Traffic Regulation Mechanism Using Heecp Protocol For D.N

CHAPTER 1
INTRODUCTION
TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 1

INTRODUCTION
1.1 OVERVIEW
Distributed network, happening earlier an insufficient year or current years the wireless tools
are developing gradually and become the unbelievable achievement. As previous relaxed
speed system have existing necessary to combine with the high quickness Systems. However
motivation behind to growing in Internet opportunity and users are possible to accepting
longer delay, for more packet damage and further routine poverty complications for of system
bottleneck. Firmly this problem was frustrated over network facility suppliers in expressions
of kind process of the continuing slow, which probably will respect such as an infeasible
description.
Computer Computer Computer

Node Node Node
Communication Network
Computer Computer
Node Node
Fig. 1.1: Distributed network/system
The network traffic of transportation connotation is very complex system to achieve and
effort owing to locality size, active nature, variations in demand and the large quantity of
heterogeneous mechanisms and devices elaborate from the group of electrical control to the
message and supply to the eventual customer.

New coming technologies such as accurate and well-timed documents scheduled the formal
of the power system increasing the possibilities to manage the system at more efficient and
responsive level and apply wider monitoring, control and protection schemes.
The problems on the networks capabilities and data traffic management are rapidly
increasing because new components and data flow. Now direction to provide well-organized
data communication and best resource usage, it is enough to classify and implement
mechanisms that can control the data traffic in a necessary routine. The suitable mechanism
for source allocation and traffic instruction is a Quality of Service, Quest implementation.
Now this technique dissimilar data flows will have held in reserve capacity on the
associations and different ordering stages without disturbing or disturbing other data flows.
The result of this would be higher reliability and accuracy in transporting brochures packages
especially under heavy network load [1].
There are three basic steps involved in implementing QoS on a network:
Identify types of traffic and their requirements: Learning the network to define the type of
traffic that is organization preceding the system then regulate the QoS necessities desirable
for the dissimilar forms of transportation.
Define traffic classes: these inactivity assemblies the traffic with similar QoS requests into
courses. For example, three classes of traffic might be defined as voice, mission-critical, and
best effort.
Define QoS policies: QoS policies meet QoS requests for intended each traffic class.
The main step in applying QoS towards identifies the traffic on network and formerly
determines QoS requirements and the importance the various traffic types. This phase
offers some high-level strategies for executing QoS in networks that maintenance for
multiple applications, delay-sensitive and bandwidth-intensive applications. These
uses may increase business processes, but bounce network resources. QoS can offer
safe, probable, quantifiable, and assured services to these requests by handling delay,
delay variation (jitter), bandwidth, and packet loss in the network.
Choose the QoS difficulties of users. Quantity the traffic on the network over and
done with congested steps. Performance CPU convention approximation on each of

the network guidelines during hard periods to govern where problems potency be
happening.
It stands a caring of computer network, in which the dishonourable information is
variation over unalike networks. This deliveries a certain communication network,
which is improved by every network equally or separately. It processes network
management software, which is answerable for achieve and display data routing,
combining and allocating network bandwidth, access control and added central
interacting procedures. For example server calculating network, where the server is
the manufacturer of resource and many remote users who are interrelated, are the
consumers who access the presentation from not the same networks. Separately from
mutual communication inside the network, a distributed network frequently also
distributes processing.
Using the increase of user made on distributed network scheme, traffic congestion is one of
necessary circumstances. Owed to vast network congestion user will face the network speed
problem, real congestion control methods are desirable and mainly to solve transportation of
todays for actual extraordinary speed networks.
We altogether are in good health known about with the congested streets, and maybe
congestion is occur in the other networks such as Internet, so it is noticeably important to
have a general understanding of how and why congestion happens in distributed networks.
On the other hand, design of flow of traffic in the network is the reputation and hard
communication between different users. For an example, in a road network we would usually
expect every one driver go to select the most appropriate route, and this best route will be be
governed by upon the delays since the car user assumes to originate over and done with on
different roads. This delays will be effort be resolute by on the changes of routes arranged by
others. As dependent on the others system we can face the delayed problems.
For example, Internet is progressively occupied using internet protocol and packet
switching; hence to growth of network performance in positions of pleasant facility to
customers is stately as inspiring tricky [2]. Currently, end of the Internet systems, controlled
admission mechanism tool are applied at transports layer. Due to delay complex environment
of software uses, the necessity functioned of significance origin for satisfying quality of
services control [3]. Network traffic prepared through the help of software requests as
identified by nature and as of accidental queuing in routers here we have planned of
occurrence overcome the problem of the delay agitations. Normally the use of congestion

control appliances, system routers are well-found through extension descent device taking
predetermined capability sequence. Once the server is exhausting, tail drop mechanism that
offers stations the incoming packets for the time being then upon sequence complete phase
has received packets are out subsequently. Separately from difficulty, the method may suffer
various complications i.e. lockout behaviour, overall organization and filled sequence [4].
The issue of filled sequence is the leading trick which can build time-consuming delay and
create this device an incorrect choice for real time uses.
By the current time, Congestion is fond of an excessive concern that increases on shared
networks when numerous operators attack for usage to the similar stuffs. Think through
contiguous freeway congestion. By way of additional numerous vehicles enter on freeway
when congestion becomes substandard. Eventually, the reaction can back up into track away,
means of transference from recognition on afterward to entirely. In packet-switched webs,
packets removable on out of the chunks and files of replacing processes as they pass
determined the system. In piece, a packet-switched system is continually confirmed to as a
"scheme of arrangements Congestion characteristically readings where many associations
provender into a single relation, such now place of where local area networks is
interconnected to the wide area network links. Congestion also enlargements at routers in
central networks where nodes that are exposed into further traffic than they are aimed to
handle. Transmission control protocol/internet protocol networks are precisely vulnerable to
congestion of their basic connection-less atmosphere. Approximately are no computer-
generated routes with convinced bandwidth. Packets are familiarized by several hosts at
certain time, and individuals packets are irregular in size, which diversity forecasting traffic
flow projects and provided that certain package unbearable. Though the fact is connectionless
systems have welfares, amount of service is not any of them.
1.2 MOTIVATION
Rate Control Mechanism, this might be reached by including self-adaptation and self-
learning capability of the internet structural design. Therefore we recommend a Rate control
Metric (RCM) based on stochastic principle that identify the traffic behaviour in a better way.
This metric provides rate control in communication between nodes and routers by register the
capability of the internet connection at that, fact of communication alleviate to the congestion
that occurs into the network.

Observing speed is unique for the critical movements of the determination of reducing the
response time of user as fit as supplying up the dynamic traffic burdens. The presently trained
methods of TCP grounded association are dedicated for particular application that organises
long live traffic flows. The rate of the traffic confirms regulatory is accomplished to the
minimum resultant to the visible enhancement for the multiple online users in distributed
network.
Therefore the rate controlling parameter in distributed network is developed to
incorporate following characteristics. Exponentially faster access to huge files as
related to conventional TCP based connection.
Would be organized on progress networking systems like visual network and should
accept high-bandwidth delay product. Have to certify stability from the network and
free from greatly dependencies on round trip time, flows, and link abilities. Must be
well-matched for integrating traffic-based strategies for certifying the observance to
congestion control schemes. Flexibility to network operators for incorporating
privileges to certain traffic flow.
Hence, it is anticipated the rate control metric should be observe the above mentioned
characteristics to ensure better congestion control strategies. The consideration for
guiding rate should approve the mechanism for assessing the rate control factor at the
end host in congestion layer that has climbed up owed to IP/TCP founded network.
The constraint for adjusting rate must also ensure execution contrast to the presently
existing transmission rate of link that has data packet impressed at the destination end.
The decreased value have transmission rate is efficient in the data packet and
promoted up from the next link. The methods guarantee to fairness in hold-up link
and can supply up the dynamic traffic requirements.
Heterogeneous Energy Efficient Congestion Aware Protocol (EECP) through this
method we can try to minimized the congestion from the future internet structural design and
recovers the quality of the internet network. This will also improve capacity of the network
over heterogeneous congestion aware protocol.
In the heterogeneous systems, outcome done expected routes that depend on the
minimum data rate of entire associations. Several packets are give directions to the route that
has minimum congestion from the network. Main objective is the heterogeneity network is, if
congestion is present into the network in that situation we can try to minimalize and remove
congestion from the specific routes and then afterward send packets from the same route. In

which we does not have need to bring up-to-date the route if congestion is existing in the
network. The protocol which used for depreciate accurate event detection, energy
effectiveness and less traffic.
EECP scheme divided into multiple channel capacity that is use according to our
requirement and many arrival packets are there that way we can utilize the entire capability
of the network as regards to our requirement they will clutch the data.
Node Configuration
Different Distance and

Energy Label
Energy Efficient Route

Chose
Total Capacity of the

Network = Total Arrival
Packets
Packets are Sending and

Traffic is Minimum
Fig. 1.2: Working of heterogeneous energy efficient congestion aware protocol.
1.3 OBJECTIVES
In this dissertation, we have aimed to quantify and improve the energy efficiency of various
heterogeneous network topologies in our environment. Calculate the delay, with respect to
time we are needed to wait for the free routes for send the packets. Furthermore maximize
throughput of the network, how much packets will have to receive from the receiver, and

minimize the real congestion transmission rate and channel capacity of the system in the
heterogeneous energy efficient protocol for distribute network.
The transmission delay and throughput would be less, so that main problem of the traffic
congestion of the distributed network to be resolute when congestion is reduced.
Heterogeneous energy efficient congestion aware protocol is proficient to minimizing
congestion to the overall increase performance of the network.
1.4 DISSERTATION ORGANIZATION

The dissertation is organized in the following manner:-
In Chapter 1, we have already discussed about the introduction of distributed network and
the congestion traffic in it, in general in terms of motivation and its objectives.
In Chapter 2, we will discuss the background information related to distributed network and
architecture of distributed network.
In Chapter 3, we will provide a literature review for this dissertation.
In Chapter 4, we will provide the description of the proposed work of the dissertation.
In Chapter 5, we will compare and analyze all simulation and outcome of the results.
In Chapter 4, the conclusion of the dissertation is presented.

CHAPTER 2
BACKGROUND

BACKGROUND
2.1 INTRODUCTION
In Information Technology (IT), the technological advancement is a continuous process and
Internet is the demonstration of this. Later ARPANET [5], the development of the internet
consumes appreciated multifold in rapidity and amount of hosts. By way of for all RFC 1958,
typical of continuous change is maybe the only standard in the Internet that should survive
intended for an infinite period. Developed design standards happening the design of the
upcoming internet are most discussed matter in the part of distributed networks. The constant
growth besides unlikely success of internet (distributed network) accommodations relaxed
detectives to devise new safety and security services, which could not be maintained by the
current Internet architecture. Such as for the entire AKARI scheme New Group
Organization Structural design AKARI Abstract Design [6], there will be heterogeneity of
nodes such as processors, stations, movable devises sensors and the measurement as large as
of 100 billion in size. Many establishment areas must remain scatty in the in performance
field of other Services, Transport, healthcare in persistent or universal way. On the same
period, consequently of the progress in video tools a larger amount of multimedia content
will also be smooth crossways the networks. Founded on the investigational remark complete
by Medeiros [7], it was create that present is massive volume of data being common among
32 millions of user, mainly from the Internet arranged audio-visual. This circumstance high
spot that resources accessible in Internet constantly raise its size with the active demands of
the huge capacity of online users. Management such data is quite a challenging mission as it
sincerely leads to worst traffic congestion. In future, that one is important congestion protocol
must be design allowing for such present experiments for the determination of increasing the
unavailability, dependability and interoperability of Internet-based facilities. Another author
Mahonen et al. [8] must also discussed about the rising traffic in Internet and recommended
that such congestion connected issue might be probably moderated by over dimensioning the
network. Parallel problems were also discoursed by Jacobson et al. [9] for the determination
of improving the network volume on the support system to provide up the developments
Internet-based services. After the current increasing evaluation lead by Pan et al. [10], it is
empirical that greatest of present exam folding bed for upcoming Internet construction
investigation in dissimilar nations are the significances of preceding research projects that are
not associated to future Internet architectures. Hence, one inconsistent question growths
established on the efficiency of the architecture as a regulatory factor to the design and

upgrading of future Internet architecture. For an example, as discussed in [11] the end-to-
end advices are inefficiently compelling to balance other criteria for functions such as routing
and congestion control. Preceding the further hand, the progression of the Internet
architecture is determined by incremental and reactive additions [12], slightly than by major
and proactive modifications. Moreover, the researchers have demonstrated that efficient
performance or functionality describe necessary but not appropriate conditions for change in
the Internet architecture (and/or its modules); hence, it demands to exhibit limits of the
current architecture [13]. Therefore, scientists and researchers from companies and research
institutes world-wide are employed towards appreciative these architectural limits thus as to
increasingly determine the principles that will enterprise this massive network to meet the
requirements in the existent scenario. We recommend a novel metric known as Rate Control
Metric (RCM), communicated through stochastic principles. The presentation of this metric
in assessment of effective congestion control in the convoluted Internet work system is
examined over simulation.
The situation has been observed that power saving will always keep on a problem to
be cracked. To answer this problem, current be necessary increasing interests in design for
wireless networks that relates on interactions between protocol stack of various layers. This is
promising solution approach of wireless sensor network called cross layer design.
Designed for network dynamics to avoid congestion in the IP framework, the

schedules without end-to-end bandwidth guarantees have to adapt efficiently to network
dynamics. Expending unicast and multicast (UDP) communications Network resources are
shared. By congestion control unicast flows, adaptation is enforced by TCP. For network
congestion, one approach is for all UDP flows to use a congestion control mechanism and
become approachable, other is to build into the network the capacity to handle unresponsive
flows effectively (e.g. by dropping packets). Both unicast and multicast best-effort flows are
addresses by a congestion avoidance mechanism, offering to all i.e. responsive or not
responsive flows involved in a congested router.
Rate adaptation process is established off the channel quality estimation. A cross-
layer design [10] explores the significant performance enhancement at MAC layer and
interference estimation of this layer for optimal routing selection at network layer.

2.2 NETWORK HISTORY
The following table shows a network history.
Table 2.1: Network history
Years Description
1966 ARPA packet switching experimentation
1969 First Arpanet nodes operational
1972 Distributed e-mail invented
1973 Non-U.S. computers linked to Arpanet
1975 Arpanet transitioned to Defense Communication Agency
1980 TCP/IP experimentation began
1981 New host added every 20 days
1983 TCP/IP switchover completed
1986 NSFnet backbone created
1990 Arpanet retired
1991 Gopher introduced
1991 WWW invented
1992 Mosaic backbone privatized
1995 Internet backbone privatized
1996 OS-3 (155 Mbps) backbone built
2.3 HISTORY OF DISTRIBUTING COMPUTING
Todays usage of distributed computing is complete likely through the huge network of
personal computers and workplaces that mark up the Internet and through help of the Internet
we can itself established the network very quickly and simply in the previous few periods.
The implements are temperately new and independent foundational to be commonly used in
the network. The idea of distributing resources inside a computer network is not novel. This
is first in progress with the usage of data access terminals on processor computers, formerly
moved into minicomputers. It is now probable in particular computers and client-server
architecture with more tiers. In which explained the working of different modules, named as,
ARPANET, The internet And last one is broadcast.

2.3.1 The ARPANET
The ARPANET stand for Advanced Research Projects Agency was generated in 1958 from
the computers networks. The highest distributed computing programs existed a pair of
programs called Creeper and Reaper which completes way ended the nodes of the
ARPANET in the 1970. The Creeper come from first and consuming of the idle CPU cycles
of computers. It generated the copy itself on the next system and then deletes them from the
earlier one. It was improved to continue on all earlier computers and the Reaper was created
which travelled from end to end the same network and deleted all duplicates of the
Creeper. The nodes would be mainframes in its place of the telephone switches recycled at
the time, so that they would be smart to decide the greatest route for transfer each message.
2.3.2 The Internet
Such as the network developed and other packet switching networks established in other
nation state, scientists initiated to search models for concerning different types of networks,
or internetworking. The Internet was born and produced from present, finally making
ARPANET by the late 1980s so that it was old in 1990. By the 1980s the Internet was static
only a fairly small group of networks for the maximum part related to military enquiry and
procedures. On the other hand, the field of computer science was increasing in institution of
higher education that required the network facilities that the ARPA sites adored. The arrival
of small, expensive personal computers in the late 1970s and particularly in 1981 with
International Business Machines (IBM) PC made number of home and office users who also
needed admission to the internet.
2.3.3 Broadband
Broadband Internet facility actually is the best used procedure of Internet access since of its
great access speeds. By way of more and more users give to Internet Service Providers, the
request improved for high-speed entrance in home environment and big business. Maximum
users were associated by narrowband analog telephone lines and a dial-up process to usage of
the mobile system. In the 1990s, the procedures of broadband service derived from
businesses that already stayed on-line to best homes and businesses in the nation state.
Broadband packages originate in all forms and dimensions, from digital Subscriber Line
(DSL) broadband to cable broadband and 3G and 4G mobile broadband.

2.4 ARCHITECTURE OF DISTRIBUTED SYSTEM/NETWORK
There are different types of distributed system architectures. That is Client Server
architecture, Layered application architecture and Distributed object architectures.
2.4.1 Client Server architecture:

It is the most important architecture of the distributed system. It is the combination of client
and server. Client sends a request from the server, and severs receive the request then send
the data to the client.
Server process
Client process
C1 C2 C3 C4
22
S1 S2
Fig. 2.1: client server architecture
2.4.2 Layered application architecture:

In the layered application architecture there is three types of layers.
Presentation layer: Concerned with presenting the results of a computation to system
users and users and with collecting user inputs. Presentation layer takings care that the data is
sent in such a way that the receiver will know the information and will be capable to use the
data.
Application processing layer: It is the top most layer of OSI Model. Operation of data in
several ways is done in this layer which supports user or software to get right to use to the
network. Concerned with providing application specific functionality e.g., in banking, in a
banking system, banking functions such as open account, and close account.
Data management layer: This layer provides error-free transfer of data frames from one node
to another over the physical layer, allowing layers above it to assume virtually error-free
transmission over the link. Concerned with the management of system database.

Presentation Layer
Application Processing Layer
Data Management Layer
Fig. 2.2: Layered system architecture
2.4.3 Distributed object architecture:-

Distributed object architectures are additional complex to design than client server
systems. There is no difference in distributed object architectures between clients and
servers. Each distributable entity is an object that provides services to other objects
and receives services from other objects. Object communication is through a
middleware system called an object request broker, request broker.
O1 O2 O3
S(O1) S(O2) S(O3)
Object request broker
O4 O5
S(O4) S(O5)
Fig. 2.3: Distributed object architecture

2.5 CONGESTION CONTROL
Congestion is a state that occurring in a network layer. A network is combination of different
types of networks, such as the Internet. Bottleneck is situations in Communicated Networks
in which lots of packets are existing in a part of the subnet Congestion occurs when packets
send from the sender is exceed the rate from its receivers. When one part of the subnet is
becomes overloaded. Congestion causes packets to be dropped on the network due to buffer
overflow and therefore leads to data loss and unreliable connections.
Due to occurring at intervals in short nature of the network packet traffic there may be
conditions where there is overflow of the queues.
Maximum carrying Perfect
Capacity of subnet
Desirable
Packet delivered
Congested
Packet send
Fig. 2.4: Basic concept of congestion

As a situation there will be re-transmission of several packets from continuously, this further
increases the network traffic. This finally leads to congestion.
Congestion Control is concerned with efficiently using a network at high load.
Several techniques can be employed. These include:
Warning bit, Choke packets, Load shedding, random early discard, Traffic shaping,
Congestion is one of the vital problems of the distributed network (Internet).
In order to improve the performance of internet (distributed network). In this thesis

first we analyses the congestion control mechanism in distributed network then study

the effect on performance of distributed network using heterogeneous energy efficient
congestion aware protocol.
Traffic congestion is one of the compulsory environments in computer network.
Congestion is a problem that occurs on shared network when multiple clients access
to the same resources.
2.6 CONGESTION CONTROL VS. FLOW CONTROL
Congestion control- congestion control is an order to control the entering data in the
network itself. It is a global issue involves every router and host within the subnet. It is router
to router mechanism. Router discards packets when it cannot serve. Sender retransmits
acknowledgment until sender does not receive. Monitor system to detect when and where
places congestion can occurs.
Flow control flow control is receiver side methods that handle the speed of
transmitting data at the receiver side. Its a point to point mechanism that controls the flow of
data between of just sender and receiver. The speed of sender is very fast in the flow control.
2.7 CONGESTION CONTROL PARADIGM
Flow types is help to control the real congestion may include a single packet, few packets, a
large number of packets, which involve light control, medium level control, and tight control,
individually. At what time a large number of nodes transmit information; their flows will
cross at intermediate nodes. This high number of bases increases the congestion But helps to
improving the consistency.
2.7.1 Contention-based Congestion:
when many nodes and packets try to send data concurrently to one another within the range at
that time loss is occur within the media due to many ways and also packet are lost. In
between packets are send from the simultaneously manner then we can say packets rate are
independent across the all nodes. [14]
2.7.2 Buffer-based Congestion:
Every single node uses a buffer for the packets for the future to be sent. The overflow of this
buffer reasons congestion and packets loss. This is due to high writing rate that differs in time

due to dynamic channel environments. The situation is presented that when using large buffer
sizes, the network load increase radically damages the event reliability, due to the partial
capacity of the mutual wireless medium. Designed for low buffer size standards, buffer
overflows leading to a larger number of packet damages but result in lower channel argument
and lower end-to-end packet potential values related to those values of greater buffer sizes.
2.8 CHARACTERISTICS OF DISTRIBUTED NETWORKING
Some key characteristics of Distributed Networking are: Concurrency, Scalability,

Openness, and Transparency etc.
2.8.1 Concurrency: Concurrency is the trend for effects to take place at the similar time in
a system. It can be referred as many users are using the same resource at the same time in
a distributed network. And many servers are responding to these requests. As soon as we
are assignment software to display and controller real-world systems, we are necessity
deal with this anticipated concurrency.
2.8.2 Scalability: Scalability is the ability of the system to smoothly handle the growing
amount of work load. Scalability in the network can be achieved by applying more than
one host in the network. It can refer to various different parameters of the system: how
much additional traffic can it handle.
2.8.3 Openness: Openness denotes that the network can be easily extendable and can be
easily modified by integrating new components. The necessity for further people with
fewer training to advantage from using geospatial data in more uses: That is, the necessity
to control reserves in software and data.
2.8.4 Transparency: Network transparency is the procedure of transfer or retrieving data

above a network in such a way that the information is not observable to users interactive
with a local or remote host, and system. If single computer system is there than present
itself to in front of users and application. The remote source retrieved by a user is
presented on another network and takes several intermediary networks earlier getting the
particular data/resource. On the other hand, customers only exist with the local or host
network; all other networks are unseen.

Table 2.2: Transparency of distributed network
Transparency of distributed system Description

Access Hides the way in which resources are
accessed and the differences in data platform.
Location Hides where resources are located.
Technology Hides different technologies such as
programming language and O.S from user.
Migration/Relocation Hides resources that may be moved to
another location which are in use.
Replication Hides resources that may be copied at several
locations.
Concurrency Hide resources that may be shared with other
users.
Failure Hides failure and recovery of resources from
user.
Persistence Hides whether a resource (software) is in
memory or disk.
2.9 CONGESTION DETECTION STRATEGIES

Many congestion detection mechanisms are used. The most used are: Packet loss, Queue
length, Wireless Channel Load, Buffer occupancy, and Packet Transmission Time Metrics.
In several circumstances, a single factor cannot specify congestion perfectly. [14]
2.9.1 Packet Loss

Packet loss in the network can be measured by the sender side. Its increase the traffic on the
network and network performance is maximized. This is become errors on the wireless
network.
2.9.2 Queue Length

When multiple packets are lost in the media, and when packets are dropped continuously, we
can try to occupy packets through retransmission of that packet and removed.
2.9.3 Wireless Channel Load

This is one method of the congestion detection; when multiple and simultaneously packets
are arrive in the network then packet load increased and take actions when the time structure
for the transmission of a distinct packets, outdoes some predefined thresholds. [15]

2.9.4 Buffer Occupancy
Buffer occupancy term led the congestion detection method, if buffer is full occupied then we
can increased the length of the buffer according to our need. It is process of controlling the
buffer data.
2.9.5 Transmission Time Metrics

This is one of the congestion detection methods for the network. The transmission time
metrics is the combination of packet service time and packet inter-arrival time.
2.10 CONGESTION CONTROL ALGORITHIM

Congestion is a problem that arises in the network. In which different types of algorithms that
are Leaky bucket algorithm and Token bucket algorithm.
2.10.1 The Leaky Bucket Algorithm
Leaky Bucket Algorithm is the most commonly used controlling mechanism of Bucket has
identified leak rate for average constricted rate of Bucket has recognized depth to provide
somewhere to stay variations in arrival rate of Incoming packet is conforming if it does not
outcome in overflow Leaky Bucket algorithm can be rummage-sale to law enforcement
agency coming rate of a package stream.
The Leaky Bucket Algorithm a buffering mechanism is presented between the host
computers used to control rate and the direction to regulate the flow of traffic in a network.
This algorithm is executed such as a single-server queue using continuous service time. If the
buffer overflows then packets are rejected. In case of buffer overflow packets are thrown
away from the network.
Assume we have a bucket in which we are driving water in a random order but we have to get
water in an immovable rate, for this we will make a hole at the bottommost of the bucket. It
will confirm that water upcoming out in some fixed rate. And also if bucket will full we will
stopover pouring in it.
A simple leaky bucket algorithm can be executed using first in first out order (FIFO) queue.
A FIFO queue controls the packets. The procedure eliminates a static number of packets from
the queue at every signal of the clock. If the traffic includes of variable-length packages, the
secure manufacture rate must be based on the amount of bytes.

Host
Computer
Packet
Unregulated
flow
Interface
containing a Bucket The bucket
leaky bucket holds packets
Regulated
flow
Network
Fig. 2.5: Leaky bucket with packets.
The leaky bucket implements a constant output rate or average rate though of the burstiness
of the input. The algorithm make ensures of unknown while input is idle.
The host inserts one packet per clock on the network. These outcomes in a constant
flow of packets, and reducing congestion.
While the packets are the same size, the one packet per moment is acceptable. For variable
length packets, it is better to allow a fixed number of bytes per pulse. As a Example1024
bytes per pulse will agree one 1024-byte packet or two 512-byte packets or four 256-byte
packets on 1 pulse.

2.10.2 Token Bucket Algorithm
The Token Bucket Algorithm production amount is varying relating to the leaky
bucket algorithm.
The token bucket algorithm in which every bucket hold the token to resend a packet
and every packet contain the one token to enter into the bucket. If packet is lost then
token is already destroy.
Every t sec Tokens are created by a clock at the rate of one token.
The Idle hosts hold the tokens up to the maximum Size of the bucket larger bursts
later in order to show into the network.
Host computer
Tokens
The bucket
One token is Bucket holds tokens
added to the
bucket every t
Network
Fig. 2.6: Token bucket

Leaky Bucket vs. Token Bucket
Leaky bucket removes the packets and Token bucket does not remove the packets
while its remove the token.
In token bucket, if they have multiple token to cover its length because of that we can
only transmitted.
The token bucket does not permit saving the packets and leaky bucket they are
allowed to save the tokens.
Leaky bucket sends packets at a normal rate, but in token bucket packets are send
sources to sink from the faster speed.
2.11 SYSTEM MODEL
2.11.1 Size of Congestion window
Boundary of Sender-side on the measure of data sender can formerly receive an

acknowledgment from the client. The cwn variable is not promoted or exchanged among the
sender and receiver in this condition, mutable preserved through the server in London.
Further, a new rule is introduced: the maximum amount of data in flight among the client and
the server is the least of the rwnd and cwnd variables.
The explanation is start slow and to raise the window size as the packets are approved: slow-
start initially, the cwnd start cost was normal to 1 network section; RFC 2581 effective this
cost to 4 segments in April 1999; most newly the value was amplified when more to 10
sections through remote function call 6928 in April 2013.
The determined quantity of data in route for a novel transmission control protocol assembly is
the minimum of the rwnd and cwnd cost; a modern server can send up to 10 network
segments to the user, at which point it must stop and wait for an acknowledgment. At that
time, for each established ack, the slow-start process displays that the server can increase its
cwnd window size by one fragment for each acknowledge packet, two new packets can be
main.

This phase of the broadcast control protocol connection is frequently acknowledged as the
"exponential growth" process (Fig. 2.7), as the client and the server are challenging to
promptly meet on the available bandwidth on the network path between them.
Fig. 2.7:- Congestion control and congestion avoidance [36]
Thus it is slow-start a significant issue to save in mind when we are building applications for
the browser. Well, hypertext transmission protocol and numerous other application protocols
run over transmission control protocol, and no substance the presented bandwidth, each
transmission control protocol assembly must go decided the slow-start phase. We cannot use
the whole capability of the link instantly.
As an alternative, It start with a minor congestion window and binary it for every cycle
i.e., exponential growth. As significance, the time is required to range a precise target
function of both the round trip time among and the client and server. The initial congestion
window size [16].
Reaching time for cwnd size of size N

Table 2.3: Contents for the window size
CONTENT FOR WINDOW SIZE VALUES

Client and server obtain spaces 65,535 bytes(64 KB)
First congestion window 10 sections (RFC 6928)
Round trip time 56 ms (London to new york)
In spite of the 64 KB obtain window magnitude; the output of an original transmission

control protocol assembly is firstly restricted through the size of the congestion window. In
circumstance, to extent the 64 KB accept window boundary, we will first necessity to develop
the congestion window size to 45 sections, which will take 168 milliseconds:
65,535 bytes /1,460 bytes 45 segments
56 ms x [log2 (45/10)]=168ms
Thats three round trips (Fig.10) to reach 64 KB of throughput among the client and server!
The fact that the client and server may be accomplished of relocating at Mbps+ data rates has
no consequence when a new connection is recognized.
Fig. 2.8: Congestion window size growth [36]
To drop the sum of time it proceeds to produce the congestion window, we can decrease the
round trip time among the user and receiver, or we can growth the preliminary congestion
window size to the new RFC 6928 value of 10 sections.

2.12 QUALITY OF SYSTEM (QOS)
Before converged networks were common, network engineering focused on connectivity. The
rates at which data came onto the network resulted in bursty data flows. In a traditional
network, data, arriving in packets, tries to acquire and use as much bandwidth as possible at
any given time. Access to bandwidth is on a first-come, first-served (FIFO) basis. The data
rate available to any one user varies depending on the number of users accessing the network
at that time. Protocols in no converged traditional networks handle the bursty nature of data
networks. Data networks can survive brief outages. For example, when you retrieve e-mail, a
delay of a few seconds is generally not noticeable. A delay of minutes is annoying, but not
serious. Traditional networks also had requirements for applications such as data, video, and
systems network architecture (SNA). Since each application has different traffic
characteristics and requirements, network designers deployed non-integrated networks. These
non-integrated networks carried specific types of traffic: data network, SNA network, voice
network, and video network.
A converged network carries voice, video, and data traffic. These flows use the same
network facilities. Merging these different traffic streams with dramatically differing
requirements can lead to a number of problems. Key among these problems is that
voice and video traffic is very time-sensitive and must get priority.
In a converged network, constant, small-packet voice flows compete with bursty data
flows. Although the packets carrying voice traffic on a converged network are
typically very small, the packets cannot tolerate delay and delay variation as they
traverse the network. When delay and delay variations occur, voices break up and
words become incomprehensible
2.13 CONGESTION-AVOIDANCE TECHNIQUES
In congestion avoidance, monitoring of network traffic do effort to forecast and ignore

congestion at mutual network and internetwork traffic jam before congestion converts in a
big problem. These techniques provide special action for best traffic once there is congestion
while simultaneously exploiting network quantity & capability use and reducing delay and
packet loss. Routers capacity also drops packets for other less mutual explanations:

Input sequence drop: In this leading control handling unit is demanding and cannot
develop packets.
Ignore: The router rings out of barrier space.
Overrun: The control processing unit is busy and cannot allocate a free buffer to the
novel packet.
Frame errors: The hardware detects an error in a frame; for example, cyclic
redundancy checks, runt, and massive.
Packet loss is frequently the outcome of congestion on a boundary. Maximum

submissions which use transmission controller protocol knowledge go-slow as broadcast
control protocol mechanically regulates to network congestion. Released transmission
control protocol sections reason TCP assemblies to decrease their window sizes. Certain
submissions do not use TCP and cannot tackle drops.
Approaches for prevent drops in complex uses:
Growth connection ability to security or avoid congestion.

Assurance sufficient bandwidth and growth defense space to put up bursts of traffic
from delicate movements. Here numerous tools presented in Cisco input output
scheme excellence of system software that cans assurance bandwidth and gives
selected promoting to drop-sensitive uses.
To prevent congestion through reducing poorer importance packets formerly
congestion occurs. Cisco value of system delivers queuing mechanisms that start
dropping lower-priority packets before congestion follows.
Cicco IOS QoS software gives the subsequent mechanisms to avoid congestion:
Traffic controlling: Traffic regulating broadcasts bursts. Once the traffic amount
ranges the arranged extreme rate, added traffic is released. The outcome is a
production rate that seems as a scored with apices and troughs.
Traffic shaping: In compare to policing, traffic shaping preserves spare packets in a
sequence and formerly programs the additional for advanced broadcast over increases
of time. The consequence of traffic determining is a round packet output rate.
There are three basic steps involved in implementing quality of system on a network:
Identify types of traffic and their requirements:

Study a network to control the type of traffic that is running on the network and then
control the quality of system requirements desirable for the different kinds of traffic.
Describe traffic programs: The action traffic collection with similar excellence of
system necessities into modules. For instance, 3 classes of traffic strength are definite
as mission-critical, voice and best power.
Define quality of system policies: quality of system strategies come across with
feature of scheme requirements for each traffic class.
2.14 TCP FACILITIES

2.14.1 Stream Data Transfer
TCP allocations a constant stream of bytes. TCP combination of bytes in TCP
segments, which are delivered to Internet Protocol for communication to the target.
TCP themself decides how to fragment the data and it may onward the data at its own
accessibility.
2.14.2 Reliability
TCP gives a sequence number to each byte communicated, TCP uses the sequence to
reorganise the sequence when segments are out of direction and assumes a positive
acknowledgment (ACK) from the sending TCP and to remove duplicate segments. If
the ACK is not received inside a timeout interval, the data is retransmitted.
2.14.3 Flow Control

Flow control shows when acknowledgment is sending that are received to the sender
side without affecting overflow in its inner buffers. This type of acknowledgment are
receives from the many times.
2.14.4 Multiplexing
In multiplexing transmission control protocol effort as a simultaneously when multiple
developers work within the one host. Its provides a fixed statements or ports within
every host in the network.

2.15 TCP CONGESTION CONTROL
TCP is unique portion of two well-known protocol criteria commonly denoted to as

TCP/IP. TCP stands on top of the IP layer and authorizations segments onto the IP layer for
extra processing. These sections are then approved onto the lower level layers and
ultimately onto the network [16]. The term of Transmission control protocol combination of
the TCP protocol using TCP must establish a TCP connection with each other and we can
transfer and exchange the data. And second one is IP protocols. It provides a connection
oriented, reliable, byte stream service. It is a full duplex protocol, meaning that each TCP
connection supports a pair of byte streams, one flowing in each direction. TCP includes a
flow-control mechanism for each of these byte streams that allow the receiver to limit how
much data the sender can transmit. TCP also implements a congestion-control mechanism.
2.16 CONGESTION CONTROL METHOD IN TCP

2.16.1 Slow start and congestion avoidance
In Slow start and congestion avoidance, when congestion occurs in the transmission
protocol its slow down the rate of transmission of packets into the network. Slow start and
congestion avoidance both methods are implemented together.
Two variables need the suggestion of the transmission control protocol.
Congestion window (cwnd)
Slow start threshold (ssthresh)
Slow start method is continuing until the congestion is occurring in the network. Congestion
occurs in the network through when timeout of a process, duplicate acknowledgment is
occurs and these times half of the congestion window saved in slow start threshold, in time
congestion avoidance proceeds place. When congestion window is less than aur equal to the
slow start threshold, then TCP is in slow start otherwise congestion avoidance take place.
2.16.2 Fast retransmit and fast recovery
TCP can produce a duplicate acknowledgment when an out- of-order segment is established.
In that case we tell about which number of segment have received and mostly segments are

received from out- of -orders. This duplicate acknowledgment outcome by the loss of packets
in the network.
In Fast recovery if packets or segments are lost in the network due to the duplicate
acknowledgment then retransmission of segments is done. When a new segment is received,
then duplicate segment has left the network and placed in the receiver's buffer.
Fast retransmit and recovery is congestion control method that is possible to rapidly recover
lost data packets. The TCP usages a timer if packet is lost at that time we retransmit the
packet. In a time out duration dont need to send retransmission of packets. If a receiver
acquires a data segment that is out of order, it immediately leads a duplicate
acknowledgement to the sender. If the sender obtains three duplicate acknowledgements, it
undertakes that the data segment shown through the acknowledgements is lost and directly
retransmits the lost segment.
2.16.3 Explicit Congestion Notification (ECN)
ECN feature provide a method to intermediary route is delivered the segment end to end
point on the TCP/IP based network. It is a non-compulsory feature that might be used among
two ECN acceptable endpoints. The benefit of this feature is the drop of delay and packet loss
in data communications.
ECN essential be present acceptable on both endpoints and all intermediately routes one end
to other ends are work properly. Explicit congestion notification leads into by designs the two
least-significant bits in the differentiated services (DiffServ) field of the Internet Protocol
header.
Table 2.4: Explicit congestion notification bits.
ECN Bits (code) Meaning

00 Nin-ECT Packet is marked as not ECN-capable
01 ECN(1)- End points of the transport protocol are ECN capable
10 ECN(0)- End points of the transport protocol are ECN capable
11 CE Congestion experienced

2.16.4 Selective acknowledgment
Selective acknowledgment is the selective retransmission method to reduce the limitation. It

drops the multiple segments in the network. In which only some selected acknowledgment is
retransmitted that are missing in the network. Receiver can notify the sender all segments that
have reached successfully, therefore the sender need retransmit only those segments that have
really been lost.

CHAPTER 3
LITERATURE REVIEW

LITRATURE REVIEW
Generally network congestion occurs once, the collective demand exceed by the greater than
as associated to the manageable ability of the properties.
Congestion and safety attacks are mutual occurrences in source controlled wireless
sensor networks, especially for distributed network, somewhere a large capacity of high bit-
rate multimedia data needs to be managed by the network. Trust-based bottleneck aware
routing in WSNs is a novel investigation subject has not lectured in writings to an unlimited
scope. T-LEACH [17] is the enhanced type of the generally known data-gathering procedure,
low energy adaptive clustering hierarchy [18], which reduces the no. of cluster bean
collections and therefore covers the generation of system, likened to that of other comparable
procedures, but the situation does not takings trust & congestion into explanation. TRANS
[19] defines the set of routing protocols prepared with trust organization. In the fuzzy c-
means clustering protocol [20], Zarei et al. have recommended a fuzzy logic-based trust
approximation system for congestion controller in wireless sensor networks. Fairness
Congestion Control on behalf of dis Trustful protocol [21] is essentially an adaptation of
FCC, in which the Threshold Trust rate is used for choice making. Our previous work, which
shows a major improvement completed the Fairness Congestion Control for a dis Trustful
procedure, is considered in TFCC [22], in which road traffic movement from the source to
sink is improved by adaptive data-rate control and data routing takes place through asset of
the Link State Routing Protocol.
Yedavalli [23], introduced wireless network leads used for congestion control and
Delay acceptance in the wireless sensor networks. Queue is implemented in the wireless
sensor network using rate models.
Reddy and Reddy [24], introduced for service rate for detect the congestion from the
multihop network. Dependency of the heterogeneous protocol is also required for detect the
congestion. Multipath direction-finding protocol always objectives to complete advanced
dependability in the network.
Sheng, et al [25], introduced various type of Technologies that are support to our
everyday life this is called Internet-of-Things is becoming more important as the need for
better know our environments and make them smart increases. It is estimated that smart

devices and systems, for example wireless sensor networks will not be isolated but associated
and combined computer networks. Thus, it becomes censoriously important to learning how
the modern methodologies to regulation in this area can be improved.
Chakravarthi and Gomathy [26], this paper focused off the congestion due to
concurrent transmission. For control and identify the congestion we are uses the message
authentication code algorithm. Thus a problem of identifying and monitoring congestion
becomes critical to increase the performance of the network. Congestion plays a smallest
character in decreasing the presentation of wireless sensor network.
Rathod and Patel [27], introduced routing aware to mobile ad-hoc network and
network congestion. Which we are finds the minimum route from a source to a destination
and we does not consider the current node.
Sheeja and Pujeri [28], introduced the nodes are in the unstructured from, nodes are
assigned at the randomly. Due to numbers of packets congestion is occurs in the network.
Sergiou et al. proposed a unique scheme that beginnings its operation on a mutual tree
that begins from the sink. We can use most suitable route. It uses the soft stage technique to
minimizing the congestion and it uses only first time. Introduced the hard-stage scheme
which attempts to force data flows to change their direction-finding path in instruction to
avoid the congestion [29].
Jaewon et al. implemented a protocol that detects congestion using the buffer
occupancy and channel load. It applications on the extra network resource adaptation and
alleviating connection hot spots in case of congestion. In this case, the traffic upcoming from
the hotspot between the original path and the indirect route path the distributor distribute the
flows while the merger combines the two flows. TARA cannot use for large-scale sensor
networks because it needs the knowledge of whole the network topology [30].
Lee and Chung [31], implemented a protocol that detects the congestion using both
traffic control methodology by reducing the packet transmission rate of sending node and
resource control line of attack by increasing the packet response rate of receiving node. It is
energy efficient congestion control scheme based on medium access control protocol which is
employed over a duty cycle for congestion avoidance in the wireless sensor networks. If the
congestion degree is overhead threshold, it informs child nodes to correct the transmission
rates and if the congestion degree is under a certain threshold, regulates its own duty cycle to
decrease congestion. It is completed based on congestion degree.

Misra et al. proposed a protocol which uses automation in regulating the incoming rate on the
nodes possibly keeping data packet arrival rate and data packet service rate equal in all
intermediates nodes. When congestion is detected LACAS uses the intermediate node to rely
the some part of the congested traffic and number of packets dropped is measured through
the optimal action. The authors made an important feature of LACAS which learns from the
previous behaviors of the nodes and controls congestion efficiently [32].
Chengdi [33], purposed to methodically adjust the present Internet congestion control
to the wireless environment at a insignificant deployment cost and interruption. Our
variations include TCP and AQM improvements, which equally help the heterogeneous
wired/wireless Internet to work in the best region of low delay and the high throughput.
Earlier the Transmission Control Protocol, we have increasing the STS model those usages a
sequential timer creation to proposal more reliable signs of packet loss and network
congestion. We have devised TCP-NCL as a practical approximation of STS, which can
serve as a unified solution for effective congestion control, sequencing control.
Mohammadizadeh [34], the objective of this research is to establish an end-to-end
cooperative congestion control algorithm in a heterogeneous wireless environment with the
Internet backbone. The end stations are prepared with multiple radio interfaces for connection
to the Internet backbone and multiple paths can be established between the end terminals for
simultaneous packet transmission. We consider cooperation among the heterogeneous
wireless networks with overlapped coverage areas as a congestion prevention method without
losing service quality. The cooperation among available networks is used for enhance QoS,
especially in the congestion scenario. The cooperative congestion control algorithm should
avoid congestion from every path created on the condition of all paths in the association.
Therefore, when congestion is predicted for one path, other paths are notified to help the
congested path by moving the traffic load away of the congested path and by increasing CW
size of non-congested paths to compensate for the CW size reduction of the congested one.
However, cooperation among heterogeneous wireless networks in congestion scenarios does
not always take place as an uncongested network might not be presented for collaboration or
cooperation may not be beneficial due to overhead such as cooperation setup time.
Landstrom [35], TCP-like features several of the numerous TCP improvements that
have been proposed during the last decade. While being more aggressive than TCP Reno, it is
reasonably fair to TCP SACK at loss rates observed from the Internet today. Simulations
show that at the shorter round trip times - up to approximately 100ms - the differences
between TCP-like and TCP SACK is more pronounced. If in addition the loss rate is high, not

having a minimum transmission timeout gives TCP-like better performance. TCP-like also
recovers faster from severe congestion as its lack of reliability makes it less dependent on
individual packets. In the future we would like to explore the delay variations generated by
the two algorithms observed from an application point of view. Also, TCP-like attempts to
regulate the acknowledgment pace when losses are detected on the return path. Features of
the numerous TCP improvements that makes sure during the last decade. While actuality
more aggressive than TCP Reno, it is reasonably fair to TCP SACK at loss rates observed
from the help of internet today. Simulations show that at shorter round trip times - up to
approximately 100ms - the differences between TCP-like and TCP SACK is more
pronounced. If addition, the loss rate is high, not having a minimum transmission timeout
gives TCP-like better performance. TCP-like also recovers faster from severe congestion as
its lack of reliability makes it less dependent on individual packets. In the future we would
like to explore the delay variations generated by the two algorithms observed from an
application point of view. Also, TCP-like attempts to regulate the acknowledgment pace
when losses are detected on the return path.
Roy [36], a variation in TCP Reno and its use in Explicit rate adjustment (ERA)
algorithm. In TCP Reno, sender adjusts its cwnd size by uncritically observing the network.
However, in improved TCP sender acquires information about network capacity from
receiver side. Receiver evaluations network bandwidth by packet pair (PP) method. ERA
algorithm is studied by using modified TCP algorithm throughput as TCP friendly rate for
estimating target rate of receivers. The throughput of improved TCP was divided analytically
and then it was matched with the systematically estimated throughput of TCP Reno. The
comparison was made in terms of average number of packet sent in one hour data transfer.

CHAPTER 4
PROPOSED WORK

PROPOSED WORK
4.1 INTRODUCTION
As the most stimulating concern in designing distributed network there is limitation from the
congestion free routes, many research efforts aim at decreasing the congestion from the
network. Consuming less congestion occurs is a primary objective in designing distributed
computing. In the proposed work we are using the heterogeneous energy efficient congestion
aware protocol for remove the congestion from the route that which we want to sends the
packets.
4.2 EXISTANCE IDEA
In order to minimize the congestion from the distributed network (Rangaswamy and
Krishnareddy(2015)[1]) proposed a traffic regulation mechanism in which they compare the
result with congestion control mechanism as defined in Cicco et al.[1] and Rate control
mechanism in term of performance analysis of throughput , bandwidth utilization and delay.
The basic terminology and short description of the existence work is as follows.
4.2.1 CICCO
CICCO et al. estimated to extent GCC streams are accomplished to follow the accessible
bandwidth, although minimizing queuing delays, and equally share the bottleneck with other
GCC or transmission control protocol flow.
GCC algorithm in runs above the user datagram protocol and it recapitulates the
audio/video frames in RTP packets. It has been executed in the open-source WebRTC that is
available in the most recent versions of the web browser Google Chrome. The google
congestion control mechanism as estimated when it flow accesses the bottleneck in
separation, this algorithm is not able to withdraw fair bandwidth utilization and flow shares
the bottleneck with transmission control protocol flow.
Recently, IETF and W3C have recognized two joint functioning groups:
1) IETF RTCWeb aims at regulating a set of protocols such as congestion control algorithm
to transport real-time flows.
2) WebRTC aims at regulating a set of HTML5 APIs to enable real-time communication
within Internet browsers.

The congestion control algorithm proposed by Google within the RTCWeb IETF WG, which
has already been implemented in the Google Chrome and Firefox browsers.
The congestion control algorithm is incapable to fully utilize the available bandwidth, fairly
share the bottleneck bandwidth with synchronised flows, and contain queuing delays.
4.2.2 Rate control mechanism(RCM)

The Rate control metric (RCM) is supposed to be incorporated in router, which will have the
capability of computing fairness in rate and time-stamped it in the headers of the control
message. The system maintains mean RTT for outbound traffic on each interface, which
performs computation of rate in every control interval. The system captures the parameters
like Traf_RateAgg and mean RTT, which are encapsulated in each control interval for the
purpose of rate evaluation.
We recommend a new metric identified as RCM, conveyed concluded stochastic
values. The routine of this metric in observation of active congestion control in the complex
Internet work scheme is evaluated complete model.
Seeing Lcap as ability of the message connection and Traf_RateAgg as projected

combined involvement traffic rate through the last appraises intermission, the experiential
remainder defines of network can be considered as:
Remnant Buffer =Lcap-Traf_RateAgg (1)
The above eq. is calculated (1) that may differ permitting to the category of network.
Similarly, recommended scheme is explored by experiment based work, a constancy
constraint is including for best correctness (). Consequently, eq. 1 develops:
Remnant Buffer = (LcapTraf_RateAgg) (2)
The accept or connects the congestion rate to the source by Iupdate factor wherever it
appraises the intermission frequently.
Remnant Buffer =R=Iupdate/Mavg( (LcapTraf_RateAgg)) (3)
The new attribute is reached built on sequence magnitude and Mavg.
NA for end host = NA= .Sequencesize/Mavg (4)
Now recommended novel metric rate control mechanism is developed allowing for
the current projected link capacity . It is definite as represented in equation. 5.
= 1+ R-NP/Lcap (5)

Then attained the recommended novel metric rate control mechanism. It is definite as
the present projected link volume ().
RCM = Runit (t)= (tIupdate) : (6)
Deliberate h1 as tester time, C is the connection volume, Tp be the system delay, N0 be the
amount of contributors, and p0 be the anticipated reducing likelihood for founding the
constancy completed the designated linkage. The frame at symmetry can be expressed as W 0
= (1/p0) and RTT in signified as R0 = (N0.W0)/C. Correspondingly, sequence length at
stability can be specified as q0 = C., although the max. & min. sequence distance is signified
as Qmax = q0 + 100 and Qmin = q0 100. Later, the eigenvalue matrix of the circulated
interacting scheme is exemplified as:
bo = max (Eo) ( +.exp -1Freq.Ro)) (7)
As above equation 7, Eo is the purpose for accomplishment eigenvalue of classification
matrix, is the matrix appraised from the no. of transmitters No, and round trip time Ro,
although is additional scheme matrix seeing reason being congestion.
= {(- N0/ R0 ^2.C) , 2.C. R0 - q0 +qmin/c.(qmax - qmin). R0 ^2} (8)
The beyond equation. 8 provide the cost of which is recommend for Eq. 7. Later, if bo > 0,
before the circulated system can be measured as unbalanced. It is required to development bo
as it is extremely non-deterministic in nature outstanding to the dynamic traffic state in
viewpoint Internet construction. Consequently, in training to project a stochastic perfect, a
distributed network association is stately in its high traffic proper as bo is a non-deterministic
prescribed.
6.3 PROPOSED IDEA (HEECP)
Earlier wireless cellular networks are typically deployed as homogeneous networks using a
rate control process. A homogeneous cellular system is a network of BSs/nodes in a planned
layout and a collection of User Equipments (UEs), in which all the BSs/nodes have similar
transmit Power levels, same distance patterns, receiver noise levels, and similar backhaul
connectivity to the data network.
Heterogeneous networks (HetNets), consisting of different types of node that can
improve both the coverage and capacity of a distributed network. However, from an energy
efficiency perspective they can perform poorly since a large scale deployment of small cells
together consume a significant amount of energy. Additionally, small cells are not as

commonly accessed as a large cell, which means that for most of the time they stand idle.
Therefore, to improve the energy efficiency of HetNets.
The heterogeneous energy efficient congestion aware protocol solution aims to
provide a uniform broadband experience to users anywhere in the network through the local
deployment of heterogeneous nodes. Such nodes can be either deployed in outdoor or in
indoor environment. Furthermore, they can turn as network APs (Access points) or conveying
the message generated by the different levels capacity of the network towards multiple
packets and users. This approach offers very high capacity and data rates in congested areas
covered by the low-power routes through HEECP. However, due to their reduced congestion
a dense deployment of local AP (Access point) may be required. Therefore, in such a novel
heterogeneous network architecture, a high number of cells of different characteristics may
share the same spectrum in a given environmental area, increasing the inter-cell interference.
In heterogeneous distributed systems, outcome done expected routes that depend on
the minimum data rate of its whole associations. The EECP protocol which used for
appreciate precise event detection, energy effectiveness and less traffic. The declaration of
protocol will deal an effective mobile agent based clustering algorithm to appreciate the
energy potency and congestion resolve. In a way of associates with numerous files amounts,
In case of high data rate node forwards more transportation to a low data rate node, near to
chance of congestion, which proposals to long line up delays in such instructions. In the
meantime hop count is used as a routing metric in outmoded direction-finding, it do not
adjust to mobile nodes. A congestion-aware direction-finding metric for mobile ad hoc
networks should include broadcast ability, dependability, and congestion about a connection.
To measure the energy consumption of networks that relate the consumed energy to
the congestion coverage area and amount of transferred data are necessary. Designing an
energy efficient network is an optimization problem that aims at minimizing the energy
consumed by the network, while maintaining a specified quality of service (QoS). The
relation between energy consumption and other performance measuring the QoS. The typical
behaviour is that the energy consumption increase with the network load while the QoS
experienced by the users decrease due to increased interference.

The algorithm to implement rate control mechanism with Heterogeneous EECP is as
follows:
Step 1. Input sample Packet

Step 2. Design congestion header (14bytes)
Step 3. Congestion Header = {bottleneck rate(x), reverse path(x),RTT}
Step 4. Design end host function
Step 5. Each node initially advertises its neighbour mobile nodes to all other neighbours
using H EECP
Step 6. If (neighbour of a node changes)
Step 7. Each node a advertises its neighbour mobile nodes to all other neighbours using H
EECP.
Step 8. Route selection.
Step 9. Only those routes are selected which have minimum traffic.
Step 10. Read inbound packet update RTT
Step 11. Timestamp rate in outbound packets.
Step 12. Estimate Initial packet processing on Arrival of packet
Step 13. In bound_Bytes + = Data_size_Byte
Step 14. If (Current_data_RTT < Max_RTT)
Step 15. RTTx - = Current_data_RTT
Step 16. Data_with_RTT + = 1
Step 17. Perform processing on outbound data
Step 18. If (Data_BW_Request > Estimate_Ctrl_rate)
Step 19. Data_BW_Request = Estimate_Ctrl_rate
Step 20. Perform evaluation (throughput, delay, congestion rate and channel capacity)
Step 21. End

Timestamp
Data packet
Start
N Update the
Route to
route
Ingress dest Exit?
Interface Create
Con. Y
N
Check queue
Communication
Existence?
Y N
Allow
Evaluate capacity of connection Wait
Traffic
link and
update
Y
Mavg Traf_RateAgg
Address Mapping
Apply Evaluate translation
Moving Round Y rule exist complete
R_unit (t)
Evaluate Evaluate Stamping Evaluate Egress
Stop
update of RCM Interface
RCM H-EECP
Fig. 4.1: Flow chart of heterogeneous energy efficient congestion aware protocol

CHAPTER 5
RESULTS

RESULTS
5.1 INTRODUCTION
The proposed congestion control scheme is implemented on a multihop network. The

network is created with randomly distributed nodes. Assume network having following
properties:
Table 5.1: Input parameters
Input parameter Values

Networks Distributed systems
Sample Period 10-3
Link Capacity 1000

Number of TCP senders 60
Desired Dropping Probability 0.1
Number of channels 4
Number of trials 7
Timestamp rate in outbound 1000
Timestamp rate in outbound 100
Simulation tool MATLAB 2014a
This segment present an advantage of the projected H-EECP system examined concluded
with wide MATLAB models. In the simulation experiment, we have taken four parameters
namely queuing delay, congestion rate, and Throughput and channel capacity of the nodes.
Primarily, we have measured nodes are all significant nodes. The confidence cost of the node
is efficient intermittently afterward t seconds. The Trust Threshold level is occupied as
0.5 while min. and max. Trust values are 0 and 1 individually. The parameters are chosen as:
Sample Period=10-3, Link Capacity=1000, number of transmission control mechanism
senders =60, Desired Dropping Probability =0.1, Number of channels =4, Number of
trials=7, Timestamp rate in outbound =1000 and Timestamp rate in outbound=100 Fig 5.1

shows the graphical view of the Throughput, we can see that the higher throughput gain by
the H-EECP.
Fig. 5.1: Simulation trials Vs Throughput for Hetro-EECP, CICCO and RCM
Throughput: It is known as the packets received positively.

Fig 5.1 Provides the throughput for three procedures once the simulation trails is amplified.
As seen from the fig. the quantity is new in the case of Heterogeneous energy efficient and
congestion aware protocol as compare to rate control mechanism and CICCO.

Fig. 5.2: Simulation trials Vs Delay for Hetro-EECP, CICCO and RCM
Fig 5.2 shows the delay time while an incident arises, when the 1st packet recognized at the
base station. This condition as first package goes over the sensor nodes nearby the path that
can break wide-awake for this work. From now, subsequent packets do not necessity to
obtain the wake-up delay at every hop, and hence the end-to-end delay for the resulting
packets is considerable smaller than that of the 1st packet.

Fig. 5.3: Frequency Vs congestion rate for Hetro.
Frequency parameter
This is usages node hop mechanisms between different channels. Two variations are
explained as:
Common hopping: If nodes decide to interchange the data packets and reply the bounding
sequence after the transmission, all nodes hop permitting to the same patterns and also remain
in the similar network.
Independent hopping: Nodes dont depend on the other nodes hopping classification.
They usually follow their own bounding sequence which allows the talk of hopping
sequences.

Fig. 5.4: Simulation trials Vs vs. used channel capacity for Hetro-EECP, CICCO and RCM
A smallest-cost straight track ranking involving all nodes can be produced to classify the
minimum-cost tracks after sensor nodes to sink. Direction-finding the facts pack in the
direction of the sink on these minimum-cost tracks is capable as long as the rate of indication
collection is low or the network capability is correctly high. The average access controller
protocol controls the mean rate at which a sensor node can connect data to its neighbour over
a wireless network. This rate is the frequency volume.

CHAPTER 6
CONCLUSION AND FUTURE WORK

CONCLUSION
The objective of this research is to improve an end-to-end congestion control algorithm in
heterogeneous wireless environment with the Internet support. In heterogeneous distributed
network congestion follows with incomplete resources and throughput through an assumed
the smallest data rate of its whole links. Current routing procedures using hop count as a
routing metric do not explain well to mobile nodes. Heterogeneous wireless networks with
overlapped coverage areas as a congestion prevention method without losing service quality.
So here is a requirement for congestion conscious routing metric which contains transmission
ability, reliability, and congestion about a link. The relationship among available networks is
used to increase quality of system, particularly in a congestion scenario.
We have advanced a hop-by-hop congestion conscious protocol which works a mutual weight
value as a routing metric, founded on the data rate, delay, congestion rate, throughput and
channel capacity.
In this thesis, the collaboration of heterogeneous wireless networks with multi-homing
capability is considered for an end-to-end congestion control. We have used an energy-
efficient and congestion-aware protocol that determines numerous disjoint routes from a
source to destination, as our base. Between the learned routes, the route with minimum cost
directory is selected, which is grounded over node weight of the entire in-network nodes
source node to the endpoint node. By model consequences, it demonstrated that projected
protocol achieves high amount and channel capacity, by decreasing the delay and packet
drop.
FUTURE WORK
The proposed system that is heterogeneous energy efficient congestion aware protocol for
distributed network can be further modified for minimization of congestion through the help
of various energy efficient congestion control methods gives improved results.

REFERENCES
[1] R. Selvarani and K. Vinodha An efficient traffic regulation mechanism for

distributed networks Rangaswamy and Krishnareddy EURASIP Journal on Wireless
Communications and Networking, vol. 10, no. 2, 2015.
[2] P. Molinero Fernandez, N. McKeown, and H. Zhang, Is IP going to take over the
world (of communications?) ACM SIGCOMM Computer Communications Review,
Vol. 33, No. 1, 2003.
[3] Bulteman, D.C.A. SMIL 2.0.2. Examples and Comparisons, IEEE Multimedia vol.
9, no. 1, pp. 74-84, 2002.
[4] B. braden, D. Clark, j. crowcroft, B. devie, and S. deering, Recommendation on
Sequence Management and Congestion Avoidance in the Internet IETF RFC, 2309,
1998.
[5] D. Morley, and D. Parker. S. C, Understanding Computers: Today & Tomorrow.
Update: Today and Tomorrow, Cengage Learning Computers, 2009.
[6] Akari: Architecture Design Project for New Generation Network. NICT. 2007.
[7] Medeiros, F: ICT: Digitally Driven, Brussels. Source Cisco VNL. 2010
[8] Mahonen, P. Trossen, D. Papadimitrou, D. Polyzos, and D. Kennedy, Future
Networked Society. EIFFEL whitepaper, 2006.
[9] V. Jacobson, D. Smetters, J. Thornton, M. Plass, N. Briggs, and R. Barnyards,
Networking Named Content. Proceeding of ACM CoNEXT. Rome, Italy. 2009.
[10] J. Pan, S. Paul, and R. Jain, A Survey of the Research on Future Internet
Architectures. IEEE Communications Magazine. 2011.
[11] T. Moors, A critical review of End-to-end arguments in system design. Proceedings
of IEEE International Conference on Communications (ICC), NewYork City (New
Jersey), USA. April/May 2002.
[12] B. carpenter, RFC 1958: Architectural principles of the internet, network working
group, 1996.
[13] T. Li, Design Goals for Scalable Internet Routing. Work in progress, draft-
irtfrrgdesign-goals-02. 2010.

[14] M. A. Kafi, D. Djenouri, J. Ben-Othman, and N. Badache Congestion Control
Protocols in Wireless Sensor Networks: A Survey IEEE communications surveys &
tutorials, vol. 16, no. 3, 2014.
[15] C. Sergiou, P. Antoniou, and V. Vassiliou, A Comprehensive Survey of Congestion
Control Protocols in Wireless Sensor Networks IEEE communication surveys &
tutorials, vol. 16, no. 4, 2014.
[16] RJ. Kilduff, Analysis of Congestion Models for TCP Networks, the Hamilton
Institute National University of Ireland Maynooth Co. Kildare, 2003.
[17] J. Hong, J. Kook, S. Lee, D. Kwon and S. Yi, T-LEACH: The method of threshold-
based cluster head replacement for wireless sensor networks, Journal Information
Systems Frontiers, Vol. 11, Issue 5, pp 513-521, 2009.
[18] W. R .Heinzelman, A. Chandrakasan and H. Balakrishnan, Energy Efficient
Communication Protocol for Wireless Microsensor Networks, Proceedings of the 33
Hawaii International Conference on System Sciences, 2000.
[19] S. Tanachaiwiwat, P. Dave, R. Bhindwale and A. Heimy, Location-centric Isolation
of Misbehaviour and Trust Routing in Energy Constrained Sensor Networks, IEEE
International Conference on Performance Computing and Communications, 2004.
[20] M. Zarei, A.M. Rahmani, A. Sasan, and M. Teshnehlab, Fuzzy based trust estimation
for congestion control in wireless sensor networks, International Conference on
Intelligent Networking and Collaborative Systems, 2009.
[21] M. Zarei, A.M. Rahmani, R. Farazkish, and S. Zahirnia,FCCTF: Fairness Congestion
Control for a distrustful wireless sensor network using Fuzzy logic, 2010 10th
International Conference on Hybrid Intelligent Systems.
[22] A.Chakraborty, S. Ganguly, M.K.Naskar, and A.Karmakar, A Trust Based Fuzzy
Algorithm for Congestion Control in Wireless Multimedia Sensor Networks (TFCC),
proceedings. Of the 2nd International Conference on Informatics Electronics & Vision,
ICIEV, Dhaka, Bangladesh, 2013.
[23] K. Yedavalli, Using Wireless Advantage for Congestion Control in Wireless Sensor
Networks, USC Technical Report, 2005.
[24] CHN. Reddy, and RL. Reddy, Congestion avoidance and mitigation in WSN.
International Journal Advance Trends Computer Science Engineering, vol. 2, no. 1,
pp. 402406, 2013.

[25] Sheng. Z, Sheng. S and Yu .Y, A survey on the IETF protocol suite for the internet
of things: standards, challenges, and opportunities. Wireless Common IEEE, vol. 20,
no. 6, pp. 9198, 2013.
[26] R . Chakravarthi, and Gomathy. C, Hop-by-hop rate control technique for congestion
due to concurrent transmission in wireless sensor network. World Computer Science
Information Technology Journal, vol. 1, no. 8, pp. 351356, 2011.
[27] NR. Rathod and HJ. Patel, An approach on coding and congestion aware routing
mechanism in MANET. International Journal Emerging Science Engineering, vol. 1,
no. 3, pp. 4144, 2013.
[28] S. Sheeja and RV. Pujeri, Effective congestion avoidance scheme for mobile ad hoc
networks. Int J Computer Network Information Security, vol. 1, pp. 3340, 2013.
[29] C. Sergiou, V. Vassiliou and A. Paphitis, Congestion control in wireless sensor

networks through dynamic alternative path selection. Journal of computer networks.
vol. 14, no. 75, pp. 226 236, 2014.
[30] K. Jaewon, Z.Yanyong and B. Nath, TARA: topology-aware resource adaptation to

alleviate congestion in sensor networks. IEEE Trans Parallel Distributed System,
vol. 18, no. 7, pp. 919 31, 2007.
[31] D. Lee and K. Chung, Adaptive duty-cycle based congestion control for home
automation networks. IEEE Trans. Consumer Electronic, vol. 56, no. 1, pp. 42 47,
2010.
[32] S. Misra, V. Tiwari and M.S. Obaida, Lacas: learning automata-based congestion
avoidance scheme for healthcare wireless sensor networks. IEEE Journal Sel Areas
Common, vol. 27, no. 4, pp. 466 79, 2009.
[33] C. LAI, Congestion Control for Transmission Control Protocol (TCP) in Wireless
Networks. August 2011.
[34] N. Mohammadizadeh, Cooperative End-to-end Congestion Control in
Heterogeneous Wireless Networks, Institute of Waterloo, Ontario, Canada, 2013.
[35] S. Landstrom, Congestion Control in Wireless Cellular Networks March 2005.
[36] G. Kaur and VK. Mishra, Performance Enhancement Heterogeneous Based EECP
for Bandwidth Utilization Proceedings of the Second International Conference on
Research in Intelligent and Computing in Engineering, vol. 10, pp. 207215, 2017.

[37] A. Roy, Modification of congestion control algorithm for TCP and its extension to
explicit rate adjustment algorithm, Indian Institute of Technology, Kanpur, 2006.

Traffic Regulation Mechanism Using Heecp Protocol For D.N

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Traffic Regulation Mechanism Using Heecp Protocol For D.N

Uploaded by

Copyright:

Available Formats

CHAPTER 1

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 1

Computer Computer Computer

Fig. 1.1: Distributed network/system

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 2

There are three basic steps involved in implementing QoS on a network:

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 3

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 4

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 5

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 6

Different Distance and

Energy Efficient Route

Total Capacity of the

Packets are Sending and

Fig. 1.2: Working of heterogeneous energy efficient congestion aware protocol.

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 7

1.4 DISSERTATION ORGANIZATION

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 8

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 9

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 10

Designed for network dynamics to avoid congestion in the IP framework, the

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 11

The following table shows a network history.

Table 2.1: Network history

2.3 HISTORY OF DISTRIBUTING COMPUTING

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 12

2.3.2 The Internet

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 13

2.4.1 Client Server architecture:

Fig. 2.1: client server architecture

2.4.2 Layered application architecture:

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 14

Application Processing Layer

Data Management Layer

Fig. 2.2: Layered system architecture

2.4.3 Distributed object architecture:-

S(O1) S(O2) S(O3)

Object request broker

Fig. 2.3: Distributed object architecture

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 15

Maximum carrying Perfect

Fig. 2.4: Basic concept of congestion

Congestion is one of the vital problems of the distributed network (Internet).

In order to improve the performance of internet (distributed network). In this thesis

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 16

Traffic congestion is one of the compulsory environments in computer network.

2.6 CONGESTION CONTROL VS. FLOW CONTROL

2.7 CONGESTION CONTROL PARADIGM

2.7.1 Contention-based Congestion:

2.7.2 Buffer-based Congestion:

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 17

2.8 CHARACTERISTICS OF DISTRIBUTED NETWORKING

Some key characteristics of Distributed Networking are: Concurrency, Scalability,

2.8.4 Transparency: Network transparency is the procedure of transfer or retrieving data

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 18

Transparency of distributed system Description

2.9 CONGESTION DETECTION STRATEGIES

2.9.1 Packet Loss

2.9.2 Queue Length

2.9.3 Wireless Channel Load

TRAFFIC REGULATION MECHANISM USING HEECP PROTOCOL FOR D.N Page 19

2.9.5 Transmission Time Metrics