Published on

IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit www.ijera.com

Published in: Technology, Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. P.Praveen Kumar, Dr.K. Rameswaraiah / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.917-920917 | P a g eImproving The Performance Of Congestion Control In WirelessNetworksP.Praveen Kumar1, Dr.K. Rameswaraiah21Asst Prof ,Dept of CSE, Samskruthi College Of Engineering & Technology, Ghatkesar, A.P, India2Professor, Dept of CSE, ST.Peters Engineering College, Maisammaguda, Hyderabad, A.P, IndiaABSTRACTThe Transmission Control Protocol (TCP)provides an end-to-end, ordered transmission ofdata over a logical connection which links twoapplications during their communication.TCP is amainly concerned with moving data betweenapplications and it uses the routing servicesprovided by IP. The most widely used Internetprotocol in the present days is TCP. It should havethe knowledge of locating the applications and itneed not know about the topology of the network.TCP actually takes the data from the sourceapplication and delivers to the destinationapplication.TCP is mainly designed for using withunreliable systems such as IP, which only gets thedata from one host to another and perform lesserror checking. The important feature of TCP withrespect to its performance is congestion controlalgorithm.TCP mainly uses a number of techniquesfor achieving high performance andavoiding congestion collapse, which makes thenetwork performance to fall by several levels ofmagnitude. The TCP congestion avoidancealgorithm is the primary basis for the control ofcongestion.Keywords: Congestion Control, TCP, InternetProtocol.1. INTRODUCTION:The main protocol of the Internet protocolsuite is Transmission Control Protocol (TCP) which isone of the two main components of the suite, and theother is Internet Protocol (IP) and hence it iscommonly referred to as TCP/IP [1] [2] [3]. Most ofthe internet applications use TCP protocol for remoteadministration and file transfer. The performance ofthe Internet depends to a great extent as the TCPcarries most internet traffic [4]. A particular version ofTCP is defined by the congestion control algorithm asit employs. To establish the connection betweenmachines, TCP uses a lower-level communicationsprotocol and the Internet Protocol [5] [6]. An interfacethat allows streams of bytes to be sent and receivedconverts the data into IP datagram packets which areprovided by this connection. Datagram cannotguarantee about the arrival of the packets at theirdestination but TCP guarantees the delivery of bytesof data [7] [8]. Even if the network errors prevents thedelivery, but TCP handles the implementationproblems such as resending packets, and alerts theprogrammer in case if there is no network host or aconnection is lost [9]. A socket represents the virtualconnection between two machines and allows the datato be sent and received. TCP sockets are connected toa single machine and allow transmission of datathrough input and output streams. User Data gramProtocol (UDP) sockets are connected to multiplemachines and only can send and receive packets ofdata [10] [11] [12].2. ADVANTAGES OF TCPAutomatic Error Control: Data transmissionthrough TCP is more reliable than transmission ofpackets through UDP. In TCP, data packets are sentthrough a virtual connection and they make sure thatthey have not been corrupted and guarantee thedelivery of data and the lost packets are retransmitted[13]. In TCP, if the timer is disabled when therecipient sends an acknowledgment and if it is notreceived before the time, the packet is retransmitted.Reliability: The datagram packets will arrive out oforder when the two machines participating in a TCPconnection is transmitted by IP datagram Fig1. Forany program reading information from a TCP socketthe order of the byte stream will be disrupted andbecome unreliable [13] [14].
  2. 2. P.Praveen Kumar, Dr.K. Rameswaraiah / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.917-920918 | P a g eFig 1 IP datagram for transport over the network.The data will be passed to the interface of thesocket. Ease of Use: To Store the information indatagram packets is not the efficient way ofcommunication between the systems which createscomplexity for the task of designing and creatingsoftware within a limited time for programmers [15].TCP makes the programmers to think in a differentway i.e. by making data packaged into datagrampackets, it is otherwise treated as a continuous inputand outputs.3. AN OVERVIEW OF CONGESTIONCONTROL AND THE ALGORITHMS:TCP sockets follows the same of Unixprogramming, where communication is treated as fileinput and output where the developer writes to anetwork socket, a data structure, or a file [16] [17][18]. Communication through TCP sockets is simplerthan communicating through datagram packets. TCPsockets performs different operations such as TCP canestablish a network connection to a remote host , it canSend the data to a remote host and can receive datafrom a remote host and can close the connection isshown in fig 2. There is a special socket providing theservice of binding to a specific port number is found inTCP [19] [20]. This socket is generally observed onlyin servers, and performs the operations such asBinding to a local port, Accepting incomingconnections from remote hosts and Unbinding from alocal port. The two sockets are combined into differentcategories and they are used by a client or a server.TCP will use a sequence number in identifying eachdata byte. The sequence number of TCP helps inidentifying the order of the bytes sent from eachsystem and it helps in reconstruction of data when thedata is lost during the process of transmission orfragmentation process. In the cumulativeacknowledgment scheme of TCP, in the segment datafield, the first byte of payload was set as the sequencenumber by the sequence number field and the receiverwill send an acceptance by specifying the sequencenumber of the next byte for which they are expectingto receive.Fig 2. TCP virtual connection to transmit dataTCP uses an end-to-end flow control protocolin order to avoid the process of making the sender tosend the data much fast for the TCP receiver to receivethe data and process it. Congestive collapse is acondition where no communication occurs because ofcongestion and attains a packet switched computernetwork Congestion collapse is observed at chokepoints in the network, where the complete incomingtraffic exceeds the outgoing bandwidth to a node.The local area network and a wide area networkconnecting points are choke points. When the networkis in congestive collapse state, the packet delay and theloss of data are at high levels and traffic demand alsohigh and poor service quality. When additional datapackets were send more than required by theintermediate routers, makes additional packetsdiscarded and expect the end points for retransmissionof the information to the network. In the earlier timesTCP implementations had worse retransmissionperformance. When the packet was lost then the endpoints sent extra packets of data which repeated theinformation which was lost, and doubled the data ratesent, exactly the reverse of what has to be done duringcongestion and this pressed the entire network into acongestion collapse making most of the packets loss.Congestion control controls the traffic entry intoa telecommunication network, forpreventing congestive collapse by avoiding oversubscription of the processing capabilities of theintermediate networks and by reducing the rate ofsending data packets. There are different ways forclassification of congestion control algorithms: Basedon the type and amount of feedback received from thenetwork. Based on the feature of performance it isaiming to improve. Based on the equality criterion ituses. Two main components are required for theprevention of network congestion. They are a systemin routers for reordering or dropping packets underoverload, End-to-end flow control system which is
  3. 3. P.Praveen Kumar, Dr.K. Rameswaraiah / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.917-920919 | P a g edesigned into the end points for giving response tocongestion. The correct end point performance usuallyshould repeat the dropped information, but graduallyrepeat the information slowly. When all end pointsperform this, the congestion lifts and makes the gooduse of the network, and all the end points obtains a faircontribution of the available bandwidth. Themechanisms used for preventing congestion collapsesare fair queuing ,scheduling algorithms, and randomearly detection (RED), in which packets are droppedrandomly and slowed down the transmission processby proactively triggering the end points before theoccurring of congestion collapse. Fair queuingmechanism is useful for mostly in the choke pointsrouters having a small number of connections passingthrough them. A larger router mainly depends onRED. Some protocols such as TCP behave betterunder congested conditions where as UDP does notcongestion control mechanism. Protocols buildingatop UDP is independent of congestion and transmitsat a fixed rate should be able to handle the congestionin their own manner. Congestion should be kept out atthe periphery position of the network in pure datagramnetworks, where the above described mechanisms aregoing to handle it. Congestion is very difficult to dealwith the Internet backbone and instead of that lowpriced fiber-optic lines have made the cost cheaper inthe Internet backbone with sufficient bandwidth forkeeping congestion at the periphery. For implementingconnection oriented protocols, such as TCP protocol,generally observes packet errors, delays for adjustingthe speed of transmit. There are different congestionavoidance processes available such as TCP congestionavoidance algorithm which is the primary basis forcontrol of congestion in the Internet. TCP uses anumber of systems for achieving high performanceand avoiding congestion collapse, which makes thefalling of network performance by several orders ofmagnitude. These mechanisms control the congestioncollapse by controlling the rate of data inflow into thenetwork, maintaining the data flow below the rate atwhich trigger collapses. Senders use theacknowledgments for sending the data and forinterfering network circumstances between sender andreceiver of TCP. TCP senders and receivers canchange the performance of the data flow by couplingwith timers and is commonly referred to as networkcongestion avoidance. Recently the research isongoing in the areas of TCP reliably handling loss,minimizing errors and managing congestion. In thepresent system, without the intermediate station thesender sends the packet data.4. PROPOSED SYSTEMBased on the consideration of traffic rates andbuffer sizes, a router detection protocol was designedfor the number of congestive loses of packet data. Afoundation was built for all presently known host tohost algorithms by the proposal of host to hostcongestion control. It includes: The basic principle ofprobing the available network resources and toestimate the congestion state in the network a Loss anddelay-based techniques are used and the Techniques todetect packet losses quickly. A sliding window basedflow control is specified by the TCP standard. Thiscontrol has several mechanisms. The buffered data arepacketized into TCP packets which contains asequence number of the first data byte in the packet.The prepared data packets are transmitted to thereceiver using the IP protocol. It transmits a newportion of packets as soon as the sender receivesdelivery confirmation for at least one data packet.Until the receiver clearly confirms the delivery of theblock, the sender is responsible for a data block.Finally, the sender may decide that a particular datablock has been lost and start recovery procedures. Thereceiver forms an ACK (acknowledgement) packetthat carries one sequence number and several pairs ofsequence numbers by admitting the data delivery. Thedata blocks which having smaller sequence numbershave already been delivered is defined by a collectiveACK (acknowledgement). The sequence numbers ofdelivered data packets are clearly indicated by aselective ACK. However, we will discuss a notion ofACK packets as a separate entity without loss ofgenerality.5. CONCLUSIONS:In this paper, we have discussed and built afoundation for host-to-host congestion controlprinciples. Based on the consideration of traffic ratesand buffer sizes, we have designed and implemented arouter detection protocol for the number of congestiveloses of packet data. In this paper, we have describedthe various approaches to TCP congestion control. Thebasic problem of eliminating the congestion collapsephenomenon to problems of using available networkresources effectively in different types ofenvironments are surveyed. In this paper, we havediscussed about congestion control proposals whichare focused on environments in which the data packetsare reordered regularly by using this proposalefficiency can be improved. New challenges for TCPcongestion control has introduced with the technologyadvances.REFERENCES:[1] J. Postel, ―RFC793—transmission controlprotocol,‖ RFC, 1981.[2] C. Lochert, B. Scheuermann, and M. Mauve,―A survey on congestion control for mobile
  4. 4. P.Praveen Kumar, Dr.K. Rameswaraiah / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.917-920920 | P a g ead hoc networks,‖ Wireless Communicationsand Mobile Computing, vol. 7, no. 5, p. 655,2007.[3] J. Postel, ―RFC791—Internet Protocol,‖RFC, 1981.[4] A. Al Hanbali, E. Altman, and P. Nain, ―Asurvey of TCP over ad hoc networks,‖ IEEECommun. Surveys Tutorials, vol. 7, no. 3, pp.22–36, 3rd quarter 2005.[5] J. Widmer, R. Denda, and M. Mauve, ―Asurvey on TCP-friendly congestion control,‖IEEE Network, vol. 15, no. 3, pp. 28–37,May/June 2001.[6] H. Balakrishnan, V. N. Padmanabhan, S.Seshan, and R. H. Katz, ―A comparison ofmechanisms for improving TCP performanceover wireless links,‖ IEEE/ACM Trans.Netw., vol. 5, no. 6, pp. 756–769, December1997.[7] K.-C. Leung, V. Li, and D. Yang, ―Anoverview of packet reordering in transmissioncontrol protocol (TCP): problems, solutions,and challenges,‖ IEEE Trans. ParallelDistrib. Syst., vol. 18, no. 4, pp. 522–535,April 2007.[8] S. Low, F. Paganini, and J. Doyle, ―Internetcongestion control,‖ IEEE Control Syst.Mag., vol. 22, no. 1, pp. 28–43, February2002.[9] G. Hasegawa and M. Murata, ―Survey onfairness issues in TCP congestion controlmechanisms,‖ IEICE Trans. Commun.(Special Issue on New Developments on QoSTechnologies for Information Networks), vol.E84-B, no. 6, pp. 1461–1472, June 2001.[10] M. Gerla and L. Kleinrock, ―Flow control: acomparative survey,‖ IEEE Trans. Commun.,vol. 28, no. 4, pp. 553–574, April 1980.[11] J. Nagle, ―RFC896—Congestion control inIP/TCP internetworks,‖ RFC, 1984.[12] C. A. Kent and J. C. Mogul, ―Fragmentationconsidered harmful,‖ in Proceedings of theACM workshop on Frontiers in computercommunications technology (SIGCOMM),Stowe, Vermont, August 1987, pp. 390–401.[13] S. Floyd and K. Fall, ―Promoting the use ofend-to-end congestion control in theInternet,‖ IEEE/ACM Trans. Netw., vol. 7,no. 4, pp. 458–472, August 1999.[14] V. Jacobson, ―Congestion avoidance andcontrol,‖ ACM SIGCOMM, pp. 314–329,1988.[15] Z. Wang and J. Crowcroft, ―Eliminatingperiodic packet losses in 4.3– Tahoe BSDTCP congestion control,‖ ACM ComputerCommunication Review, vol. 22, no. 2, pp. 9–16, 1992.[16] V. Jacobson, ―Modified TCP congestionavoidance algorithm,‖ email to the end2endlist, April 1990.[17] M. Allman, V. Paxson, and W. Stevens,―RFC2581—TCP congestion control,‖ RFC,1999.[18] S. Floyd and T. Henderson, ―RFC2582—theNewReno modification to TCP’s fastrecovery algorithm,‖ RFC, 1999.[19] S. Floyd, T. Henderson, and A. Gurtov,―RFC3782—the NewReno modification toTCP’s fast recovery algorithm,‖ RFC, 2004.[20] M. Mathis, J. Mahdavi, S. Floyd, and A.Romanov, ―RFC2018—TCP selectiveacknowledgment options,‖ RFC, 1996.BIOGRAPHY:P.Praveen Kumar obtained hisB.Tech graduation from JNTUH in2008, M.Tech degree from JNTUHin 2011.Having 2 years of Industryexperience and 2 years of teachingexperience, working as AssistantProfessor in Samskruthi College OfEngineering & Technology, Ghatkesar, Ranga Reddy,AP-India, Presented 2 international papers at variousjournals and attended workshops at variousinstitutions. Interested areas of research are: DataMining and Ware housing, Network Security SoftwareEngineering, Information Security.Dr.K.Rameshwaraiah obtained hisB.Tech graduation from JNTUH in2001,M.S degree from UK in 2004andPh.D from SBU,UK in 2011.Having12 years of teaching experience ,working as professor in ST.PetersEngineering College, Maisammaguda, Hyderabad,AP-India ,Presented more than dozen international andnational papers at various journals and attended morethan 20 conferences and workshops across the globe.Interested areas of research Software Engineering,Image Processing, cyber terrorism&Security ,NetworkSecurity ,Information Security Data Mining and Warehousing and MANETS..