You are on page 1of 7

PACK: Prediction-Based Cloud Bandwidth and Cost

Reduction System
ABSTRACT:
In this paper, we present PACK (Predictive ACKs), a novel end-to-end traffic
redundancy elimination (TR) system, desi!ned for cloud computin! customers"
Cloud-#ased TR needs to apply a $udicious use of cloud resources so that the
#andwidth cost reduction com#ined with the additional cost of TR computation
and stora!e would #e optimi%ed" PACK&s main advanta!e is its capa#ility of
offloadin! the cloud-server TR effort to end clients, thus minimi%in! the
processin! costs induced #y the TR al!orithm" 'nli(e previous solutions, PACK
does not re)uire the server to continuously maintain clients& status" This ma(es
PACK very suita#le for pervasive computation environments that com#ine client
mo#ility and server mi!ration to maintain cloud elasticity" PACK is #ased on a
novel TR techni)ue, which allows the client to use newly received chun(s to
identify previously received chun( chains, which in turn can #e used as relia#le
GLOBALSOFT TECHNOLOGIES
IEEE PROJECTS & SOFTWARE DEVELOPMENTS
IEEE FINAL YEAR PROJECTS|IEEE ENGINEERING PROJECTS|IEEE STUDENTS PROJECTS|IEEE
BULK PROJECTS|BE/BTECH/ME/MTECH/MS/MCA PROJECTS|CSE/IT/ECE/EEE PROJECTS
CELL: +91 9!9" #9$"% +91 99&&' #"(% +91 9!9" "(9$% +91 9($1! !$!$1
V)*)+: ,,,-.)/012304546738+*-649 M0)1 +6:)333.)/01*3:546738+*;9:0)1-86:
GLOBALSOFT TECHNOLOGIES
IEEE PROJECTS & SOFTWARE DEVELOPMENTS
IEEE FINAL YEAR PROJECTS|IEEE ENGINEERING PROJECTS|IEEE STUDENTS PROJECTS|IEEE
BULK PROJECTS|BE/BTECH/ME/MTECH/MS/MCA PROJECTS|CSE/IT/ECE/EEE PROJECTS
CELL: +91 9!9" #9$"% +91 99&&' #"(% +91 9!9" "(9$% +91 9($1! !$!$1
V)*)+: ,,,-.)/012304546738+*-649 M0)1 +6:)333.)/01*3:546738+*;9:0)1-86:
predictors to future transmitted chun(s" *e present a fully functional PACK
implementation, transparent to all TCP-#ased applications and networ( devices"
+inally, we analy%e PACK #enefits for cloud users, usin! traffic traces from
various sources"
EXISTIN S!STE":
Traffic redundancy stems from common end-users& activities, such as repeatedly
accessin!, downloadin!, uploadin! (i"e", #ac(up), distri#utin!, and modifyin! the
same or similar information items (documents, data, *e#, and video)" TR is used
to eliminate the transmission of redundant content and, therefore, to si!nificantly
reduce the networ( cost" In most common TR solutions, #oth the sender and the
receiver e,amine and compare si!natures of data chun(s, parsed accordin! to the
data content, prior to their transmission" *hen redundant chun(s are detected, the
sender replaces the transmission of each redundant chun( with its stron! si!nature"
Commercial TR solutions are popular at enterprise networ(s, and involve the
deployment of two or more proprietary-protocol, state synchroni%ed middle-#o,es
at #oth the intranet entry points of data centers"
#ISA#$ANTAES %& EXISTIN S!STE":
Cloud providers cannot #enefit from a technolo!y whose !oal is to reduce
customer #andwidth #ills, and thus are not li(ely to invest in one"
The rise of -on-demand. wor( spaces, meetin! rooms, and wor(-from-home
solutions detaches the wor(ers from their offices" In such a dynamic wor(
environment, fi,ed-point solutions that re)uire a client-side and a server-side
middle-#o, pair #ecome ineffective"
cloud load #alancin! and power optimi%ations may lead to a server-side
process and data mi!ration environment, in which TR solutions that re)uire
full synchroni%ation #etween the server and the client are hard to accomplish
or may lose efficiency due to lost synchroni%ation
Current end-to-end solutions also suffer from the re)uirement to maintain
end-to-end synchroni%ation that may result in de!raded TR efficiency"
PR%P%SE# S!STE":
In this paper, we present a novel receiver-#ased end-to-end TR solution that relies
on the power of predictions to eliminate redundant traffic #etween the cloud and its
end-users" In this solution, each receiver o#serves the incomin! stream and tries to
match its chun(s with a previously received chun( chain or a chun( chain of a
local file" 'sin! the lon!-term chun(s& metadata information (ept locally, the
receiver sends to the server predictions that include chun(s& si!natures and easy-to-
verify hints of the sender&s future data" /n the receiver side, we propose a new
computationally li!htwei!ht chun(in! (fin!erprintin!) scheme termed PACK
chunking" PACK chun(in! is a new alternative for Ra#in fin!erprintin!
traditionally used #y R applications"
A#$ANTAES %& PR%P%SE# S!STE":
/ur approach can reach data processin! speeds over0 1#2s, at least 345
faster than Ra#in fin!erprintin!"
The receiver-#ased TR solution addresses mo#ility pro#lems common to
)uasi-mo#ile des(top2 laptops computational environments"
/ne of them is cloud elasticity due to which the servers are dynamically
relocated around the federated cloud, thus causin! clients to interact with
multiple chan!in! servers"
*e implemented, tested, and performed realistic e,periments with PACK
within a cloud environment" /ur e,periments demonstrate a cloud cost
reduction achieved at a reasona#le client effort while !ainin! additional
#andwidth savin!s at the client side"
/ur implementation utili%es the TCP /ptions field, supportin! all TCP-
#ased applications such as *e#, video streamin!, P3P, e-mail, etc"
*e demonstrate that our solution achieves 045 redundancy elimination
without si!nificantly affectin! the computational effort of the sender,
resultin! in a 345 reduction of the overall cost to the cloud customer"
S!STE" ARC'ITECT(RE:
+i!" 6" +rom stream to chain"
+i!ure 3- /verview of the PACK implementation"
S!STE" C%N&I(RATI%N:-
'AR#)ARE C%N&I(RATI%N:-
Processor - Pentium 7I8
9peed - 6"6 1h%
RA: - 3;< :=(min)
>ard ?is( - 34 1=
Key =oard - 9tandard *indows Key#oard
:ouse - Two or Three =utton :ouse
:onitor - 981A
S%&T)ARE C%N&I(RATI%N:-
Operating System : Windows XP
Programming Language : JAVA
Java Version : JDK 1. ! a"ove.
RE&ERENCE:
yal @ohar, Israel Cidon, and /snat :o(ryn-- PACKA Prediction-=ased Cloud
=andwidth and Cost Reduction 9ystem.-IEEE*AC" TRANSACTI%NS %N
NET)%RKIN +,-./

You might also like