
alarm rings.

C. GSM Modem interfacing segment:

MCU makes HIG H on RCO which in turn, activates
MAX232 IC. Then MCU starts sending AT
commands to the GSM modem through the pins RCI
and RC2. The commands are sent through the
interface to the modem. The modem receives the
commands and sets up a call to a pre-defined number.
The call is not disconnected until the call time - up or
the recipient disconnects the call. After the call is
disconnected, MCU goes to SLEEP, i.e. low power
consuming mode. Before going into SLEEP, MCU
enables the external interrupt in software. When the
mechanical switch is open, an interrupt occurs and
MCU is brought out of SLEEP mode.

3.2. SOFTWARE

The whole system is built around a MCU. MCU
requires to be burned with software written for
specific applications. The code is written using
ASSEMBLY language and compiled using MPLAB.
MPLAB generated a hex file which is burned using a
burner into the IC. This section demonstrates the
flowchart of the software which helps to visualize the
coding steps as shown in fig.3. At the beginning of
the program, external interrupt of MCU is disabled in
software. Therefore, any signal input on the pin RBO
cannot generate interrupt. Then, MCU looks for the
switch whether it is closed or open. When the switch
is open the signal is LOW and when the switch is
closed, the signal is HIGH on pin RBO. If the signal
is LOW, MCU repeatedly checks for the switch
status.

When the signal gets HIGH, MCU converts the
analog signal from the temperature sensor to the
binary equivalent and checks repeatedly if the
temperature of the surrounding is greater or equal to
15 Celsius. When the temperature rises to 15
Celsius or more, MCU waits for a pre-defined time
before executing any instruction. This wait state is
introduced to ensure proper evacuation of the place
where the system is to run. After the wait state is
over, MCU starts checking for any signal from the
PIR sensor module. When there is no signal from the
sensor, MCU checks the status of the switch. If the
switch is still closed, it continues to check for sensor
signal.Wait state allows the call to be completed
successfully. After that, MCU enables the external
interrupt and goes to SLEEP mode. Enabling the
external interrupt prior to SLEEP mode ensures that
MCU will wake from the SLEEP mode whenever
there is a HIGH to LOW transition on RBO, i.e.
when the switch gets opened. When an external
interruption occurs, MCU wakes up from sleep mood
and disable the external interrupt and the program
goes to the beginning of the algorithm.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
87





Fig.3.Software flowchart

IV. IMPLEMENTATION AND EXPERIMENT
RESULTS

In the experiment results we found that an ultrasonic
signalwould be affected by environment sounds and
the amplitude ofthe reference voltage. Those factors
affect the transmissiondistance and the error rate of
detecting. We therefore put thetransmitter and the
receiver on both ends of the sensing areaand make
sure the intruder passes through if the outside
grouphas detected an individual.

Fig.4 shows an ultrasonic signal being interfered with
byanother ultrasonic frequency. As ultrasonic signal
interferencecauses difficulties in setting up the
amplitude of the referencevoltage of the circuit,
therefore we use the coding signal toreduce ultrasonic
signal interference based on thecharacteristics of the
ultrasonic signal.

Fig.4. Relationship between reliability and
number of bits of the ultrasonic signal code



Fig.5. Ultrasonic signal interfered by another
frequency

In Fig.6 we see that the signal has been coded by
oursystem. The coded signal is not affected by
another frequencybecause our design receives a
signal through the code insteadof through its signals
frequency.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
88





Fig.6. Our ultrasonic coding signal in the scope

Fig.7 shows the signal judgment of our experiment.
Whenjudging a coding signal, one method of our
design counts therising edge number. If the rising
edge number is equal to two, itmeans the signal is
correct. Lines A and B of Fig.7 show acoding
interval.


Fig.7. Judgment from received ultrasonic coding
signal

Fig.8 shows our design that consists of the
internalsoftware module and the home embedded
system softwaremodule. When an intruder has been
detected, the MCU wakesup the majority decision to
test the threshold and then turn onthe power supply
for the indoor sensors. If the indoor sensorsdetect no
intruder when the outdoor sensors are misjudging,the
MCU turns off the power of the indoor sensors and
goesback to the alert state. If the indoor sensors
detect an intruder,the MCU turns on the Web camera
to capture images inkeeping with the decision of the
MVM.


Fig.8. Software flowchart of embedded
surveillance system [12]

After the intruder has been detected outdoors, the
MCUsoftware submodule of the ultrasonic coding
signal is executedas shown in Fig.9. Our design
transmits the ultrasoniccoding signal, and the receiver
checks whether the codingsignal is correct or not. If
the coding signal is correct, theultrasonic sensor
group continues for some seconds to makesure that
there is no intruder. If the coding signal is not
correct,our design also starts the majority voting
mechanism (MVM)to make sure whether there is any
detection.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
89




Fig.9. Flowchart of detection by ultrasonic coding
signal

In the experiment the ultrasonic signal of the receiver
hasthe same direction as the transmitter, and we find
that if theamplitude of the voltage waveform has
been reduced toapproximately the same as the
comparator's reference voltageour design can work
normally. The scattering causes theamplitude of the
voltage waveform to become gradually lowerthan the
reference voltage. To increase the amplitude of
thevoltage waveform we place a PET bottle at the
front end forfocusing. Fig.9 shows the distribution of
scattering afteradding a PET bottle focus. We have
found that the ultrasonicsignal increases at the central
point and achieves the focusingeffect.


Fig.10. Curves showing distribution of scattering
after adding PET bottle focus

Fig.10 shows the arrangement of our
experimentalenvironment that detect intruders in a
suitable place. We placethe PIR sensor on the ceiling
or above the detection area.Transmitter and receiver
of the ultrasonic sensor module areplaced in a line
direction. When an intruder enters thedetection area,
the ultrasonic coding signal will be blockedand the
PIR sensors will detect temperature changes.


Fig.11. Arrangement of our experimental
environment

Table I compares our coding signal and non-coding
signal.The coding signal is not interfered with by
other frequenciesunless their patterns are similar. It is
easier to break anon-coding signal than a coding
signal. When we add to thebits of the ultrasonic
coding signal, the message type rise with2
N
A non-
coding signal transmits just two types of messages,0
and 1. N means number of bits. With more message
types,more codes can be used in the same design to
decrease theprobability of breaking a signal. In our
design, when N isequal to 8, the message
combination of the ultrasonic codingsignal is 128
times better than that of non-coding signal, andthe
reliability is enhanced from 0.5 to 0.996.

TABLE I
COMPARISON OF OUR CODING SIGNAL
AND NONCODING SIGNAL


V. CONCLUSION

Our experiment shows two different types of sensors
whichare enhancing the overall sensing probability
by using theMVM to reduce the shortcomings of both
the ultrasonicsensors and the PIR sensors. By adding
an ultrasonic codingsignal our design reduces the
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
90



miss rate of the receiver withultrasonic sensors by
different patterns, improving thereliability of the
overall system.

VI. REFERENCES

[1] Ying-Wen Bai, Chen-Chien Cheng and Zi-Li Xie,
Use of Ultrasonic Signal Coding and PIR Sensors
toEnhance the Sensing Reliability of an
EmbeddedSurveillance System, 2013 IEEE.

[2] Jun Hou, Chengdong Wu, Zhongjia Yuan, Jiyuan
Tan, Qiaoqiao Wangand Yun Zhou, Research of
Intelligent Home Security SurveillanceSystem Based
on ZigBee, International Symposium on
IntelligentInformation Technology Application
Workshops, Shanghai, 21-22 Dec.
2008, pp. 554-57.

[3] Xiangjun Zhu, Shaodong Ying and Le Ling,
Multimedia sensornetworks design for smart home
surveillance, Control and DecisionConference,
2008, Chinese, 2-4 July 2008, pp. 431-435.

[4] L. Lo Presti, M. La Cascia, Real-Time Object
Detection in EmbeddedVideo Surveillance Systems,
Ninth International Workshop on ImageAnalysis for
Multimedia Interactive Services, 7-9 May 2008,
pp.151-154.

[5] Wen-Tsuen Chen, Po-Yu Chen, Wei-Shun Lee
and Chi-Fu Huang,Design and Implementation of a
Real Time Video Surveillance Systemwith Wireless
Sensor Networks, VTC Spring 2008. IEEE
VehicularTechnology Conference, 11-14 May 2008,
pp. 218-222.
[6] MikkoNieminen, TomiRaty, and
MikkoLindholm, Multi-SensorLogical Decision
Making in the Single Location Surveillance
PointSystem, Fourth International Conference on
Systems, France, 1-6March 2009, pp. 86-90.

[7] Ying-Wen Bai, Li-Sih Shen and Zong-Han Li,
Design andImplementation of an Embedded
Surveillance System by Use ofMultiple Ultrasonic
Sensors, The 28th IEEE International Conferenceon
Consumer Electronics, Las Vegas, Nevada, USA, 11-
13 Jan. 2010,11.1-3, pp. 501-502.

[8] Yang Cao, Huijie Zhao, Na Li and Hong Wei
Land-CoverClassification by Airborne LIDAR Data
Fused with Aerial OpticalImages, International
Workshop on Multi-Platform/Multi-SensorRemote
Sensing and Mapping (M2RSM), Jan. 2011, pp. 1-6.

[9] Hai-Wen Zhao, Hong Yue, and He-Gao Cai,
Design of a DistributedUltrasonic Detecting System
Based on Multiprocessor for AutonomousMobile
Robot, Proceedings of tbe 2007 WSEAS Int.
Conference onCircuits, Systems, Signal and
Telecommunications, Gold Coast,Australia, January
17-19, 2007, pp. 59-64.

[10] Zi-LI Xie and Zong-Han Li, Design and
implementation of a homeembedded surveillance
system with ultra-low alert power,
IEEETransactions on Consumer Electronics, Feb.
2011, pp. 153-159.

[11] Francesco Alonge, Marco Brancifortem and
Francesco Motta, A novelmethod of distance
measurement based on pulse position modulationand
synchronization of chaotic signals using ultrasonic
radar systems,IEEE Transactions on Instrumentation
and Measurement, Feb.2009, pp.318-329.

[12] ShragaShoval and Johann Borenstein, Using
Coded Sognals to Benefitfrom Ultrasonic Sensor
Crosstalk in Mobile Robot Obstacle
Avoidance,IEEE International Conference on
Robotics and Automation, Seoul,Korea, 21-26 May,
2001, vol.3, pp. 2879-2884.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
91


9

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
92
Modified Artificial Potential Fields Algorithm for Mobile Robot Path Planning
Amrita.K, Keerthana.S.T and Vasanthkumar.B
Department of Robotics and Automation,
PSG College of Technology,
Coimbatore,
India.
amritakrishnaraj@gmail.com keerthanaast@gmail.com vasanthbe@gmail.com

Abstract The problem of path planning is
studied for the case of a mobile robot moving in
an environment filled with obstacles whose
shape and positions are not known. Path
planning is dynamic when the path is continually
recomputed as more information becomes
available. A computational framework for
dynamic path planning is proposed which has
the ability to provide navigational directions
during the computation of the plan. Path
planning is performed using a potential field
approach. This paper introduces a new
algorithm, bidirectional artificial potential fields,
capable of planning paths in unknown, partially
known, and changing environments in an
efficient, optimal, and complete manner. The
algorithm uses the potential field values
iteratively to find the optimum points in the
workspace in order to form the path from start
to destination. In the algorithm, motion planning
is done continuously (dynamically), based on the
systems current position and on its feedback.
Keywords Mobile robot, Path Planning,
Artificial Potential Field, Collision Free Path.

I. INTRODUCTION
Obstacle avoidance is one of the key issues to
successful application of mobile robot systems.
Based on the configuration space and the
destination generating a path (finding a continuous
route) in a 2D environment with unknown obstacles
represent still a fundamental problem to be solved.


The abominable and potentially dangerous objects
encountered by the robot in its route to destination
are the obstacles and must be avoided. The capacity
to move without collision in the uncertain
environment taking into consideration the
perception of the system is a fundamental problem
to be solved in the autonomous mobile robot field.
We study the problem of robust navigation for
indoor mobile robots.
Numerous algorithms and methods have been
proposed for path planning of mobile robots.
Artificial potential fields (APF) are the method
most widely used due to its mathematical simplicity
and ease of implementation and high efficiency.
The Artificial potential field method was proposed
by Khatib, which is a virtual force field method
[10]. The basics of artificial field method are
finding a function that represents the energy of the
system and forces the robot to move towards the
destination which posses the minimum energy
value [12]. The robot is made to travel from high-
potential to low-potential state. Moving the robot
from source point to the destination in a downhill
approach is mathematically termed as the gradient
descent (i.e.)
) (
k k
x f x x (1)
The motion terminates as the gradient vanishes.
Although these methods are fast and efficient, they
have the following drawbacks and limitations as
discussed in [7]:
a) Trap situations due to local minima.
b) No passage between closely spaced
obstacles.
c) Oscillations in the presence of obstacles.
d) Oscillations in narrow passages.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
93
e) Incapability to avoid dynamic obstacles.
While trying to solve the above stated problems,
researchers have introduced an additional potential
so that the destination becomes the global minima.
Still others have made changes to the path and tried
to avoid the minimum points. A few researchers
have tried to incorporate artificial intelligence
methods into the potential methods.
The aim of this paper is to use the artificial
potential field method to detect and avoid the static
as well as the dynamic obstacles. The navigation
architecture is made of a global planner, local
planner and a cost map. Localization of the robot is
done with the visualization obtained from amcl.
The proposed algorithm was implemented in the
robot operating system (ROS), a LINUX based
framework used for controlling robots. The
reminder of the paper is arranged as follows:
section1 discusses the algorithm for sensing and
avoiding the static obstacles. Section2 examines the
performance of this method through simulations. In
section 3, new APF based method for dynamic path
planning is explained. Finally section 4 discusses
the conclusions and further work.

II. STATIC OBSTACLE AVOIDANCE
Since the model under study is a mobile robot
possessing freedom along the X and Y and rot
(rotation about Z) motion, the gross workspace is a
2D space. Each point of the workspace is
considered and the potential at each point is
determined. The notion of cell is introduced for
further use in this paper which represents each pixel
point in the camera transformation matrix. The
workspace is considered to be a discrete space of N
coordinate cells X=(x, y) and each cell is either
obstacle or free space.
The potential is given as
otal
(x)=
source
(x)+
destination
(x)tacle
(x) [1]
(2)
The individual terms are expressed as

source
(x) =/(x, a) (3)

destination
(x) =/(x, b) (4)
tacle
(x) =/(x, c) (5)
Where, (x, a) is the Euclidean distance between
the locations of robot and the cell. (x, b) is the
Euclidean distance between the destination of robot
and the cell. (x, c). is the minimum distance
between the affected areas of obstacle and the
location of the robot. and , the gains
coefficients of attraction and repulsion functions
respectively are positive constants.
Using the expression (2) the potential for each cell
in the workspace is determined. These determined
potentials are then sorted in descending order. The
cells with high potential are found to occupy
positions close to the source point and destination
point. Of the sorted potentials, the top 60% is
considered to determine the threshold value, a
notion to be continued for the rest of the paper. A
threshold value is picked from the sorted list and all
the cells with potential values greater than the
threshold is studied. If the set threshold value is
large, two distinct clusters of marked cells around
start and destination points is obtained. As the
threshold value is gradually decreased, these two
clusters get bigger and bigger until they run into
each other
The threshold values are altered until a particular
value is reached that guarantees that there is one
and only one cell that connects the source point
cluster to the destination point cluster. Values
smaller than this threshold value would make a
connection between two clusters, but this threshold
value is the biggest value that by using all cells with
potential value bigger than this value makes it
possible for having a path between start and end
points. The point that connects these clusters is
called the midpoint. It is observed that, there is a
path from the start point to the midpoint, and there
is a path from midpoint to the destination within the
range of the clusters. This is implementing by using
the breadth first search algorithm.
The next step is to find the midpoint between the
source and the current midpoint assuming that only
a few cells of the workspace are available. This
process is iteratively repeated until a straight line
path exists between consecutive points.
III. SIMULATION RESULTS
Two examples with two different environments are
studied. If a collision is detected, the robot due to
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
94
the prime repulsive force is deviated. Using this
method, a smooth path with reasonable distance
form obstacles is identified while keeping the path
as short as possible. The algorithm has been
simulated in the ros4mat an open source
client/server library which supports the Kinect
interface. Ros4mat allows a platform independent
connection between the robot and multiple clients.




Fig. 1 Simulation Results of Static Obstacle
Sensing

IV. DYNAMIC OBSTACLE AVOIDANCE
The proposed algorithm is based on the principles of
APF path planning and our proposed ideas. These ideas
are suggested to avoid collision with moving entities and
to ensure a reliable on-line path planning where the
robot employs continuous sensing and acting. The
proposed algorithm is presented as a sequence of
following steps.
A.Analysis of Obstacle Behaviour
First step in the analysis of behaviour of
dynamic objects is learning from the data obtained
about their previous states. The most basic
conceptualization of a moving objects spacetime
behaviour is a geo-spatial lifeline [5] also referred
to as a movement path or trajectory, which
describes a sequence of visited locations in space, at
regular or irregular temporal intervals [9]. This
process is known as spatiotemporal data mining a
field still under persistent research and only limited
researchers are working on the human interface
spatial data analysis. Several prediction techniques
are used to predetermine the location of moving
entities: neural networks ([2][12]), Markov models
[6] Ehrenfest chains or Kalman filter ([8][3][4]).
A.Markov Models
Though the motion of obstacle is continuous, the
imaging and other communication technologies support
sampling only at discrete time. These sample datum
obtained are analyzed to determine the behaviour of
moving entities. A Markov Chain consists of a countable
(possibly finite) set S (called the state space) together
with a countable family of random variables X
0
, X
1
, X
2
,
... with values in S such that
P[X
l+1
=s | X
l
=s
l
, X
l1
=s
l1
, , X
0
=s
0
] =
P[X
l+1
=s | X
l
=s (6)
This fundamental equation is referred to as the
Markov property.
Consider a autonomous mobile system that may
be described at any time as being in one of the N
countable states S
0
S
1
S
n
. At regularly spaced
discrete times, the system undergoes a change of
state within the workspace. The time instants
associated with the change of state is denoted as t =
0,1,... , and the actual state at time t is given as St
[19]. X is a set of stochastic variables {X
t,
t T}.
The random variables X
0
, X
1
, X
2
, ... are dependent.
Markov property expresses the assumption that the
knowledge of the present (i.e., Xl = sl) is relevant to
predictions about the future of the system, however
additional information about the past (Xj = sj , j l
1) is irrelevant.
For the following consideration it is assumed
that the chains are time-homogeneous:
A
ij
=P (X
t+1
=i|X
t
=j) =P(X
t
=i|X
t-1
=j), t
T, i, j S . (7)
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
95
Unless stated to the contrary, all Markov chains
considered in this paper are time homogeneous and
we simply represent the matrix of transition
probabilities as P =(Pij ). P is called the transition
matrix. The non-homogeneous case is generally
called time inhomogeneous or non-stationary in
time.
For a homogeneous Markov chain the transition
probabilities can then be noted in a time
independent stochastical matrix A:
A=[a
ij
], a
ij


0 i, j S, a
ij
=1 (8)
Where A is called the transition matrix .It can be
shown that the probability of getting in m steps to
state j, starting from state i
a
ij
=P(X
t+m
=j | X
t
=i) (9)
can be computed as the m-th power of the transition
matrix a
ij
=A[i,j]. Recapitulating, a first-order time-
homogeneous Markov Chain can be defined as a 3-
tuple, consisting of the set of states S, the transition
matrix A and the initial distribution vector
=(S, A, ) (10)
The embedded-renewal process is considered to
determine the time at which the moving entity
intrudes the path if the mobile robot. The time of
the nth visit to point y is denoted by
T
y,n
=min{kN+:Ny,k=n} (11)
The function is expressed as
N
y,n
=i=1n1(X
i
=y) (12)
Where N
y,n
is the number of visits
The behaviour of the system thus analyzed from the
sampled data and is simultaneously expounded to
obtain the accurate time and point of collision. The
feedback data obtained from the odometers paired
with the mobile robot wheel helps in the
localization of the robot

A. Collision Avoidance
It is ensured that the mobile robot avoids the
obstacles by passing them with safe potential
distance. The analyzed behaviour determines the
exact point of collision of the robot and the
obstacle, henceforth providing a time frame in
advance to take the necessary precautions. The
velocity of the robot can be altered (i.e.) increased
or decreased with an allowable scaled factor as
measure. This step depends on the velocities of the
robot and the obstacles. An alternative proposed
method utilizes the earlier calculated potentials. The
map at the location of the expected collision point is
retrieved from the global map and is further
analyzed by applying the iterative midpoint method.
Size of the retrieved map depends on the direction
of motion and the dimensions of obstacles.

V. CONCLUSION
In this paper, modified artificial potential field
algorithm is used to solve the path planning in an
unknown and dynamic environment. The
simulation results show that the proposed algorithm
is fast and efficient. In addition, it overcomes the
drawbacks and limitations of traditional artificial
potential field. The proposed algorithm has the
capabilities like escapes from dynamic obstacles.
The variations in the velocity of the obstacle might
lead to changes in behaviour and result in collision.
As a further work the dynamic path planning in the
ROS platform and real time implementation is
under study.

.
REFERENCES

[1] H.Adeli, M.H.N.Tabrizi and A. Mazloomian, Path
Planning for Mobile Robots using Iterative
ArtificialPotential Field Method, IJCSI International
Journal of Computer Science, vol. 4,no. 2, pp. 28-32 , July
2011.
[2] S.Akoush and A.Sameh, Mobile User Movement
Prediction Using Bayesian Learning for Neural Network,
International conference on Wireless communications and
mobile computing, 2007.
[3] J.M.Franois, G.Leduc and S.Martin, Learning movement
patterns in mobile networks: a generic approach',
Proceedings of European Wireless, pp. 128-134, 2004.
[4] J.Froehlich and J.Krumm, Route Prediction from Trip
Observations, Society of Automotive Engineers (SAE)
World Congress, 2008.
[5] K.Hornsby and M.J.Egenhofer, Modeling moving objects
over multiple granularities, Annals of Mathematics and
Artificial Intelligence, vol. 36,pp. 177194, 2002.
[6] Y.Ishikawa, Data Mining for Moving Object Databases,
Mobile Computing and Computational Intelligence, John
Wiley & Sons , Chapter 12 (to appear), 2008.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
96
[7] Y.Koren and J.Borenstein, Potential Field Methods and
Their Inherent Limitations for Mobile Robot Navigation,
Proceedings of the IEEE Conference on Robotics and
Automation, pp. 1398-1404, April 1991.
[8] P.Kontkanen, P.Myllymki, T.Roos, H.Tirri, K.Valtonen
and H.Wettig, Probabilistic Methods for Location
Estimation in Wireless Networks, In Emerging Location
Aware Broadband Wireless Ad Hoc Networks, Springer US,
pp. 173-188, 2005.
[9] P.Laube, S.Imfeld and R.Weibel,Discovering relative
motion patterns in groups of moving point objects,
International Journal of Geographical Information Science,
vol. 19,pp. 639668,2005.
[10]M.Mohamed and M.Abbas, Obstacles Avoidance for
Mobile Robot Using Enhanced Artificial Potential Field,
Al-Khwarizmi Engineering Journal, vol. 9, pp. 80-91,2012.
[11]L.R.Rabiner, A Tutorial on Hidden Markov Models and
Selected Applications in Speech Recognition, Proceedings
of the IEEE, vol. 77,1989.
[12] K.Uyanik, A study on Artificial Potential Fields,
KOVAN Research Lab, 2004.
[13] L.Vintan, A.Gellert, J .Petzold and T.Ungerer,Person
Movement Prediction Using Neural Networks, Workshop
on Modelling and Retrieval of Context, 2004.


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
97
Protecting Privacy Preserving For Cost Effective Adaptive Actions
Using Memitic Algorithm
Ulsa Suresh Babu
M.Tech, Dept. of CSE, Audisankara College of Engineering & Technology, Gudur, A.P, India.

Abstract We present feasible algorithms to
solve this difficult optimization problem, and point
an end-to-end system based on our prior work on
the PREvent (prediction and prevention based on
event monitoring) framework, which noticeably
indicates the usefulness of our model.
Index term Memetic Algorithm, Local Search,
Branch and Bound.
I. INTRODUCTION
I propose a generic Cloud computing stack that
classifies Cloud technologies and services into
different layers. In this paper, we contribute to the
state of the art by formalizing this tradeoff as an
optimization problem, with the goal of minimizing
the total costs for the service provider. We argue
that this formulation better captures the real goals
of service providers. Additionally, we present
possible algorithms to solve this optimization
problem efficiently enough to be applied at
composition runtime. We evaluate these
algorithms within our PREVENT (prediction and
prevention based on event monitoring) framework.
The below table show the SLO's of service level
agreements.
Table 1
SLO's of service level agreements
# SLO Name Description
1 Time to Offer
Time between Receiving the
RFQ and responding with an
offer(in working days)
2
Order
Fulfillment
Time
Time between receiving the
order and finishing the
process (in working days)
3
Process Lead
Time
Time between initializing
the process and finishing it
(excluding activities at
customer side) in working
days.
4
Cost
Compliance
Cost overrun with regard to
the offer in % of the offer;
5
Product as
Specified
Product is exactly as
specified.

II. PROBLEM FORMULATION
In this section, we formalize the problem of
selecting the most cost-effective adaptation actions
to prevent one or more predicted SLA violations.
We consider an interaction of the service
composition with a given client, who has a given
SLA with the composition provider. Let I be the
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
98
set of all possible composition instances of this
client, and let i I be concrete instances that we
can monitor. As part of the SLO definition, a
penalty function is associated with all SLOs using
the PREVENT tooling.
Penalty functions for SLOs can take many
different shapes. The most important ones are:
1. constant penalty (a constant payment needs to
be made if a certain SLO threshold value is
surpassed),
2. staged penalty (similar to a constant penalty, but
with different levels of penalty),
3. linear penalty (the penalty is linearly increasing
with the degree of violation), and
4. linear penalty with cap (the penalty is linearly
increasing up to a maximumvalue).
Even though these functions span many different
types of mathematical functions, they share two
essential characteristics.
i. First, SLA penalty functions are always
monotonically increasing.
ii. Second, SLO penalty functions always have a
point discontinuity in a special violation
threshold point.
This also means that penalty functions are
generally discontinuous. Furthermore, this
property signifies that there is no incentive for the
service provider to apply further adaptation and
improve an SLO value below because all further
improvements do not further reduce his costs (they
are already 0 for this SLO).
To prevent violations, we are able to apply a
number of possible adaptations to an instance.
Selecting the most cost-effective adaptation
actions means finding the adaptation actions that
minimize the total costs for the service provider.
The total costs TC are defined in Equation 1 as the
sumof the costs of SLA violations after adaptation
(VC) and the costs of adaptation (AC).
AC is the sumof the costs of all applied adaptation
actions and VC is defined as the sumof all penalty
functions applied to an instance. However, the
SLO Predictor provides estimations for SLOs
based on instance data. However, not all
combinations of adaptation actions are legal. Some
adaptation actions are mutually exclusive while
others depend on each other.
III. ALGORITHMS
Now discuss different approaches for finding
solutions to this problem. These algorithms are
implemented in the Cost-Based Optimizer
component. Optimization is always triggered by a
predicted violation of at least one SLO, and
receives as input a list of monitored facts and
estimates of the current instance.
1. Branch-and-Bound
Branch-and-bound is a very general deterministic
algorithmfor solving optimization problems. The
high-level idea of this approach is to enumerate
the solution space in a smart way so that at least
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
99
some suboptimal solutions can be identified and
discarded prematurely.

Fig. 1 General Branch-and-Bound Algorithm
We describe our general Branch-and-Bound
algorithm in Fig. 1. The algorithm is easy to
understand. What is the most important is the
implementation of Line 13, the rules for pruning
the search tree.
In our Branch-and-Bound approach, we prune a
partial solution in two cases: 1) if the partial
solution already contains at least one conflict or 2)
if the partial solution already prevents all SLA
violations. In general, this approach is suboptimal.
Even though the order in which we investigate
actions has no impact on the quality of our
solution, the order may have an impact on the
number of solutions we are able to prune.
Therefore, we can conclude that it is beneficial to
investigate actions in a specific order that
maximizes the number of solutions that can be
pruned. We specify two possible criteria for this
ordering: 1) the impact of an action on the SLOs,
and 2) the utility of an action. We will now define
those two orderings.
2. Local Search
While the Branch-and-Bound algorithm discussed
above has the advantage of always finding the
optimal set of actions for any composition
instance, the execution time of the algorithm
increases exponentially with the number of
available actions. Even though we can reduce the
runtime using impact- or utility-based sorting of
actions, the complexity still remains exponential.
Hence, there is an evident need to find strong
heuristics.
A simple heuristic that is often used to very good
ends is Local Search. Local Search is a
metaheuristic. The general idea is that in each
iteration the algorithm searches a specified
neighborhood for better solutions than the current
one. If at least one such solution is found, the
algorithmprogresses to the next iteration with one
of the better solutions. If no better solution can be
found in the neighborhood, the algorithm has
converged to a local optimum and is terminated.
Usually, this algorithm is repeated multiple times
with different starting solutions.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
100
This kind of algorithm typically depends on the
definition of: 1) a suitable neighborhood and 2) a
senseful selection of starting solutions. We use the
following neighborhood definition: A complete
solution vector is in the neighborhood of an
original solution if the two solutions represented
as binary vectors have a Hamming distance of 1.

Fig. 2 GRASP Construction
For selecting the start solutions, we use two
different approaches. The first and primitive one is
to select n start solutions with mbits set to 1 at
random. Alternatively, we propose to use an
algorithm commonly referred to as GRASP
(greedy randomized adaptive search procedure).
GRASP is essentially a variation of local search,
in which the start solutions are constructed using a
greedy heuristic. The idea is that GRASP can
converge to a better solution than a simple local
search because the start solutions are already
better than random start solutions. We have
sketched the construction heuristic that we have
used in our implementation of GRASP. The
heuristic is based on similar concepts that we have
already used in our
discussion of Branch-and-Bound: The idea is to
stop adding actions if either no more SLOs are
violated or no senseful actions are available
anymore (the RCS is empty), and to prefer adding
actions which have a high utility value. Hence, in
every step, the RCS consists of the r (maximum
size of the RCS) actions with highest nonnegative,
which have not yet been added and which do not
lead to a conflict.
3. Genetic Algorithm
As an alternative to locality-based heuristics (local
search, GRASP) we also present a solution based
on the concept of evolutionary computation. More
precisely, we use genetic algorithms (GAs) [23] as
a more complex, but potentially also more
powerful heuristic to generate good solutions to
the cost-based optimization problem. The overall
idea of GA is to mimic the processes of evolution
in biology, specifically natural selection of the
fittest individuals, crossover, and mutation.
We use the termfit to describe solutions with a
good (low) target function value. First, we
generate a random start population. For this, we
use the same primitive construction scheme stated
in local search. Every following iteration of the
algorithmessentially follows a three-step pattern.
First, we select a set of solutions from the
population to survive into the next generation. In
our genetic algorithm implementation, the fittest
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
101
solution is selected deterministically (elitism),
while all remaining slots in the next generation
population are selected using a process called
tournament selection. Second, crossover is used to
produce new solutions based on the selected ones
fromthe last generation. The main challenge of
implementing a strong crossover mechanism is to
ensure that the crossover product of two fit
solutions is also likely to be fit. This simple
procedure ensures that characteristics of both
original solutions are preserved.
Third, we use mutation to introduce entirely new
features into the population. The need for mutation
can be illustrated easily: Assume that a given
action a is not applied in any solution in the
population. After crossover, we may randomly flip
every bit in every solution in the population with a
very small probability. This means that most
solutions in the population are not mutated, but
sometimes new actions are applied, which are not
the product of crossover. Unfortunately, this
canonical GA implementation takes a significant
amount of time to converge against a solution,
because the solution space is searched solely
through the means of crossover and mutation.
This leads us to an adapted algorithm, which we
have sketched in Fig. 3. In such combinations of
GA and local search are often referred to as
memetic algorithms (MA). The main changes of
MA (as compared to GA) are as follows. First, a
new Local Optimization operator is introduced
after crossover. Local optimization applies the
local search algorithm as discussed above to each
solution in the generation, basically reducing the
population to a set of locally optimal solutions.
Second, we remove the mutation operator fromthe
algorithm. The main reason is that given that all
solutions in the population are already locally
optimal, randomly mutating one bit in a solution
can only lead to a worse solution.

Fig. 3 Memetic Algorithm
Furthermore, the main motivation for having
mutation in the first place was that it is the only
way of introducing new actions in the canonical
GA. This is no longer the case, because local
search can do the same thing. Generally, MA is
slower than GA, because more solutions are
evaluated in each generation. However, the
algorithm potentially converges against a very
good solution in a low number of generations.
Hence, we argue that in practice MA improves on
the canonical form most of the time for our
problem.


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
102
CONCLUSION
In this paper, we have modeled this problem as a
one-dimensional discrete optimization problem.
Furthermore, we have presented both,
deterministic and heuristic solution algorithms.
We have evaluated these algorithms based on a
manufacturing case study and have shown which
types of algorithms are better suited for which
scenarios. The main current limitation is that
adaptation is only considered on instance level,
which is, for each composition instance separately.
We believe that the PREVENT adaptation model
can be extended to this kind of SLOs and actions,
but new approaches to predict violations and
impact models are needed to this end. This need is
taken as the future work.
REFERENCES
[1] M.P. Papazoglou, P. Traverso, S. Dustdar, and
F. Leymann, Service-Oriented Computing:
State of the Art and Research Challenges,
Computer, vol. 40, no. 11, pp. 38-45, Nov.
2007.
[2] A. Lenk, M. Klems, J . Nimis, S. Tai, and T.
Sandholm, Whats Inside the Cloud? An
Architectural Map of the Cloud Landscape,
Proc. ICSE Workshop Software Eng.
Challenges of Cloud Computing (CLOUD
09) pp. 23-31, 2009.
[3] A. Dan, D. Davis, R. Kearney, A. Keller, R.
King, D. Kuebler, H. Ludwig, M. Polan, M.
Spreitzer, and A. Youssef, Web Services on
Demand: WSLA-Driven Automated
Management, IBM Systems J ., vol. 43, no. 1,
pp. 136-158, J an. 2004.
[4] L. Bodenstaff, A. Wombacher, M. Reichert,
and M.C. J aeger, Analyzing Impact Factors
on Composite Services, Proc. IEEE Intl
Conf. Services Computing (SCC 09), pp.
218-226, 2009.
[5] B. Wetzstein, P. Leitner, F. Rosenberg, S.
Dustdar, and F. Leymann, Identifying
Influential Factors of Business Process
Performance Using Dependency Analysis,
Enterprise Information Systems, vol. 4, no. 3,
pp. 1-8, July 2010.
[6] P. Leitner, B. Wetzstein, F. Rosenberg, A.
Michlmayr, S. Dustdar, and F. Leymann,
Runtime Prediction of Service Level
Agreement Violations for Composite
Services, Proc. Third Workshop Non-
Functional Properties and SLA Management
in Service-Oriented Computing (NFPSLAM-
SOC 09), pp. 176-186, 2009.
AUTHORS
Ulsa Suresh Babu has received
his B.Tech degree in Computer
Science and Engineering from
Brahmaiah College of
Engineering, Nellore affiliated to J NTU,
Anantapur in 2012 and pursuing M.Tech degree in
Computer Science & Engineering at Audisankara
College of Engineering & Technology, Gudur
affiliated to J NTU, Anantapur in (2012-2014).

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
103




Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
104
Use of Mobile Communication in Data-Intensive
Wireless Networks
Nuthanapati J yothsna
1
Reddy Sagar A C
2
Ravi Kumar G
3

1
PG Student, Siddartha Institute of Science and Technology, Puttur chittoor(d) A.P, jyothsna_oct30@yahoo.com
2
Assistant Professor, Siddartha Institute of Science and Technology, Puttur chittoor(d) A.P, sagar.akkaru@gmail.com
3
Assistant Professor, Siddartha Institute of Science and Technology, Puttur chittoor(d) A.P, ravikumar.gunduru@gmail.com
Abstract: Wireless Networks are increasingly
used in different types of data-intensive
applications scenarios such as micro-weather
monitoring, meticulousness agriculture, and
audio/video observation. The sensor nodes are
minute and limited in power. Sensor types vary
according to the application of WN. Whatever be
the application the key challenge is to broadcast
all the data generated within an applications life
span to the base station in the face of the fact that
sensor nodes have limited power supplies. The
concept of mobile communication is that the
mobile nodes change their locations so as to
minimize the total energy consumed by both
wireless transmission and locomotion. The
predictable methods, however, do not take into
account the energy level, and as a result they do
not always extend the network lifetime.
Keywords: Data-intensive; Energy;
communication; routing tree; WN

1. Introduction

Sensors have the capabilities of doing sensing,
data processing, and wirelessly transmitting
collected data back to base stations by way of
multiple-hop relay. Sensor itself supplies
necessary operation with limited battery energy.
Those operations that consume energy are
transmitting and receiving data, running
applications, measuring power, and even staying
in standby mode. Among others, data transmission
consumes the most energy. A wireless sensor
network (WSN) consists of spatially distributed
autonomous sensors to monitor physical or
environmental conditions,
such as temperature, sound , vibration , pressure
humidity, motion or pollutants and to
cooperatively pass their data through the network
to a main location[1,2]. The more modern
networks are bi-directional, also enabling control
of sensor activity. The development of wireless
sensor networks was
motivated by military applications such as
battlefield surveillance; today such networks are
used in many industrial and consumer
applications, such as industrial process monitoring
and control, machine health monitoring, and so
on. Figure 1 shows an example of a wireless
sensor network.


Figure 1: An example of Wireless Sensor Network
Recent advancement in mobile sensor platform
technology has been taken into attention that
mobile elements are utilized to improve the
WSNs performances such as coverage,
connectivity, reliability and energy efficiency.
The concept of mobile relay is that the mobile
nodes change their locations so as to minimize
the total energy consumed by both wireless
transmission and locomotion. The conventional
methods, however, do not take into account the
energy level, and as a result they do not always
prolong the network lifetime.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
105

2. Related Work

Analyzing the three different approaches: Mobile
base stations, data mules and mobile relays. All
the three approaches use mobility to reduce
energy consumption in wireless networks.

A. Mobile Base Station

A mobile base station is a sensor node collects the
data by moving around the network from the
nodes [4]. In some work, in order to balance the
transmission load, all nodes are performing
multiple hop transmissions to the base station. The
goal is to rotate the nodes which are close to the
base station. Before the nodes suffer buffer
overflows, the base station computes the mobility
path to collect data fromthe visited nodes. Several
rendezvous based data collection algorithms are
proposed, where the mobile base station only
visits a selected set of nodes referred to as
rendezvous points within a deadline and the
rendezvous points buffer the data fromsources.
High data traffic towards the base station is
always a threat to the networks life time [5]. The
battery life of the base station gets depleted very
quickly due to the sensor nodes which are located
near to the base station relay data for large part of
the network. The proposed solution includes the
mobility of the base station such that nodes
located near base station changes over time. All
the above approaches incur high latency due to the
low to moderate speed of mobile base stations.
Figure 2 shows Mobile base station


Figure 2: Mobile base station


B. Data Mules

Data mules are another form of base stations.
They gather data fromthe sensors and carry it to
the sink. The data mule collects the data by
visiting all the sources and then transmits it to the
static base station through the network. In order to
minimize the communication and mobility energy
Consumption the mobility paths are determined.
In paper [6] the author analyses an architecture
based on mobility to address the energy efficient
data collection problem in a sensor network. This
approach utilizes the mobile nodes as forwarding
agents. As a mobile node moves in close
propinquity to sensors, data is transmitted to the
mobile node for later dumps at the destination. In
the MULE architecture sensors transmit data only
over a short range that requires less transmission
power. However, latency is increased because a
sensor has to wait for a mule before its data can be
Delivered. Figure 3. The three tiers of the MULE
architecture. The Mule architecture has high
latency and this limits its applicability to real time
applications (although this can be mitigated by
collapsing the MULE and access point tiers). The
systemrequires sufficient mobility. For example,
mules may not arrive at a sensor or after picking
the data may not reach near an access-point to
deliver it. Also, data may be lost because of radio-
communication errors or mules crashing. To
improve data delivery, higher-level protocols need
to be incorporated in the MULE architecture. Data
mules also introduce large delays like base
stations since sensors have to wait for a mule to
pass by before initiating their transmission.


Figure 3: The three tiers of the MULE architecture

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
106


C. Mobile Relay

In this approach, the network consists of three
nodes such as mobile relay nodes along with static
base station and data sources. To reduce the
transmission cost relay nodes do not transport data
rather it will move to different locations. We use
the mobile relay approach in this work. In [7]
author showed that an iterative mobility algorithm
where each relay node moves to the midpoint of
its neighbors converges on the optimal solution
for a single routing path This paper presents
mobility control scheme for improving
communication performance in WSN. The
objectives of the paper are

1) Analyze when controlled mobility can improve
fundamental networking performance metrics
such as power efficiency and robustness of
communications
2) Provide initial design for such networks.
Mobile nodes move to midpoint of the neighbors
only when movement is beneficial.

Unlike mobile base stations and data mules, our
approach reduces the energy consumption of both
mobility and transmission. Our approach also
relocates each mobile relay only once
immediately after deployment. The paper study
the energy optimization problem that accounts for
energy costs associated with both communication
and physical node movement. Unlike previous
mobile relay schemes the proposed solution
consider all possible locations as possible target
locations for a mobile node instead of just the
midpoint of its neighbors.

3. Proposed work

We use low-cost disposable mobile relays to
reduce the total energy consumption of data-
intensive WSNs. Different from mobile base
station or data mules, mobile relays do not
transport data; instead, they move to different
locations and then remain stationary to forward
data along the paths from the sources to the base
station. Thus, the communication delays can be
significantly reduced compared with using mobile
sinks or data mules. Moreover, each mobile node
performs a single relocation unlike other
approaches which require repeated relocations.
Figure 4 shows Proposed Network.


Figure4: Proposed Network

The network consists of mobile relay nodes along
with static base station and data sources. Relay
nodes do not transport data; instead, they move to
different locations to decrease the transmission
costs. We use the mobile relay approach in this
work. Goldenberg et al. [13] showed that an
iterative mobility algorithmwhere each relay node
moves to the midpoint of its neighbors converges
on the optimal solution for a single routing path.
However, they do not account for the cost of
moving the relay nodes. In mobile nodes decide
to move only when moving is beneficial, but the
only position considered is the midpoint of
neighbors.

The sink is the point of contact for users of the
sensor network. Each time the sink receives a
question from a user, it first translates the question
into multiple queries and then disseminates the
queries to the corresponding mobile relay, which
process the queries based on their data and return
the query results to the sink. The sink unifies the
query results from multiple storage nodes into the
final answer and sends it back to the user.

The source nodes in our problem formulation
serve as storage points which cache the data
gathered by other nodes and periodically transmit
to the sink, in response to user queries. Such a
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
107
network architecture is consistent with the design
of storage centric sensor networks. Our problem
formulation also considers the initial positions of
nodes and the amount of data that needs to be
transmitted fromeach storage node to the sink.

We consider the sub problem of finding the
optimal positions of relay nodes for a routing tree
given that the topology is fixed. We assume the
topology is a directed tree in which the leaves are
sources and the root is the sink. We also assume
that separate messages cannot be compressed or
merged; that is, if two distinct messages of lengths
m1 and m2 use the same link (s
i
, s
j
) on the path
froma source to a sink, the total number of bits
that must traverse link (s
i
, s
j
) is m
1
+m
2
.

a) Energy Optimization Framework

In this section, we formulate the problem of
Optimal Mobile Relay Configuration (OMRC) in
data-intensive WSNs. Unlike mobile base stations
and data mules, our OMRC problemconsiders the
energy consumption of both mobility and
transmission. The Optimal Mobile Relay
Configuration (OMRC) problem is challenging
because of the dependence of the solution on
multiple factors such as the routing tree topology
and the amount of data transferred through each
link. For example, when transferring little data,
the optimal configuration is to use only some
relay nodes at their original positions.

Assume the network consists of one source s
i
1,
one mobile relay node s
i
and one sink s
i
+1. Let
the original position of a node sj be o
j
=(p
j
, q
j
),
and let u
j
= (x
j
, y
j
) its final position in
configuration U. According to our energy models,
the total transmission and movement energy cost
incurred by the mobile relay node s
i
is
c
i
(U) =ku
i
o
i
+am+bu
i
+1 u
i

2
m

Now We need to compute a position ui for si that
minimizes C
i
(U) assuming that u
i
1 =o
i
1 and
u
i
+1 =o
i
+1; that is, node s
i
s neighbors remain at
the same positions in the final configuration U.
We calculate position u
i
=(x
i
, y
i
) for node s
i
by
finding the values for xi and y
i
where the partial
derivatives of the cost function C
i
(U) with respect
to x
i
and y
i
become zero. Position U
i
will be
toward the midpoint of positions u
i
1 and u
i
+1.
The partial derivatives at x
i
and y
i
, respectively
are defined as follows:
C
i
(U)
---------- =2bm(x
i
+1 x
i
) +2bm(x
i
x
i
1)
x
i

(x
i
p
i
)
+k --------------------------------------
(x
i
- p
i
)
2
+(y
i
- q
i
)2

C
i
(U)
---------- =2bm(y
i
+1 y
i
) +2bm(yi yi1)
yi
(yi qi)
+k --------------------------------------
(x
i
- pi)2 +(yi - qi)2

and rms do not have to be defined. Do not use
abbreviations in the title or heads unless they are
unavoidable.
b) Tree Optimization Algorithm
In this section, we consider the sub problem of
finding the optimal positions of relay nodes for a
routing tree given that the topology is fixed. We
assume the topology is a directed tree in which the
leaves are sources and the root is the sink. We
also assume that separate messages cannot be
compressed or merged; that is, if two distinct
messages of lengths m1 and m2 use the same link
(s
i
, s
j
) on the path from a source to a sink, the
total number of bits that must traverse link(s
i
, s
j
)
is m
1
+m
2
. Let the network consists of multiple
sources, one relay node and one sink such that
data is transmitted from each source to the relay
node and then to the sink. We modify our solution
as follows. Let s
i
be the mobile relay node, S(s
i
)
the set of source nodes transmitting to s
i
and sd
i

the sink collecting nodes from s
i
. The cost
incurred by s
i
in this configuration U is:

c
i
(U) =ku
i
o
i
+ am
i
+bm
i
u
i
+1 u
i

2


Algorithm1
procedure OPTIMALPOSITIONS(U
0
)
converged false;
j 0;
repeat
anymove false;
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
108
j j +1;
Start an even iteration followed by an odd
iteration
for idx =2 to 3 do
for i =idx to n by 2 do
(u
j
i
,moved) LOCALPOS(o
i
, S(s
i
), sd
i
);
anymove anymove OR moved
end for
end for
converged NOT anymove
until converged
end procedure

Our algorithm starts by an odd/even labeling step
followed by a weighting step. To obtain consistent
labels for nodes, we start the labeling process
fromthe root using a breadth first traversal of the
tree. The root gets labeled as even. Each of its
children gets labeled as odd. Each subsequent
child is then given the opposite label of its parent.
We define mi, the weight of a node s
i
, to be the
sum of message lengths over all paths passing
through s
i
. This computation starts from the
sources or leaves of our routing tree. Initially, we
know mi =Mi for each source leaf node s
i
. For
each intermediate node s
i
, we compute its weight
as the sumof the weights of its children. Once
each node gets a weight and a label, we start our
iterative scheme. In odd iterations j, the algorithm
computes a position u
j

i
for each odd-labeled node
s
i
that minimizes C
i
(U
j
) assuming that
u
j

i
1 =u
j1
i
1 and u
j

i+1
=u
j1
i+1
that is, node
s
i
s even numbered neighboring nodes remain in
place in configuration U
j
. In even-numbered
iterations, the controller does the same for even-
labeled nodes. The algorithm behaves this way
because the optimization of u
j

i
requires a fixed
location for the child nodes and the parent of s
i
.
By alternating between optimizing for odd and
even labeled nodes, the algorithm guarantees that
the node s
i
is always making progress towards the
optimal position ui. Our iterative algorithm is
shown in algorithm1.

4. Conclusion
The main objective of this paper is energy
conservation which is holistic in that the total
energy consumed by both mobility of relays and
wireless transmissions is minimized, which is in
contrast to existing mobility approaches that only
minimize the transmission energy consumption.
The tradeoff in energy consumption between
mobility and transmission is exploited by
configuring the positions of mobile relays. We
develop two algorithms that iteratively refine the
configuration of mobile relays. The first improves
the tree topology by adding new nodes. It is not
guaranteed to find the optimal topology. The
second improves the routing tree by relocating
nodes without changing the tree topology. It
converges to the optimal node positions for the
given topology. Our algorithms have efficient
distributed implementations that require only
limited, localized synchronization.

References
[1] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam,
and E.Cayirci, A Survey on Sensor Networks,
IEEE Comm.Magazine, vol. 40, no. 8, pp. 102-
114, Aug. 2002.
[2] Shio Kumar Singh, M.P. Singh, and D.K.
Singh,Applications, Classifications, and
Selections of Routing Protocols for Wireless
Sensor Networks International J ournal of
Advanced Engineering Sciences and Technologies
(IJ AEST), November 2010, vol. 1, issue no. 2, pp.
85-95.
[3] Ganesan, B. Greenstein, D. Perelyubskiy, D.
Estrin, and J .Heidemann, An Evaluation of
Multi-Resolution Storage for Sensor Networks,
Proc. First Intl Conf. Embedded Networked
Sensor Systems (SenSys), 2003.
[4] Fatme El-Moukaddem, Eric Torng, Guoliang
Xing" Mobile Relay Configuration in Data-
intensive Wireless Sensor Networks" in IEEE
Transactions on Mobile computing, 2013.
[5] J . Luo and J .-P. Hubaux, J oint Mobility and
Routing for Lifetime Elongation in Wireless
Sensor Networks, in INFOCOM, 2005.
[6] S. J ain, R. Shah, W. Brunette, G. Borriello,
and S. Roy,Exploiting Mobility for energy Cient
Data Collection in Wireless Sensor Networks,
MONET,vol. 11, pp. 327339, 2006.
[7] K. Goldenberg, J . Lin, and A. S. Morse,
Towards Mobility as a Network Control
Primitive, in MobiHoc, 2004, pp. 163174.
[8] Tang and P. K. McKinley, Energy
Optimization Under Informed Mobility, IEEE
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
109
Trans. Parallel Distrib. Syst., vol. 17,pp. 947962,
2006.
[9] Fatme El-Moukaddem, Eric Torng, and
Guoliang Xing, Member, IEEE "Mobile Relay
Configuration in Data- Intensive Wireless Sensor
Networks".
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
110
Confidential Multiparty Computation with
Anonymous ID Assignment using Central Authority
P Babitha
1
P Yogendra Prasad
2
A C Reddy Sagar
3
Assist Professor Assist Professor Assist Professor
SISTK SISTK SISTK


ABSTRACT- For giving out private data securely
between several parties an algorithm has been
used. An anonymous ID assignment technique is
used iteratively to assign the nodes with ID
numbers ranging from 1 to N. This technique
enhances data that are more complex to be
shared securely. The nodes are assigned with the
anonymous ID with the help of a central
authority. The algorithm has been compared
with the existing algorithm. In this paper we
propose an algorithm has been developed based
on Sturms theorem and Newtons identities.
The numbers of iterations are found out with the
help of Markov chain.
KEYWORDS: Anonymous Id, AIDA-Anonymous Id
Assignment.
1. INTRODUCTION
The anonymous message plays a very significant
role in internets popularity for both individual
and business purposes. Cloud based website
management tools enable the servers to analyze
the users behavior. The disadvantages of giving
out private data are being studied in detail.
Other applications for anonymity are various viz
Doctor medical records, social networking,
electronic voting and many more. In secure
multiparty computation which is a new form of
anonymity allows several multiple parties to
share data that remains unknown? A secure
computation function enables multi parties to
compute the sum of their inputs rather than
revealing the data. This method is very much
well-liked in data mining operations and enables
classifying the complexities of secure multiparty
computation.
Our main algorithm is built on top of a method
that shares simple data anonymously and yields
a technique that enables sharing of complex
data anonymously. With the help of permutation
methods the assigned ID are known only to the
nodes which are being assigned IDs. There are
several applications where network nodes needs
self-motivated unique IDs. One such application
is grid computing where the services are
requested without disclosing the identities of
the service requestor. To differentiate between
anonymous communication and anonymous ID
assignment, think about a situation where N
parties wish to exhibit their data in total, in N
slots on a third party site, anonymous ID
assignment method assigns N slots to the users
whereas anonymous communication allows the
users to conceal their identities In our network
the identities of the parties are known but not
the true identity. In this project we use an
algorithm for sharing simple integer data which
is based on secure sum. This algorithm is used in
every iteration of anonymous ID assignment.
Here we consider all the nodes to be half direct.
Even though they follow a set of rules for
message if they happen to see information they
might intrude.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
111
In existing system, the information about each
node will be shared along with the data. It is
usually encrypted along with the data. The usage
of Newtons polynomial cannot be avoided as it
increases the number of rounds of iterations
that are used to compute the secure sum and
power sum. Hence the performance of the
system also decreases. It only focuses on the
sum inputs whereas our project deals with the
number of rounds. The usage of secure
multiparty computation is being avoided with
the usage of Sturms theorem to make sure that
the information about the nodes are not
revealed. In the current system the main goal is
to provide anonymous id for each node. Each
node will have a secure communication of
simple and complex data. Those datas may be
from static data or dynamic data. By
implementing secure sum hides permutations
method and anonymous id assignment (AIDA)
method the permutation methods are kept
anonymous to each other. Hence here encoding
technology is used to create anonymous ID and
the ID is being assigned to the user by the
central authority and the receiver receives the
data and decodes it with the key that is known
only to the sender and the receiver which might
not be known to the other semi honest node
that might intrude.
II MODULES DESCRIPTIONS
AUTHENTICATION
The process of identifying an individual usually
based on a username and password. In security
systems, Authentication merely ensures that the
individual is who he or she claims to be, but says
nothing about the access rights of the individual.
1 LOGIN:
In User and Admin login we are going to check
whether the system is trusted machine or
distrust machine. If the machine is trusted then
the user or admin is allowed with n attempts. If
the machine is distrusted machine then the user
is allowed with single attempt. Process Involved
is to Check the login name and password Then
allows the authorized user to use these pages. If
the unauthorized user attempts to access user
login then restrict that user and give the
information.
A FORGET PASSWORD:
When the users forget their password then the
user can access this forget password. It is used to
create a new password. To ensure that user
accessing forget password is a legitimate user,
the user will be asked a question. These
questions and their answers are created, while
the user is registering to the site. If the user
enters the answer then the entered text will be
matched with the database. If the result is true,
then the user will be allowed to enter the new
password to access the site. If the result is false,
user will not be allowed to enter the new
password to access the site.
B REGISTRATION:
When a new user is creating an account he
needs to register here by giving the sufficient
information. Registration might also contain
some private data that will be kept confidential
so that the information about username and
password is retrieved when it is forgotten.
2 ADMIN:
In this module when the admin attempts to login
we need to find whether the machine is trusted
or it is distrusted machine. It is found by user id
and pass word. Admin can provide AID to all
nodes. With the help of that AID each node can
share the data in internet. Admin can generate
that AID for individual nodes. So the sharable
data can be kept in a sharable database. As a
result their own private data will be maintained
secret.
A GENERATES AID:
In this module admin wants to create the AID for
individual nodes. Admin can make this unique
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
112
AID for each user presented in network. With
the help of this AID user can share his data and
also he can keep his own private data as secret.
B ASSIGN AID:
Admin can provide unique AID to all nodes.
Nodes presented in network will be
communicating by using AID. AID can not
contain any private information.AID helps to
keep personal information as more secret.
3. USERS
Users can login by entering the given username
and password. Then he/she may go for
corresponding page. User can keep his own
information in a sharable data base. And also he
can retrieve data shared by other users. User has
to use his AID for sharing information.
A SHARE DATA:
New user has to get AID from admin. Admin will
assign that AID for every node. So data shared by
user can be kept in a sharable database and it
can be shared by all users. Each node will have
unique AID with the help of that unique AID any
user can store and retrieve from sharable
database.
B RETRIEVE SHARED DATA:
In this module user can retrieve the shared data.
Shared data may be stored by him or by any
other node. So it will be easy to make own
private data as secret by implementing
anonymous id algorithm.
A SHARE DATA:
New user has to get AID from admin. Admin will
assign that AID for every node. So data shared by
user can be kept in a sharable database and it
can be shared by all users. Each node will have
unique AID with the help of that unique AID any
user can store and retrieve from sharable
database.
B RETRIEVE SHARED DATA:
In this module user can retrieve the shared data.
Shared data may be stored by him or by any
other node. So it will be easy to make own
private data as secret by implementing
anonymous id algorithm.
Secure sum Algorithm
Given nodes each holding an data item from a
finitely represent able abelian group, share the
value among the nodes without revealing the
values .
1) Each node , chooses random values such that
ri,1+ri,N =di
2) Each random value is transmitted from
node to node . The sum of all these random
numbers is, ofcourse, the desired total .

3) Each node totals all the random values
received as:

si,j =r1,j++rN,j
4) Now each node simply broadcasts to all other
nodes so that each node can compute:

T =s1 +sN
AIDA Algorithm Implementation:
Given nodes n1,,nN, use distributed
computation (without central authority) to find
an anonymous indexing permutation s : {1,,N}
{1,,N}.
1) Set the number of assigned nodes A =0.
2) Each unassigned node chooses a random
number in the range 1 to S. A node assigned in a
previous round chooses r1 =0.
3) The random numbers are shared
anonymously. Denote the shared values by
q1,,qN.
4) Let q1,,qk denote a revised list of shared
values with duplicated and zero values entirely
removed where k is the number of unique
random values. The nodes ni which drew unique
random numbers then determine their index si
from the position of their random number in the
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
113
revised list as it would appear after being sorted:
si =A +Card{qj : qj <=ri}
5) Update the number of nodes assigned: .
6) If A <N then return to step (2).
Basic formula
P (collision) =1-e^(-N^2/(2*H))=alpha
N=sqrt (2*ln (1/1-alpha))*sqrt (H)
N: number of evenly distributed hashes to
compare
H: the Size of the element count of all possible
hashes
H=2^(IDlength *ln (36)/ln (2))
Hashes calculated with SHAI (IDlength<=30).
K.Communications Requirements of AIDA
Methods
Consider the required number of data bits for
each of the three variant methods just
described. This is the number of data bits that
would be transmitted in each packet by the
secure sum algorithm introduced earlier. The
required numbers of data bits B are slightly
overestimated by the formulae: Bprimc =N.
[log2(P+1)] Bprimc =N. [log2(N)] /2.[log2(S)]
Bprimc =S. [log2(N+1)] (2) The computational
requirements of the slot selection appear, at
first, to be trivial. However, for every root that
the prime modulus method must check, slot
selection. L. The Completion Rate after R
Rounds Two nodes might make identical choices
of random numbers, or slots as they will be
termed in this section. One can only guarantee
that a complete assignment of nodes using
possibilities for slots or random number choices
and rounds will occur with at least a desired
probability. The formulae are derived by
assuming that N -1 node have chosen slots and
looking at the next choice. The Nth node into
choose a slot resulting in assignments and
conflicts. The slot it chooses could be
unassigned, already in conflict with multiple
occupants, or already assigned with exactly one
occupant.
III. EXPERIMENTAL RESULTS
SI.NO N H P
1 189 45 390
2 210 55 400
3 290 95 442
4 549 225 667
5 999 455 1109
Table 1: Anonymous ID creation Algorithm
Results are given in terms of identities. N
represents the number of evenly distributed
hashes to compare. H indicates the size of the
element count of all possible hashes. P Indicates
the value of anonymous ID. This is unique to
every members of the group. Anonymous ID
received is unknown to other members of the
group. Anonymous id is examined between
communication and computational requirement.
IV. CONCLUSION
Proposed paper greatly decreases
communication overhead. By using private
communication channel that is anonymity
router to transmit the data more securely.
To overcome the problem of identifying
details and changing information
anonymous id was utilized. Random serial
number is used to identify whether the data
requesting person is a correct authorized
person or hackers. The use of the Newton
identities greatly decreases communication
overhead. This can enable the use of a larger
number of slots with a consequent
reduction in the number of rounds required.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
114
The solution of a polynomial can be avoided
at some expense by using Sturms theorem.
The development of a result similar to the
Sturms method over a finite field is an
enticing possibility. With private
communication channels, the algorithms are
secure in an information theoretic sense.
Apparently, this property is very fragile. The
very similar problem of mental poker was
shown to have no such solution with two
players and three cards. The argument can
easily be extended to, e.g., two sets each of
N colluding players with a deck of 2N+1
cards rather than our deck of 2N cards. In
contrast to bounds on completion time
developed in previous works, our formulae
give the expected completion time exactly.
All of the no cryptographic algorithms have
been extensively simulated, and the present
work does offer a basis upon which
implementations can be constructed. The
communications requirements of the
algorithms depend heavily on the underlying
implementation of the chosen secure sum
algorithm. In some cases, merging the two
layers could result in reduced overhead.
REFERENCES
[1] SarbanesOxley Act of 2002, Title 29,
Code of Federal Regulations,Part 1980,
2003.
[2] Y.Zhang, W.Liu, and W.Lou .Anonymous
communications in mobile ad hoc networks.
In INFOCOM 2005, 24th annual joint
conference the IEEE Computer Societies.
Proceeding IEEE, volume 3, pages 1940
1951, March 2005.
[3] D. Chaum, Untraceable electronic mail,
return address and digital pseudonyms,
Commun. ACM, vol. 24, no. 2, pp. 8488,
Feb.1981.
[4] C. Clifton, M. Kantarcioglu, J. Vaidya, X.
Lin, and M. Y. Zhu, Tools for privacy
preserving distributed data mining, ACM
SIGKDD Explorations Newsletter, vol. 4, no.
2, pp. 2834, Dec. 2002.
[5] J. Wang, T. Fukasama, S. Urabe, and
T.Takata, A collusion-resistant approach to
privacy-preserving distributed data mining,
IEICE Trans.Inf.Syst. (Inst. Electron. Inf.
Commun. Eng.), vol. E89-D, no. 11, pp.2739
2747, 2006.
[6] F. Baiardi, A. Falleni, R. Granchi, F.
Martinelli, M. Petrocchi, and A.Vaccarelli,
Seas, a secure e-voting protocol: Design
and implementation, Comput. Security, vol.
24, no. 8, pp. 642652, Nov. 2005.
[7] Sanil, A. P., Karr, A. F., Lin, X., and Reiter,
J. P. (2007). Privacy preserving analysis of
vertically partitioned data using secure
matrix products.
[8] J. Kong and X. Hong. Anodr: anonymous
on demand routing with untraceable routes
for mobile ad-hoc networks. InMobihoc03:
Proceedings of the 4th ACM international
symposium on Mobile ad hoc networking
ages 291302, New York, NY, USA,2003.
[9] J. Yoon and H. Kim, A new collision-free
pseudonym scheme in mobile ad hoc
networks,5th workshop on Resource
allocation, Cooperation and competition in
wireless networks. June 2009.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
115
[10] A. Friedman, R. Wolff, and A. Schuster,
Providing k-anonymity in data mining,
VLDB Journal, vol. 17, no. 4, pp. 789804,
Jul. 2008.
[11] J. Domingo-Ferrer, A new privacy
homomorphism and applications,
Information Processing Letters, vol. 60, no.
5, pp. 277282, December1996.[Online].
Available: citeseer.nj.nec.com/290190.html
[12] J. Domingo-Ferrer, A provably secure
additive and multiplicative privacy homo
morphishm, in Information Security, ser.
Lecture Notes in Computer Science, A. Chan
and V. Gligor, Eds., vol. 2433. Springer
Verlag, 2002, pp. pp. 471483.
[13] D. M. Goldschlag,M. G. Reed, and P. F.
Syverson, Hiding routing information,
inProc. Information Hiding, 1996, pp.
137150,Springer Verlag
[14] A. Karr, Secure statistical analysis of
distributed databases, emphasizing what we
dont know, J. Privacy Confidentiality, vol.
1, no.2,pp. 197211, 2009.
[15] Q. Xie and U. Hengartner, Privacy-
preserving matchmaking for mobile social
networking secure against malicious users,
in Proc. 9thAnn. IEEE Conf. Privacy, Security
and Trust, Jul. 2011, pp. 252259.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
116
DESIGN OF RF BASED VOICE-CONTROLLED MULTI-TERRAIN ROBOT TO
TRAVEL ON UNEVEN SURFACES

C.Venkatesh T.Venkata Subbaiah
Assistant Professor, Department of ECE, AITS, Department of ECE, AITS,
Rajampet, Andhra Pradesh, India Rajampet, Andhra Pradesh, India
E-mail: venky.cc@gmail.com E-mail: t.v.subbaiah1994@gmail.com

G.Sarath Kumar
Department of ECE, AITS,
Rajampet, Andhra Pradesh, India
E-mail: sarath.g66a@gmail.com


Abstract Ambulation in rough and unstructured terrain
is naturally very difficult for small animals or robots. The
ability to climb is found widely among small animals
.Mobile robots are more and more leaving the protected
lab environment and entering the unstructured and
complex outside world, e.g. for applications such as
environmental monitoring. A substantial portion of the
Earth is inaccessible to any sort of wheeled mechanism
natural obstacles like large rocks, loose soil, deep ravines,
and steep slopes conspire to render rolling locomotion
ineffective. Hills, mountains, shores, seabeds, as well as the
moon and other planets present similar terrain challenges.
In many of these natural terrains, the terrain robot is well-
suited to travel in uneven surfaces In robotics, various
climbing systems have been presented so far using
different approaches for climbing but only very few are
capable of up righting themselves after landing. Inspired
by the terrain climbing capability, a small terrain robot
has been developed. In this paper the working and design
of terrain robot is presented. This robot utilizes four
wheels to drive over just about any terrain. It works on
any indoor surface and outdoor surfaces. This is an
intelligent robot that can easily move on uneven surfaces
too. This robot uses RF technology controlled by RF
remote. This can be moved forward and reverse direction
using geared motors of 60RPM. Also this robot can take
sharp turnings towards left and right directions.


Keywords Ambulation, Seabeds, Terrain, Locomotion.

I. INTRODUCTION
In Recent technology advances mechanical, electronics,
communication, and networking make it possible to build
small robots and organize many of them to forma mobile
sensor network. This mobile sensor network has various
applications such as search and rescue, surveillance, and
environmental monitoring. The small terrain robot are
developed for such applications. Compared with the
traditional wheeled sensors. This ability makes jumping an
ideal locomotion method in the areas of rugged terrain and
natural obstacles. To use jumping as a locomotion method, the
robot must be able to jump continuously; therefore, the
following two capabilities are needed: (1)The robot can jump
with a certain takeoff angle; (2) After landing on the ground,
the robot can self-right fromany possible landing posture.
Besides these two capabilities, it would be beneficial that the
robot can also change the jumping direction or the takeoff
angle.In nature, millimeter-sized insects often address both
obstacles and efficiency through jumping. While jumping is
often seen as an energetically costly escape mechanism, has
shown that as insect size shrinks, jumping becomes more
advantageous due to the higher takeoff velocities allowed.
Because small jumpers are more mechanically efficient than
their larger counterparts, they require less muscle tissue to
make themmore energy efficient as well.
This paper presents the design and working of terrain
robot with experimental results
II. DESIGN STUDY
The block diagram of the robotic systemconsists of
Transmitter and Receiver sections. The Transmitter section
consists of PC with commands, RF Module and the Receiver
section consists of RF module, ARM processor (LPC2148),
H- Bridge and DC Motor.
A. Transmitter:
The Transmitter section having the four switches
placed around the neck and RF Module. Initially the switches
are at logic1. When the switch is pressed the concerned switch
level goes to logic0.The switches are the inputs to RF
transmitter through the RF encoder is shown in fig.1
Fig. 1. Transmitter Section
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
117
B. Receiver
The received signal from the transmitter is fed to the
RF decoder (Serial input and parallel output) .The output of
the decoder is given to H-Bridge through the ARM processer
shown in fig2.The output of H-Bridge drives the DC motors.


Fig. 2. Receiver section
III. RF TECHNOLOGY
Radio frequency [RF] has a frequency range about 3Hz to
300GHz. This range corresponds to frequency of alternating
current electrical signals used to produce and detect radio
waves. Since most of this range is beyond the vibration
rate that most mechanical systems can respond to RF
usually refers to oscillations in electrical circuits or
electromagnetic radiation. When an RF current is supplied to
an antenna, it gives rise to an electromagnetic field that
propagates through space. Electrical currents that oscillate at
RF have special properties not shared by direct current signals.
One such property is the ease with which it can ionize air
creates a conductive path through air. Another property is the
ability to appear to flow through paths that contain insulating
material, like the dielectric insulator of a capacitor. The
degree of effect of this property depends on the frequency of
the signals. RF is a radio frequency technology which uses
frequencies in the range of 3MHZ to 300 MHZ in general.
Here in this RF system, we are using the frequency of
433MHZ which is in the Frequency range. The distance of
this radio frequency range is up to 100min general. In this
project, the distance is up to 100m.
A. RF TRANSMITTER:
The STT--433 is ideal for remote control applications as
shown in Fig.1 where low cost and longer range is required.
The transmitter operates froma 1.5-12v supply, making it
ideal for battery powered applications. The transmitter
employs a SAW-stabilized oscillator, ensuring accurate
frequency control for best range performance. The
manufacturing friendly SIP style package and low cost
make the STT-433 suitable for high volume applications.


Fig. 3. RF Transmitter STT-433MHZ

Features: 433.92Hz Frequency, Low cost, 1.5-12V operation,
small size.
PIN OUT:
GND: Transmitter ground. Connect to ground plane.
DATE: Digital date input. This input is CMOS compatible
and should be driven with CMOS level inputs.
VCC: Operating voltage for the transmitter.VCC should be
bypassed with a .01uF ceramic capacitor and filtered with
a 4.7uF tantalum capacitor . Noise on the power supply will
degrade transmitter noise performance.
ANT: 50ohmantenna output. The antenna port impedance
affects output power and harmonic emissions. Antenna can
be single core wire of approximately 17cm length or
PCB trace antenna.

B. RF RECEIVER:
The data is received by the RF receiver from the
antenna pin and this data is available on the data pins.
Two data pins are provided in the receiver module. Thus data
can be used for further application.
This output is capable of driving one TTL or
CMOS load .It is a CMOS compatible output. The data
transmitted in to the air is received by the receiver. The
received data is taken from the data line of the receiver
and is fed to the decoder.


Fig. 4. RF Receiver STT-433MHZ.
Features:
* Low current (max.100ma)
* Low voltage (max.65v).
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
118

C. ENCODER AND DECODER:
The encoder and decoder used here is HT12E and HT12D
respectively from HOLTEK SEMICONDUCTORS INC. The
HT12E encoder ICs are series of CMOS LSIs for remote
control systemapplications. They are capable of encoding 12
bit of information which consists of N address bits and 12-N
data bits.



Fig. 5. RF Encoder Connections
Encoder IC (HT12E) receives parallel data in the form of
address bits and control bits. The control signals from remote
switches along with 8 address bits constitute a set of 12
parallel signals. The encoder HT12E encodes these parallel
signals into serial bits. Transmission is enabled by providing
ground to pin14 which is active low. The control signals are
given at pins 10-13 of HT12E.


Fig. 6. RF reciver connections
Features:
* Operating voltage 2.4V~12V.
* Low power and high noise immunity CMOS
Technology.
* Low standby current.
IV. JUMPING MECHANISM:
Traditionally, most mobile robots have been equipped with
wheels. The wheel is easy to control and direct. It provides a
stable base on which a robot can maneuver and is easy to
build. One of the major drawbacks of the wheel, however, is
the limitation it imposes on the terrain that can be successfully
navigated. A wheel requires a relatively flat surface on which
to operate. Rocky or hilly terrain, which might be found in
many applications as forestry, waste clean-up and planetary
exploration, imposes high demands on a robot and precludes
the use of wheels. A second approach to this problem would
be to use tracked wheel robots. For many applications this is
acceptable, especially in very controlled environments.
However, in other instances the environment cannot be
controlled or predicted and a robot must be able to adapt to its
surroundings.
V. TERRAIN ROBOT HARDWARE TOOLS
Robotic colonizer main board is the brain of mechanical
robot which receives the commands fromPC through ZigBee
wireless connection and processes these commands to perform
pattern motion control. The main part in the robot main board
is ATmega168 microcontroller which generates two PWM
signals for each wheel of DC motor. The two active wheels of
the robot are actuated by two independent servo motors
modified for continuous rotation. In particular, The robot is
powered by 12 V battery. The hardware tools of the robotic
system are: ZigBee, ATmega 168 Microcontroller, L293D
Driver Circuit, DC Motors, PH Meter, Metal Detector, GPS
The hardware components of the robotic systemis as shown in
Fig. 4.
A) ARM Processor
ARM7TDMI is a core processor module embedded in many
ARM 7 microprocessors including LPC2148. The
ARM7TDMI core is a 32-bit embedded RISC processor. Its
simplicity results in a high instruction throughput and
impressive real-time interrupt response from a small and cost-
effective processor core.
Founded in November 1990, it is spun out of a corn
computers, it designs the ARM range of RISC processor cores.
Licenses ARM core designs to semiconductor partners who
fabricate and sell to their customers. ARM dose not fabricate
and sell to their customers it also develop technologies to
assist with the design in of the ARM architecture. Software
tools, boards, debug hardware application software, bus
architectures, peripherals etc. The ARM processor core
originates within a British computer company called acorn.
Instruction set. Essentially, the ARM 7TDMI-S processor has
two instruction sets:
*The standard 32 bit ARM set.
*A16-bit thumb set.
ARM stands for advanced RISC machines. It is a 32 bit
processor core used for high end applications. The LPC 2148
micro controller are based on a 16 bit/32-bit ARM7 TDMI-S
CPU with real time emulation and embedded trace support
that combined the micro controller with embedded high speed
flash memory running from32 KB to 512 KB.
ARM (Advance RISC machine)
T- The thumb 16 bit instruction set.
D-ON chip debug support.
M- Embedded multiplier.
I- Embedded ICE hard ware.
S-Synthesizable.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
119
The ARM has seven basic operating modes
User: unprivileged mode under which most tasks run
FIQ: entered when a high priority (fast) interrupt is raised.
IRQ: entered when a low priority (normal) interrupt is raised.
Supervisor: entered on reset and when a software
interrupt instruction is executed.
Abort: used to handle memory access violations.
Undef: used to handle undefined instructions.
System: privileged mode using the same registers as user
mode.
The ARM architecture provides a total of 37 register ,
all of which are 32 bits long . however these are
arranged in to several banks, with the accessible bank
being governed by the current processor mode. In each
mode, the core can access; A particular set of 13 general
purpose registers (r0 r12),A particular r13 -which is
typically used as a stack pointer, r14 is used as a link
register for branching.

B) LPC2148 CONTROLLER:
The LPC2141/42/44/46/48 microcontrollers are based on
a 16-bit/32-bit ARM 7TDMI-CPU With realtime emulation
and embedded trace support, that combine microcontroller and
embedded high speed flash memory ranging from32KB to
512KB. Serial communications interfaces ranging froma USB
2.0 fullspeed device, Multiple UARTs, SPI,SSP to 12C- Bus
and on-chip SRAM of 8KB up to 40KB, make these devices
very well suited for communication gate ways and protocol
converters, soft modems, voice recognition and low end
imaging, providing both large buffer size and high processing
power. Various 32 bit timers, or dual 10- bit ADC, 10-bit
DAC, PWM channels and 45fast GPIO lines with up to nine
edge or level sensitive external interrupt pins make these
microcontrollers suitable for industrial control and medical
systems.

C) L293D Driver Circuit:
Motor driver is basically a current amplifier which takes a
low-current signal fromthe microcontroller and gives out a
proportionally higher current signal which can control and
drive a motor. In most cases, a transistor can act as a switch
and perform this task which drives the motor in a single
direction[4].
D) DC Motors:
DC motors are configured in many types and sizes,
including brush less, servo, and gear motor types. A motor
consists of a rotor and a permanent magnetic field stator. The
magnetic field is maintained using either permanent magnets
or electromagnetic windings. DC motors are most commonly
used in variable speed and torque.
Motion and controls cover a wide range of components that in
some way are used to generate and/or control motion.
E) Ultrasonic Sensor Motors:
An ultrasonic transducer is a device that converts energy
into ultrasound, or sound waves above the normal range of
human hearing. While technically a dog whistle is an
ultrasonic transducer that converts mechanical energy in the
form of air pressure into ultrasonic sound waves, the term is
more apt to be used to refer to piezoelectric
transducers or capacitive transducers that convert electrical
energy into sound. Piezoelectric crystals have the property of
changing size when a voltage is applied; applying
an alternating current (AC) across them causes them to
oscillate at very high frequencies, thus producing very high
frequency sound waves. The location at which a transducer
focuses the sound can be determined by the active transducer
area and shape, the ultrasound frequency, and the sound
velocity of the propagation medium. Some systems use
separate transmitter and receiver components while others
combine both in a single piezoelectric transceiver.Non-
piezoelectric principles are also used in construction of
ultrasound transmitters. Magnetostrictive materials slightly
change size when exposed to a magnetic field; such materials
can be used to make transducers. A capacitor microphone uses
a thin plate which moves in response to ultrasound waves;
changes in the electric field around the plate convert sound
signals to electric currents, which can be amplified.
VI. EXPERIMENTAL RESULTS
The autonomous manoeuvre sailing robot
for oceanographic research is used to explore all the details on
the surface of the water. This robot is used for locating the
position of the systemusing GPS & GSM, detects metals
present in the ocean, and measures the depth and boundaries,
used for surveillance and rescue operation.



Fig. 7. Robot is moving aty sandy surface


Fig. 8. Robot is moving in sandy and rocky surface
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
120

Fig. 9. Robot is moving in rocky surface

VII. CONCLUSION
In this paper, we introduce a successful working prototype
model of terrain robot is designed to travel in uneven surfaces.
An autonomous terrain robot offers major advantages
compared to other vehicles. In this paper we also discuss the
development process of a robotic system for environmental
monitoring in search &rescue and demining applications.
Further development is required to demonstrate the feasibility
of a terrain robot for long termuse in open area and helpful
for the scientists.

REFERENCES
[1] Alina Conduraru , Ionel Conduraru, Emanuel Puscalau, Geert De
Cubber, Daniela Doroftei, Haris Balta Development of an autonomous
rough-terrain robot
[2] G. De Cubber and D. Doroftei and H. Sahli and Y. Baudoin, Outdoor
Terrain Traversability Analysis for Robot Navigation using a Time-Of
Flight Camera, RGB-D Workshop on 3D Perception in Robotics, 2011
[3] D. Doroftei and G. De Cubber and K. Chintanami, Towards
collaborative human and robotic rescue workers, Human Friendly
Robotics, 2012.
[4] K. Richardson, Rescue robots - where were they in the J apanese quake
relief efforts?, Engineering and Technology Magazine, vol.6, nr. 4, 2011
[5] M.P. Khorgade Application of MEMS in Robotics and Bio MEMS,
Proceedings of the UK sim13
th
international conference.
[6] Mobile Robotic Navigation and control for large-scale wireless sensor
network repair, by Kyle lathy in north Carolina state university on
may6, 2009
[7] S. Mehta, ET. Al. CMOS Dual -Band Tri-Mode chipset for IEEE
802.11a/b/g wireless LAN, IEEE RF IC Symposium, pp 427-
430,2003
[8] B. Razavi, RF Transmitter Architectures and circuits, IEEE CICC,
pp.197-204, 1999.
[9] M.P. Khorgade Application of MEMS in Robotics and Bio MEMS,
Proceedings of the UK sim13
th
international conference.
[10] Mobile Robotic Navigation and control for large-scale wireless sensor
network repair, by Kyle lathy in north Carolina state university on
may6, 2009
[11] L. Fortuna, M. Frasca, M.G. Xibilia, A. A. A , M. T. R s , M p pu A m
s, P gs T 1s International Conference on Energy, Power, and Control,
College of Engineering, The University of Basrah, Basrah, Iraq , pp.12-
15, Nov. 30 to Dec. 02, 2010.
[12] I. Hayas , N. I suk, M m g s, I P . O I . Symposium on
Micromechatronics and Human Science, Nagoya, 1998, pp.41-50.
[13] T. H. L , F.H. F. L u g, P.K.S. T m, P s ms us g uzz g , IEEE
international Conference on Industrial Electronics Socielty, vol. 3, no. 2,
pp.525-528, 1999.
[14] Embedded systems 2
nd
edition - RajKamal
[15] Programming AVR microcontroller - Dhananjay v.Gadre
[16] Robotics reference guide - J oseph A. Angelo
[17] Data networks and interfacing - Hawkins
[18] Texas - linear ICs manual
[19] Segnetics Digital ICs manual ZigBee Specification

Author Profile

C.Venkatesh received B.Tech
Degree from J .N.T.University,
Hyderabad and M.Tech Degree from
J .N.T.U.A.,Anantapuramu.
Presently he is with
Annamacharya institute of Technology &
Sciences, Rajampet, Andhra Pradesh,
India, working as an Assistant Professor
in Department of ECE. His research
interests include Embedded Systems, Signal Processing and Digital
Imaging. He presented many research papers in National &
International Conferences. He is a member of professional societies
like MISTE,MIE,IACSIT(Singapore), IAENG(Hong Kong),
UACEE(India), ISOC(Switzerland), APCBEES(China) and SIE
(Singapore).He acted as reviewer for many International conferences
and journals


Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
121



Zigbee For Vehicular Communication Systems
J .RamHarish Yadav
1
, K.Dhanunjaya
2

1
PG Student, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India.
2
Head of the Deepartment, Department of Electronics & Communication Engineering, ASCET, Gudur, A.P, India

1
r amhar i sh. j @gmai l . com


Abstract:Vehicular communication is a popular
topic in theacademia and the car industry. The aim of
this growing interestis to develop an effective
communication system for theIntelligent
Transportation System (ITS). In this paper
wepresented the model of wireless base station
goodput evaluation.We used wireless access point
model as a queuing system withvariable requests and
the auto traffic model. The performance ofthe
wireless networks can be impacted from a variety
ofparameters, such as radio communication range,
availablebandwidth and bit rate, the number of clients
in wireless networkrange and vehicle speed. The
basic parameters were analyzedand presented in this
paper.

Index Terms:Short Range Vehicle Network;
802.11n; wirelessnetwork; goodput; network
performance; transport; mobilestations; auto traffic;
vehicle speed; Markov chain.

I. INTRODUCTION

The needs to enhance road safety, traffic efficiency
and to reduce environmental impact of road transport
are seriouschange for both academics and industry.
Researchers aregreatly interested to develop
vehicular communication andnetworking technology
in two realistic ways vehicle tovehicle (V2V) in ad
hoc mode and vehicle to infrastructure(V2I) with
fixed nodes along the road. The potency toexchange
information wireless via V2X is a foundationstone
for building powerful Intelligent Transport
Systems(ITS). In Europe, USA and Japan are great
efforts madefrom automakers and governments to
reach single standardsthrough the several and
common projects such as CAR 2CAR
Communication Consortium, Vehicle Safety
Communication Consortium, and EUCAR SGA etc.
Result fromcommon effort is an international
standard, IEEE802.11p[2], also known as Wireless
Access for VehicularEnvironments (WAVE). This
standard will be used as thegroundwork for
Dedicated Short Range Communications(DSRC).


This type of communication has potential toimprove
safety on the road, traffic flow and provide
comfortfor passengers and drivers with expedited
applications suchas INTERNET, network games,
automatic electronic tollcollection, drive-through
payments, digital map update,wireless diagnostic and
flashing etc. DSRC is the one stepin the future,
because it lets inter-vehicle and vehicle
toinfrastructure wireless communication.

Wireless networking based on IEEE802.11
technology[3] has recently become popular and
broadly available atlow-cost for home networking
and free Wi-Fi orcommercial hotspots. The DSRC
starting idea was to equipvehicular network nodes
with off-the-shelf wirelesstechnology such as
IEEE802.11a. This technology is costeffective and
has potential to grow and new versions havebeen
recently produced. The latest standard of
wirelesslocal area network (WLAN) is IEEE802.11n
[4]. The IEEE802.11n standard promises to improve
and extend mostpopular WLAN standards by
significantly increasingthroughput, reliability and
reach.

Nowadays dispositions of WLAN-based
accesstechnology are predominantly to stationer
indoor andoutdoor users who are most slowly moving
and in rangelimited. Despite the fact that the standard
has not beendeveloped for fast dynamic usage,
nothing limits it to beevaluated for vehicular
communication systems. Themotivation is to
understand the interaction between thevehicle speed
and goodput of WLAN-based network.Realizing
field trials for goodput evaluation ofvehicular
wireless communication systems is very difficultand
costly because many vehicles and
communicationequipments need to be purchased or
rented, and also manyexperimenters need to be
employed. Given such problems,it is highly desirable
to obtain a mathematical descriptionof process with
real data from small scale scenarios ofpractical
measurement results and performance
evaluationsprior to conducting field trials as it is
made in this work.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
122



This paper constructs as follows: After introducing
theproblem in Section 1, Section 2 provides some
Vehicular Communication Systems. Then, in Section
3 SeVeCom Implementation and demonstrating the
analysis results inSection 4. Section 5 summarize and
concludes this paperwith a brief description on future
works.

II. VEHICULAR COMMUNICATION
SYSTEMS

There are significant differences between devices
such asmobile phones or desktop computers
connected to the Internetand devices in a VC system.
Differences in development,production, and
operation, determine VC-specific constraintsand
conditions:

1) Vehicles have a long life span, lasting several
yearsin most cases. This makes it hard to change on-
boardsystems as reaction to new upcoming risks to
the vehiclesafety.

2) Owners have constant physical access to and full
controlover vehicles. In spite of the involved safety
risks,many users might try to modify or enhance
theirvehicles. From a manufacturers point of view,
the riskof hardware tampering cannot be neglected.

3) No technical expertise on vehicle electronics or
VCsecurity aspects is expected from a user that runs
avehicle. Hence, the vehicular security measures
have tooperate autonomously with no need for
intervention orfeedback from the user.

4) Robustness requirements and time constraints are
demanding.Functions necessary, for example, for
drivingor alerts received via the VC system must be
processedin real-time: delays or errors could lead to
vehicle malfunctions,driving errors, and consequently
to physicaldamages and injuries.

5) Liability and conformance require precise
formulationof legal issues. Differing regulations and
requirements invarious countries make it even more
difficult to addressthese challenges.

These observations have consequences on the
implementationof a VC security system. Due to the
long vehicle lifecycle, it cannot be ensured that all
threats are thwarted at thetime of development.
Therefore, the VC security mechanismsshould be
flexible, adaptable, and extensible, to allow
lateradjustments to changing security requirements.
To address this need, we propose a component-based
security architecture forVC systems, which allows
adding, replace, and reconfigurecomponents (for
example, substitute cryptographic
algorithms)throughout the life cycle of the vehicle.

The large number and the variety of vehicles have
tobe taken into account. Even for a single car type,
differentproduction and equipment lines lead to many
distinct versionsand variants. Nonetheless, it should
be possible to integratea security system into all those
platforms. In addition, thecommunication stack and
security measures might be designedby different
teams or vendors; a situation that clearly
requireswell-defined but still flexible interfaces.
These reasons ledto the development of the so called
hooking architecture,which introduces special
hooks at the interface between everylayer of the
vehicular communication system. The
hookingarchitecture introduces an event-callback
mechanism into thecommunication stack which
allows adding security measureswithout the need to
change the entire communication system.The security
system in a vehicle has to fulfill real-time ornear real-
time requirements. For the underlying
cryptographicprimitives, this implies optimized
cryptographic hardware,in order to guarantee the near
real-time performance. Thepotential trade-off
between security and performance has tobe well
balanced.To enable VC systems to withstand future,
yet unknownattacks, besides the traditional
prevention-oriented approach,functionalities to detect
attacks, such as intrusion detectioncapabilities, and to
recover after an attack, are needed. In thelong run, the
goal is to enhance the resilience of the system.

III. SEVECOM IMPLEMENTATION

The SeVeCom project defines baseline
securityarchitecture for VC systems. Based on a set
of designprinciples, SeVeCom defines an architecture
that comprisesdifferent modules, each addressing
certain security and privacyaspects. Modules contain
components implementing one partof system
functionality. The baseline specification providesone
instantiation of the baseline architecture, building on
wellestablishedmechanisms and cryptographic
primitives, thusbeing easy to implement and to
deploy in upcoming VCsystems.

A. Baseline Architecture: Deployment View

The SeVeCom baseline architecture addresses
different aspects,such as secure communication
protocols, privacy protection,and in-vehicle security.
As the design and development ofVC protocols,
system architectures, and security mechanismsis an
ongoing process, only few parts of the overall system
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
123



areyet finished or standardized. As a result, a VC
security systemcannot be based on a fixed platform
but instead has to beflexible, with the possibility to
adapt to future VC applicationsor new VC
technologies.

To achieve the required flexibility, the SeVeCom
baselinearchitecture consists of modules, which are
responsible fora certain system aspect, such as
identity management. Themodules, in turn, are
composed of multiple components eachhandling a
specific task. For instance, the Secure
CommunicationModule is responsible for
implementing protocols forsecure communication
and consists of several components,each of them
implementing a single protocol. Components
areinstantiated only when their use is required by
certain applications,and they use well-defined
interfaces to communicatewith other components.
Thus, they can be exchanged by morerecent versions,
without other modules being affected.



Fig.1. Baseline Architecture: Deployment View.

As shown in Fig. 1, the Security Manager is the
centralpart of the SeVeCom system architecture. It
instantiates andconfigures the components of all
other security modules andestablishes the connection
to the Cryptographic Support Module.To cope with
different situations, the Security Managermaintains
different policy sets. Policies can enable or
disablesome of the components or adjust their
configuration, forexample, to enhance or relax the
parameters for a pseudonymchange under certain
circumstances.

B. Communication Stack Integration
To be independent of the actual communication
stack, theintegration of the SeVeCom security system
into the protocolstack is based on a hooking concept,
inspired by similararchitectures such as the Linux
Netfilter kernel subsystem.Inter Layer Proxies (ILPs)
are inserted at several points in thecommunication
stack. Every ILP maintains a list of callbackhandlers
that are to be notified of certain events.During
initialization, the SeVeCom components can
registerat an ILP, subscribing for certain message
types and direction(up or down the stack). Therefore,
they have to implement anevent listener interface and
use the registerHandler() methodto connect to an ILP.
Some components may have to register atmultiple
ILPs, subscribing for different kinds of packets.
Whena message arrives at an ILP, an event callback
is triggered forall components that have registered for
this message type andtheir eventHander() method is
called. The callback includes areference to the
received message, and the component is thenable to
inspect or modify it. By the return value the
componentindicates if the message was modified, if it
should be reinsertedinto the stack, or if it should be
simply dropped by the ILP. The Secure Beaconing
Component, for example, connects tothe ILP above
the MAC layer and checks the signatures of
allincoming beacon messages. Beacons with invalid
signaturesare either discarded or tagged. Using this
hooking architecture,it is possible to transparently
integrate security functionalityinto an existing
network stack with minimal modifications.

Whereas events are triggered by the communication
stack,the security system can also access the stack by
means ofcommand calls using a well-defined API
offered by stacklayers. Command calls could, e.g.
instruct the MAC layer toset its MAC address to that
of a new pseudonym.The hooking concept makes
certain assumptions aboutthe network stack. It
assumes a layered architecture, wherethe ILPs can be
inserted in between, and the stack has toimplement a
certain command API, e.g. for change of
MACaddresses. To be able to port the SeVeCom
architecture tomany different communication
platforms, we also providean additional convergence
layer: This defines an abstractioninterface that
proxies call between the communication systemand
the security components. Whenever the SeVeCom
systemis ported to a new platform, besides adapting
to differentpacket formats, only the ILPs and the
convergence layer haveto be modified, while all other
components remain unaffectedboth in terms of
security and communication.

C. Hardware Security Module

As explained, the purpose of the Hardware
SecurityModule (HSM) is to provide a physically
protected environmentfor the storage of private keys
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
124



and for the execution ofcryptographic operations
using them. Clearly, the full implementationof a
HSM is beyond the scope of the SeVeComProject,
but we can summarize the main requirements
thatsuch an implementation should meet in order to
be applicablefor securing vehicle communication
systems.First of all, the HSM must be tamper
resistant, to someextent. High-end tamper resistant
modules (such as the IBM4758 Cryptographic
Coprocessor) are too expensive to beadded to every
vehicle. At the same time, we observe thatLow-end
tamper resistant devices (such as smart cards) donot
provide all the functionality that we need. In
particular,commercially available low-end devices do
not have built-inbatteries, and consequently, cannot
provide a trusted internalclock. As pointed out,
without a trusted source of time,such devices are not
able to produce time-stamps that canbe trusted by
other participants of the system. Therefore, weneed
an HSM implementation somewhere between high-
endand low-end devices. A potential approach is to
implementthe HSM as an Application-Specific
Integrated Circuit(ASIC)with some special coating
that provides a certain level oftamper resistance.
Such a customized device can provide allthe
necessary functionality by design and it can be
producedin large quantity at sufficiently low costs.

Second, the HSM must have an API, through which
itcan provide services to the other modules of the
securityarchitecture that run on the OBU. This API
should supportthe digital signature and timestamping
service, the decryptionservice, as well as the key and
device management servicesdescribed. We specified
such an API in the SeVeComProject, however,
lacking the appropriate HSM hardware; weonly
implemented it in the form of a software library
runningon a general purpose computer. Nevertheless,
besides beinguseful for demonstration purposes, our
implementation canalso serve as a reference for
future implementations on realHSM devices. In our
implementation, we used ECDSA fordigital signature
generation, ECIES with HMAC-SHA1 andAES-CBC
for encryption, and we fully implemented the
keymanagement services of the HSM described.

Finally, we note that some examples published
showthat physically secure modules can successfully
be attackedthrough their weakly designed API. For
this reason, we usedformal verification techniques to
verify the SeVeCom HSMAPI. Our method is based
on the applied pi-calculus and anautomated
verification tool called Prove it. We proved that akey
generated by an adversary cannot be implanted as a
newroot key in the HSM through the API.
Additionally, short-termand long-term private keys
are proved not to be revealed asthe result of possible
series of function calls.

D. In-Vehicle Security

In order to achieve their full potential, VC systems
need access to the in-car network and sensors that
observe thecurrent status of the vehicle and the
environment. This enablesthe VC system to process
signals such as emergency braking,airbag activation,
and slippery road detection, thus greatlycontributing
to the avoidance of accidents and improvementof
road safety.On-board system signals are transferred
inside the carthrough different networks and
domains. Usually, the networkarchitecture and the in-
car gateways restrict the signals tothe defined
network segments and prevent information
fromleaving its dedicated domains. This clear
architecture andstrict separation is one measure that
ensures the entire vehicle,especially its vital functions
(brakes, engine or airbag control),always operate
reliably and cannot be attacked from the outside.

If this were to be changed into a more open
architecture,for example, allowing reading out sensor
information fromin-vehicle networks or displaying
and reacting to warningmessages from external
sources, it would be absolutely necessaryto ensure
that in-vehicle systems are protected from
anyexternal malicious influence.The In-vehicle
Security Module protects the interface betweenthe in-
car networks and the wireless communicationsystem.
It controls external access to the in-car networks,on-
board control units and vehicle sensor data, but it
alsoensures that data and services required by other
V2V andV2I applications are provided correctly.
Within the in-vehiclesecurity module, two main
components are provided: (i) Afirewall that controls
the data flow from external applicationsto the vehicle
and backwards, and (ii) an Intrusion DetectionSystem
(IDS) that constantly monitors the status of the in-
carsystems and provides real-time detection of
attacks.

The firewall realizes a packet or application based
firewallapproach. Its rule-based table states which
application isallowed to access each kind of data or
service the IDS candynamically add rules to the
firewall table, in order to denyaccess for a specific
application or disable a service.The IDS is based on
an anomaly detection approach,which implies that
normal on-board system behavior is clearlydefined
and specified. If an event results in an on-board
systemstate that is not part of the standard
specification, a potentiallydangerous situation is
detected. Depending on the source andtype of the
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
125



event, appropriate reactions are taken to get
thesystem back to a secure and safe state.

IV. ANALYTICAL MODEL

Realizing field trials for good put evaluation of
vehicular wireless communication systems is very
difficult and costly. Numerous vehicles and
communication equipments need to be involved, and
also many experimenters need to be employed. In this
case, it is highly desirable to obtain theoretical
analysis with real data from small scale scenarios of
practical measurement results and perform an
evolution prior to conducting field trials. In terms of
analysis methods, were mapped previous
approximations of vehicle mobility and good put into
Markov M/M/1/N chain model. Use of Markov
model is novel for evaluation of IEEE802.11n
standard in a mobwith legacy standard (i.e.
IEEE802.11g)

A. System model descript

For this model computation, was considering the case
where the access points transmission data rate is
variable through the access point coverage
range.Primitive packets flow from finite wireless
mobile users N and arrive to an infinite buffer of the
system and are served by the server or wireless
access point.

In this case our system is expressed by the Kendall
notation like M/M/1//N where first M-defines
exponential inter arrival times between packet
distribution (Poisson process), second M- defines
exponential data packets transmission time
distribution , next number defines the transmission
channel and N- represents the number of packet
sources.Queuing models for M/M/1/N systems are
very elegant in analysis of wireless data networks in
transmission channel with no packet loss and vehicles
simultaneously under the coverage of the access point
speed-N (v) (i.e, (v). Based on this M/M/1/N
queuing model the average good put by a vehicle can
be computed as follows:



where
0
represent the probability of the idle system



where j = 1,2,3,.,N(v), the data packet
transmission rate of the channel between vehicular
and base station, is the packet arrival rate in the
coverage range of the access point

B. Results

In the computation of the analytical model in the
previous subsection was constructed a topology with
an access point sending file data to all vehicles within
coverage range of an access point. In the
computation, each vehicle maintains its speed as it
derives through the access point coverage range. The
computation compares the results derived from trial
field tests with analytical model for the single-lane in
vehicle traffic.

The range of good put that a vehicle can receive
from the access point per pass shown in Fig.2. The
results here are for the case where there are two types
of vehicles, i.e, wireless-equipped and non-wireless-
equipped vehicles. The type of vehicles can be
interpreted as the penetration rate of wireless-
equipped vehicles for use.

From the Fig.2 can make the following observations:

At low traffic density corresponding to high vehicle
speed, there are few vehicles and as such there is a
few connections using the access point resource and
the value of good put is close to maximum. It is about
two times less than plausible maximum good put.


Fig.2. Average good put of a vehicle at different
speed and WLAN penetration rate

On low velocity increase value of vehicles and
bandwidth connections increases leading to lower
values of good put for the individual user.

Despite reduction of maximum good put due to
mobility at a velocity from 50 km/h to 100 km/h
improves the good put value of a vehicle.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
126



Penetration rates specify the possible optimal values
of WLAN performance.


V. CONCLUSION AND FUTURE WORK

In this article was presented field trial
evaluationstogether with theoretical analyses of the
IEEE802.11nstandard comparing with legacy
standard in the vehicleenvironment. The trial field
test was performed in thecontext of simple scenario
of one vehicle and access point.At various velocities
has been testing the performance ofWLAN. Wireless
network link under fluent number ofvehicles
respectively active users simultaneously
realizingsuch field trials for goodput evaluation is
very difficult andcostly. Therefore a simple
mathematical model for goodputevaluation of
vehicular communication systems in V2Iscenario was
presented and analyzed for understanding thebasic
processes in wireless data networks prior
toconducting larger field trials.

We mark that while numbers of necessary real time
application of vehicular networks are the
dissemination ofsafety and traffic condition
messages, we can assume Wi-Fi
for vehicle communication systems in the near future
willalso be requested to provide different
applications, for e.g.web browsing, video streaming,
VoIP, downloading files,WiFi radio, etc. These types
of applications have arequirement for high
throughput during connections to theaccess point and
existing mobile communication systemsexcept
WLAN arent able to provide growing needs.And it
is also important to note that the results wereshowing
her serve as information for future analysis anddesign
of vehicle networking systems.

VI. REFERENCES

[1] Janis Jansons, Ernests Petersons,
NikolajsBogdanovs, WiFi for Vehicular
Communication Systems, 2013 27th International
Conference on Advanced Information Networking
and Applications Workshops.

[2] "Part 11: Wireless LAN Medium Access Control
(MAC) and PhysicalLayer (PHY) Specifications
Amendment 6: Wireless Access inVehicular
Environments",
http://ieeexplore.ieee.org/servlet/opac?punumber=55
14473, IEEE,15.July, 2010.

[3] IEEE 802.11, The Working Group for WLAN
Standards,http://grouper.ieee.org/groups/802/11/ ,
April, 2006.

[4] Part 11: Wireless LAN Medium Access Control
(MAC) and PhysicalLayer (PHY) Specifications
Amendment 5: Enhancements for
HigherThroughput,
http://ieeexplore.ieee.org/servlet/opac?punumber=53
0729, IEEE, 29.October, 2009.

[5] J. P.Singh, N. Bambos, B. Srinivassan and D.
Clawin, Wireless LANperformance under varied
stress conditions in vehicular trafficscenarios,
proceedings of Vehicular Technology Conference,
2002,Vol. 2, pp. 24-28.

[6] J. Ott, D. Kutscher, Drivethru Internet: IEEE
802.11b forAutomobile Users, IEEE Infocom, Hong
Kong, 2004.

[7] R. Gass, J. Scott, C. Diot, Measurements of In
Motion 802.11Networking, WMCSA '06.
Proceedings, 2006, pp. 69-74.

[8] M. Wellens, B. Westphal, P. Mhnen
Performance Evaluation ofIEEE 802.11based
WLANs in Vehicular Scenarios, Proc. VTCSpring,
2007. pp. 11671171.

[9] M. Rubinstein, F. Ben Abdesslem, S.
RodriguesCavalcanti, M. EliasMitreCampista, R.
Alves dos Santos, L. Costa, M. Dias de Amorim,O.
Duarte, Measuring the capacity of incar to incar
vehicularnetworks IEEE Communications
Magazine, Vol. 47., Iss. 11, 2009.,pp. 128136.

[10] P. Richards Shock waves on the highway.
Operations Research 4,1956, 4251.

[11] M. Lighthill and G. Whitham On kinematic
waves: II. A theory oftraffic on long crowded roads.
Proc. Roy. Soc. of London A 229,1955, pp 317345.

[12] B. Kerner and P. Konhauser, Structure and
parameters of clusters intraffic flow, Physical
Review E 50, 1994, pp 5483.

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
127


7




































































































































Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
128
An Efficient and Large data base using Subset Selection Algorithm
for Multidimensional Data Extraction
M.Nagendra
M.Tech, Dept. of CSE, Audisankara College of Engineering & Technology, Gudur, A.P, India

Abstract- Feature subset selection is an effective way for reducing dimensionality, removing irrelevant
data, increasing learning accuracy and improving results comprehensibility. This process improved by
cluster based FAST Algorithm and Fuzzy Logic. FAST Algorithm can be used to Identify and removing
the irrelevant data set. This algorithm process implements using two different steps that is graph theoretic
clustering methods and representative feature cluster is selected. Feature subset selection research has
focused on searching for relevant features. The proposed fuzzy logic has focused on minimized redundant
data set and improves the feature subset accuracy.The FAST algorithmworks in two steps. In the first
step, features are divided into clusters by using graph-theoretic clustering methods. In the second step, the
most representative feature that is strongly related to target classes is selected fromeach cluster to forma
subset of features. Features in different clusters are relatively independent; the clustering-based strategy
of FAST has a high probability of producing a subset of useful and independent features. To ensure the
efficiency of FAST, we adopt the efficient minimum-spanning tree clustering method.
Index Terms Feature Extraction, MinimumSpanning Tree, Clustering
I. INTRODUCTION
The performance, robustness, and usefulness of classification algorithms are improved when relatively
few features are involved in the classification. Thus, selecting relevant features for the construction of
classifiers has received a great deal of attention.With the aimof choosing a subset of good features
with respect to the target concepts, feature subset selection is an effective way for reducing
dimensionality, removing irrelevant data, increasing learning accuracy, and improving result
comprehensibility. Many feature subset selection methods have been proposed and studied for
machine learning applications.
i. Existing System
The embedded methods incorporate feature selection as a part of the training process and are usually
specific to given learning algorithms, and therefore may be more efficient than the other three categories.
Traditional machine learning algorithms like decision trees or artificial neural networks are examples of
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
129
embedded approaches. The wrapper methods use the predictive accuracy of a predetermined learning
algorithmto determine the goodness of the selected subsets, the accuracy of the learning algorithms is
usually high. However, the generality of the selected features is limited and the computational complexity
is large. The filter methods are independent of learning algorithms, with good generality. Their
computational complexity is low, but the accuracy of the learning algorithms is not guaranteed. The
hybrid methods are a combination of filter and wrapper methods by using a filter method to reduce search
space that will be considered by the subsequent wrapper. They mainly focus on combining filter and
wrapper methods to achieve the best possible performance with a particular learning algorithm with
similar time complexity of the filter methods.
ii. Proposed System
Feature subset selection can be viewed as the process of identifying and removing as many irrelevant and
redundant features as possible. This is because irrelevant features do not contribute to the predictive
accuracy and redundant features do not redound to getting a better predictor for that they provide mostly
information which is already present in other feature(s). Of the many feature subset selection algorithms,
some can effectively eliminate irrelevant features but fail to handle redundant features yet some of others
can eliminate the irrelevant while taking care of the redundant features. Our proposed FAST algorithm
falls into the second group. Traditionally, feature subset selection research has focused on searching for
relevant features. A well-known example is Relief which weighs each feature according to its ability to
discriminate instances under different targets based on distance-based criteria function. However, Relief
is ineffective at removing redundant features as two predictive but highly correlated features are likely
both to be highly weighted. Relief-F extends Relief, enabling this method to work with noisy and
incomplete data sets and to deal with multiclass problems, but still cannot identify redundant features.
iii. Framework of Feature Extraction

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
130
Fig: Framework of Feature Cluster-Based Extraction Algorithm
II. FEATURE CLUSTER BASED EXTRACTION ALGORITHM
Irrelevant features, along with redundant features, severely affect the accuracy of the learning machines.
Thus, feature subset selection should be able to identify and remove as much of the irrelevant and
redundant information as possible. Moreover, good feature subsets contain features highly correlated
with the class, yet uncorrelated with each other. Keeping these in mind, we develop a novel algorithm
which can efficiently and effectively deal with both irrelevant and redundant features, and obtain a good
feature subset. We achieve this through a new feature selection framework which composed of the two
connected components of irrelevant feature removal and redundant feature elimination. The irrelevant
feature removal is straightforward once the right relevance measure is defined or selected, while the
redundant feature elimination is a bit of sophisticated.
In our proposed FAST algorithm, it involves (i) the construction of the minimumspanning tree (MST)
froma weighted complete graph; (ii) the partitioning of the MST into a forest with each tree representing
a cluster; and (iii) the selection of representative features from the clusters. Feature subset selection can
be the process that identifies and retains the strong T-Relevance features and selects R-Features from
feature clusters. The behind heuristics are that
1) Irrelevant features have no/weak correlation with target concept;
2) Redundant features are assembled in a cluster and a representative feature can be taken out of the
cluster.
III. ALGORITHM AND TIME COMPLEXITY ANALYSIS
The proposed FAST algorithm logically consists of three steps: (i) removing irrelevant features, (ii)
constructing a MST fromrelative ones, and (iii) partitioning the MST and selecting representative
features.
i. Algorithm: FAST
Inputs: (F
1
,F
2
,,F
m
,C) - the given data set 0 - the I Rclc:oncctrcsolJ.
Output: S sclcctcJcoturcsubsct.
1 ori =1 tomJo
2 I Rclc:oncc =Su(F

,C)
3 iI Rclc:oncc >0tcn
4 S =S {F

};
5 0 =NuII;
6 orcocpoirocoturcs|F

i
,F
]
i
| SJo
7 F Corrclotion =Su(F

i
,F
]
i
)
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
131
8 AJJF

i
und
o
F
]
i
to0witFCorrclotion
ostcwcigtotccorrcsponJingcJgc
9 minSponIrcc =Prim(0);
10 Forcst =minSponIrcc
11 orcoccJgcE
]
ForcstJo
12 iSu(F

i
,F
]
i
) <Su(F

i
,C) Su(F

i
,F
]
i
) <Su(F
]
i
,C)tcn
13 Forcst =Forcst E
]

14 S =
15 orcoctrccI

ForcstJo
16 F
R
]
=orgmox
P
k
|
1
i
Su(F
k
i
,C)
17 S =S |F
R
]
|;
18 rcturnS

ii. Time Complexity
The major amount of work for Algorithm 1 involves the computation of SU values for T-
Relevance and F-Correlation, which has linear complexity in terms of the number of instances in a given
data set.
The first part of the algorithm has a linear time complexity O(m) in terms of the number of
features m. Assuming k(1 k m) features are selected as relevant ones in the first part, when k =1, only
on feature is selected. The second part of the algorithmfirstly constructs a complete graph fromrelevant
features and the complexity is O(k
2
), and then generates a MST fromthe graph using PrimAlgorithm
whose time complexity is O(k
2
). The third part partitions the MST and Chooses the representative features
with the complexity of O(k). Thus when 1<k m, the complexity of the Algorithmis O(m)
CONCLUSION
This project presented a novel clustering based feature extraction algorithm for high
dimensional data. The algorithminvolves 1) removing irrelevant features, 2) constructing a minimum
spanning tree from relative ones, and 3) partitioning the MST and extracting representative features.
The purpose of cluster analysis has been established to be more effective than feature extraction
algorithms. Since high dimensionality and accuracy are the two major concerns of clustering, we have
considered themtogether in this paper for the finer cluster for removing the irrelevant and redundant
features. The proposed supervised clustering algorithm is processed for high dimensional data to improve
the accuracy and check the probability of the patterns. Retrieval of relevant data should be faster and
more accurate. This displays results based on the high probability density thereby reducing the
dimensionality of the data.
FUTURE ENHANCEMENT
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
132
In the near feature, we plan to analyze the distinct types of relationship measures and some
formal properties of feature space.


REFERENCES
[1] Almuallim H. and Dietterich T.G., Algorithms for Identifying Relevant Features, In Proceedings of
the 9th Canadian Conference on AI, pp 38-45, 1992.
[2] Arauzo-Azofra A., Benitez J .M. and Castro J .L., A feature set measure based on relief, In Proceedings
of the fifth international conference on Recent Advances in Soft Computing, pp 104-109, 2004.
[3] Bell D.A. and Wang, H., A formalism for relevance and its application in feature subset selection,
Machine Learning, 41(2), pp 175-195, 2000.
[4] Biesiada J . and Duch W., Features election for high-dimensionaldataa Pearson redundancy based
filter, AdvancesinSoftComputing, 45, pp 242C249, 2008.
[5] Butterworth R., Piatetsky-Shapiro G. and Simovici D.A., On Feature Selection through Clustering, In
Proceedings of the Fifth IEEE international Conference on Data Mining, pp 581-584, 2005.
[6] Cardie, C., Using decision trees to improve case-based learning, In Proceedings of Tenth International
Conference on Machine Learning, pp 25-32, 1993.
[7] Butterworth R., Piatetsky-Shapiro G. and Simovici D.A., On Feature Selection through Clustering, In
Proceedings of the Fifth IEEE international Conference on Data Mining, pp 581-584, 2005.
[8] Cardie, C., Using decision trees to improve case-based learning, In Proceedings of Tenth International
Conference on Machine Learning, pp 25-32, 1993.
AUTHORS
MosaNagendrahas received his B.Tech degree in Information Technology from
J agans College of Engineering & Technology, Venkatachalamaffiliated to J NTU,
Anantapur in 2012 and pursuing M.Tech degree in Computer Science & Engineering
at Audisankara College of Engineering & Technology, Gudur affiliated to J NTU,
Anantapur in (2012-2014).
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
133
A Review on ECG Arrhythmia Detection
based on DD-DW Transformation

1
B.Maheswari
2
T.Nirmala M.Tech
3
A.Rajani M.Tech
M .Tech Student Assistant professor Assistant professor
Department of ECE,
Annamacharya Institute of Technology and Sciences, Tirupati, India-517520.
1
mahi sol mon@gmai l . com
2
ni mmohan12@gmai l . com
3
r aj ani r evant h446@gmai l . com
AbstractMachine learning of Electrocardiogram
(ECG) is a core component in any of the ECG-based
healthcare informatics system. Since the ECG is a
nonlinear signal, the subtle changes in its amplitude and
duration are not well manifested in time and frequency
domains. Therefore, a machine-learning approach to
screen arrhythmia from normal sinus rhythm from the
ECG is proposed. The methodology consists of R-point
detection using the double density discrete wavelet
transformation (DD-DWT) decomposition, statistical
validation of features. The average accuracy of
classification is used as a benchmark for comparison.
Support vector machine (SVM) kernel is used as
classifier for better classification purpose. The proposed
method will provide better results compared to state-of-
art criteria like Signal quality indices (SQI) based
feature extraction method. DD-DWT based ECG
feature extraction is used in clinical purpose like
intensive care unit (ICU) monitors for diagnosis of
abnormalities in heart beat and is used for
psychological analysis of human activities.

Index TermsElectrocardiogram (ECG), intensive
care unit (ICU), signal quality.

I. INTRODUCTION
In certain signal processing applications, like de
noising, over complete transforms can offer a better
tradeoff between performance and complexity,
compared to critically sampled transforms. A
distinguished member of the family of over complete
discrete wavelet transforms (DWT) is the double
density (DD) DWT [1], based on the filter bank
shown in Figure 4. The input signal is split in three
channels, each decimated by a factor of two. The
signal on the first channel is processed by an identical
filter bank.
The DD-DWT [1] is expansive with a factor of
two, compared to the critically sampled DWT
arrhythmias based on their characteristics features
extracted from ECG signals. Intended for
Premature ventricular contraction (PVC) beat
exposure RR interval ratio and power of beat is
calculated. Dominant notched R wave, dominant S
wave, QRS duration and direction of T wave are
used for the detection of Left bundle branch block
(LBBB) and Right bundle branch block ( RBBB).
The rest of this paper is organized as follows:
presents the ECG signal processing, ECG De noising,
ECG parameter calculation, DD-DW Transformation,
Feature Extraction and Selection. Describes the
detection of arrhythmia. Finally, the summarizes
the result & conclusion of this work.

II. EXISTING APPROACH
A. Pre-processing of ECGs
Each channel of ECG was filtered to remove
baseline wander and low frequency noise using a
high pass filter with a cut-off at 1 Hz. QRS detection
was performed on each channel individually using
two open source QRS detectors (eplimited and wqrs).
since eplimited is less sensitive to noise, we prefer
eplimited detector.

B. Signal quality indices
Six signal quality indices (SQIs) were chosen
based on earlier work and run on each of the m =12
leads separately, producing 72 features per recording:
1. lSQI: The percentage of beats detected on each
lead which were detected on all leads.
2. bSQI: The percentage of beats detected by wqrs
that were also detected by eplimited.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
134
3. fSQI: The ratio of power P(5-20Hz)/P(0-fnHz),
where fn=62.5 Hz is the Nyquist frequency.
4. sSQI: The third moment (skewness) of the
distribution.
5. kSQI: The fourth moment (kurtosis) of the
distribution.
6. pSQI: The percentage of the signal xm which
appeared to be a flat line (dxm/dt < where
=1mV).

C. Support Vector Machine Classification
In this section, a brief description of the
two and multi-class SVM classification concept is
reviewed. Support Vector Machines (SVMs) [17]
are very popular and powerful in pattern learning
because of supporting high dimensional data and
at the same time, providing good generalization
properties. Moreover, SVMs have many usages in
pattern recognition and data mining applications,
phoneme recognition, 3D object detection, image
classification, bio-informatics etc. At the beginning,
SVM was formulated for two-class (binary)
classification problems. The extension of this
method to multi-class problems is neither straight
forward nor unique. DAG SVM is one of the
methods that have been proposed to extend SVM
classifier to support multi-class classification [17].

D. Binary Support vector machine formulation
Let be a set of n
training samples, where is an m-
dimensional sample in the input space. In a support
vector machine, the optimal hyper plane is obtained
by maximizing the generalization ability of the
SVM. However, if the training data are not linearly
separable, the obtained classifier may not have high
Generalization ability, even though the hyper planes
are determined optimally. To enhance linear
separability, the original input space is mapped into a
high-dimensional dot-product space called the
feature space. Now using the nonlinear vector
function that maps the m-dimensional input vector x
into the l-dimensional feature space.

E. Multi Class Support vector machine
As described before, SVMs are intrinsically
binary classifiers but the classification of ECG
signals often involves more than two classes.
In order to face this issue, a number of multiclass
classification strategies can be adopted.

III. PROPOSED CRITERIA
A.ECG Signal Processing
The proposed method includes processing
and parameter calculation of ECG and then detection
of cardiac arrhythmia using an algorithm developed
in MATLAB 7.12 simulation tool. The algorithm is
tested over MIT-BIH Arrhythmia database.

B.ECG De noising
ECG signals are usually corrupted by several
noises like 50 Hz power line interferences, baseline
wander and electro mayogram (EMG). Therefore, the
signal needs to be preprocessed before applying any
detection algorithm. Wavelet de noising and S- Golay
Filter is used for removal of baseline wander and
high frequency noise. ECG unfiltered data is passed
through baseline wandering removal function,
followed by wavelet based high frequency noise
removal. The data is then smoothed. Further using S-
Golay filter. All the modules are implemented and
simulated in MATLAB.

C. ECG Parameter Calculation
The purpose of the feature extraction process
is to select and retain relevant information from
original signal. The Feature Extraction stage extracts
diagnostic information from the ECG signal. In order
to detect the peaks, specific details of the signal are
selected. The detection of R peak is the first step of
feature extraction. R peak is detected by using Pan-
Tompkins algorithm. The intervals QRS, PR and QT
are calculated by searching for corresponding onset
and offset points in the wave. The separate logic was
implemented for identifying P-onset, Q and S points,
once R-peak was located using Pan-Tompkins
algorithm.
The window is selected around R-wave and
the minimum of the points within this window are
declared as Q and S points. In the differentiated
signal ECG, a window of 155 ms is defined starting
225 ms before R-peak position. In this window, we
search for maximum and minimum signal value. The
P-wave peak is assumed to occur at the zero-
crossings between maximum and minimum values
within the selected window. Once P-wave peak is
found, we proceed to locate waveform boundary-P-
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
135
wave onset. Similarly T wave ms with abnormal
morphology. Premature i.e. occurs earlier than
would be expected for the next sinus impulse.
Discordant ST segment and T wave changes. There
are five different types Of PVC, first Bigeminy
every other beat is PVC, second Trigeminy every
third beat is a PVC, third Quadrigeminy every fourth
beat is a PVC, forth Couplet two consecutive PVCs
and last Triplet three consecutive PVCs. The main
characteristic of PVCs is its premature occurrence.
This characteristic is measured by relating the RR
interval lengths of heart cycles adjacent to the PVC.
In case of a PVC, these lengths should be notoriously
different .The method for classifying the abnormal
complexes from the normal ones is based on the
concepts of RR interval ratio of detected R peaks
and energy analysis of ECG signal.

D. Feature Extraction and Selection
In this section explain the characteristics
of the extracted feature from the ECG signals and
the procedure designed for the extraction. Figure 1,
presents the block diagram of the proposed
arrhythmia classification.

Fig.1.Block diagram of proposed arrhythmia
classification

i) Dataset Description
In this experiments were conducted on the ECG
data as the basic signal for classification. In recent
researches, the annotated ECG records, available at
the MIT-BIH arrhythmia database, have been widely
used for the evaluation of the performance of
different classifiers. The database has 48 records with
each record being an ECG signal for the duration of
30 minutes.

Fig 2: Sample signal of Normal, Paced, LBBB and
RBBB
In particular, the considered beat types refer to
following classes: normal sinus rhythm (N), right
bundle branch block (RB), left bundle branch block
(LB), and paced beat (P). In Figure 2, sample of four
N, RB, LB, and P beats are noticeable.

ii) Feature Description
For each signal nineteen temporal features such as R-
R interval, PQ interval, PR interval, PT interval and
three morphological features are recognized. These
features are manually extracted for each beat and put
into a separate vector. Each vector is tagged with one
the four possible labels N, P, LB, RB.

Fig 3: sample features, ST interval, TP interval
and RR interval

The three morphological features by computing the
maximum and the minimum values of a beat in ECG
signal. Signals of each beat are scaled, using the
following formula, such that the range of every
signal is between zero and one.


The minimum and maximum voltages between the
first and the second R feature is computer first and
the normalization action is performed [0 1]. As
mentioned before, we considered percent that are
higher than 0.2, 0.5 and 0.8 as three features. Six of
the 22 features called basic features are: R1, S, T, P,
Q, R2 and the rest are called derived features. The
derived features are calculated using the basic
features via a semiautomatic procedure. We suggest
first and second R point to expert using an algorithm
based on maximum-minimum. Then the expert
distinguishes appropriate points(R, S, T, P, Q, and
R).

Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
136
E.Double Density Wavelet Transform
1)1-D Double Density DWT (DD)
The DD was consisted of two stages of filter banks as
shown in Fig. 4:
(i)Analysis
In the analysis filter banks, three filters were
implemented and the original signals were down-
sampled by 2 in order to decompose the signals into
three sub-bands. The low frequency sub-band c(n)
was produced by low pass filter h
0
(-n) and the two
high frequency sub-bands d
1
(n) and d
2
(n) were
produced by high pass filters h
1
(-n) and h
2
(-n).

(ii) Synthesis
The synthesis filter banks were the inverse of
analysis filter banks where the three sub-bands were
up-sampled by 2, filtered by the high pass filter h
0
(n)
and the two low pass filters h
1
(-n) and h
2
(-n). The
filtered signals were combined to form the output
signal x(n).

Fig.4.Filter bank diagram of Double Density DWT

2) Double Density Complex DWT (DDC)
The input data, were processed by two parallel
iterated filter banks h
i
(n) and g
i
(n) where i =0, 1, 2.
The real part of a complex wavelet transform [2] was
produced by the sub-band signals of the upper DWT
and the imaginary part was produced by the lower
DWT.
Fig.5.Filter bank diagram of Double Density
Complex DWT
F.DETECTION OF ARRHYTHMIA
A premature ventricular contraction (PVC) also
known as a premature ventricular complex,
ventricular premature contraction (complex or
complexes) (VPC), is Broad QRS complex ( 120)
range [0, 1]. Therefore, S SQI and k SQI were
normalized by subtracting the mean and dividing by
the standard deviation, with both of which computed
on the training set.
Sensitivity (Se) measures the proportion of poor
quality signals that have been correctly identified as
such. Specificity (Sp) measures the proportion of
good quality signals that have been correctly
identified as acceptable, and accuracy (Ac)
corresponds to the proportion of signals that have
correctly been classified. These statistical measures
are calculated from the number of true positive (TP),
true negative (TN), false positive (FP) and false
negative (FN) with Se = TP/(TP+FN), Sp =
TN/(FP+TN) and Ac =(TP+TN)/(TP+TN+FP+FN).
The energy of ECG signal is calculated for each beat
and RR interval ratio is also calculated. The threshold
for energy is taken as 65% of maximum energy and
for ratio 70% of maximum ratio value. If RR interval
ratio and energy is less than threshold PVC beat is
detected.

IV. EXPERIMENTAL RESULTS
To report the two examples of design. The SDP
problem has been solved using the library SeDuMi.
The optimization problem has been solved with the
large scale version of the Matlab function fmincon,
with default parameters, excepting TolCon (the
tolerance for satisfying the constraints), which has
been set. The tolerance from has been set , which
means that the PR constraints are satisfied with very
good precision. The results obtained after 10
refinement iterations, although in the second example
fewer iterations give approximately the same result.

V. CONCLUSION
The contribution of this is twofold. Firstly, the
Hilbert transform [4] FIR approximation problem
can be expressed as an SDP problem and solved
reliably. Secondly, a complete algorithm to compute
the second filter bank of a dual-tree double-density
DWT. The algorithm comprises an initialization step
based on SDP, followed by iterative refinement via
nonlinear optimization.
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
137

REFERENCES
[1] I.W. Selesnick, The Double-Density Dual-Tree
DWT, IEEE Trans. Signal Processing, vol. 52, no.
5, pp. 13041314,May 2004.
[2] N.G. Kingsbury, Complex wavelets for shift
invariant analysis and filtering of signals, J. Applied
and Computational Harmonic Analysis, vol. 10, no.3,
pp. 234253,May 2001.
[3] I.W. Selesnick, R.G. Baraniuk, and N.G.
Kingsbury,The Dual-Tree ComplexWavelet
Transform,IEEE Signal Proc.Magazine, vol. 22, no.
6, pp. 123151, Nov. 2005.
[4] I.W. Selesnick, Hilbert Transform Pairs of
Wavelet Bases, IEEE Signal Proc. Letters, vol. 8,
no. 6, pp. 170173, June 2001.
[5] R. Yu and H. Ozkaramanli, Hilbert Transform
Pairs of Orthogonal Wavelet Bases: Necessary and
Sufficient Conditions, IEEE Trans. Signal
Processing, vol. 53, no. 12, pp. 47234725, Dec.
2005.
[6] B. Dumitrescu, SDP Approximation of a
FractionalDelay and the Design of Dual-Tree
Complex Wavelet Transform, IEEE Trans. Signal
Proc.,2008, to appear.
[7] B. Dumitrescu, Positive trigonometric
polynomials and signal processing applications,
Springer, 2007.
[8] T.N. Davidson, Z.Q. Luo, and J.F. Sturm, Linear
Matrix Inequality Formulation of Spectral Mask
Constraints with Applications to FIR Filter Design,
IEEE Trans. Signal Proc., vol. 50, no. 11, pp. 2702
2715, Nov. 2002.
[9] B. Alkire and L. Vandenberghe, Convex
optimization problems involving finite
autocorrelation sequences, Math. Progr. ser. A, vol.
93, no. 3, pp. 331359, Dec. 2002.
[10] B. Dumitrescu, I. Tabus, and P. Stoica, On the
Parameterization of Positive Real Sequences and MA
Parameter Estimation, IEEE Trans. Signal Proc.,
vol. 49, no. 11, pp. 26302639, Nov. 2001.
[11] Y. Genin, Y. Hachez, Yu. Nesterov, and P. Van
Dooren, Optimization Problems over Positive
Pseudopolynomial Matrices, SIAM J. Matrix Anal.
Appl., vol. 25, no. 1, pp. 5779, Jan.2003.
[12] J.F. Sturm, Using SeDuMi 1.02, a Matlab
Toolbox for Optimization over Symmetric Cones,
Optimiza-tion Methods and Software, vol. 11, pp.
625653,1999.
[13] J. Gao, H. Sultan, J. Hu and W. W. Tung, De
noising nonlinear time series by adaptive filtering and
wavelet shrinkage: a comparison, IEEE Sig. Proc.
Let., vol. 17, no. 3, 2010, pp. 237-240.
[14] Fraser HSF, Blaya J. Implementing medical
information systems
in developing countries, what works and what
doesnt.
AMIA Annu Symp Proc 2010;2010:232236.
[15] Waegemann CP. mHealth: the next generation
of
telemedicine? Telemed J E Health 2010;16(1):2325.
[16] Li Q, Mark RG, Clifford GD. Robust heart rate
estimation
from multiple asynchronous noisy sources using
signal quality
1. indices and a Kalman filter. Physiological
measurement
2008;29(1):1532.
[17] Bishop CM. Pattern Recognition and Machine
Learning.
Springer Verlag, 2006.
[18] Monasterio V, Clifford GD. Robust apnoea
detection in the neonatal intensive care unit. Annals
of Biomedical Engineering 2011;In Submission.
Previous method
Using SQI
Proposed
method using
DDDWT
Specificity 0.7500 0.92
Sensitivity 0.8000 0.87
Accuracy 0.7931 0.82
Precision 0.9524 0.97
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
138

You might also like