Analytical Model for Congestion Control and Throughput with TCP CUBIC connections
|
|
- Kristin Skinner
- 6 years ago
- Views:
Transcription
1 Analytical Model for Congestion Control and Throughput with TCP CUBIC connections Sudheer Poojary and Vinod Sharma Department of Electrical Communication Engineering Indian Institute of Science, Bangalore Abstract We develop a Markov model for a TCP CUBIC connection. Next we use it to obtain approximate expressions for throughput when there may be queuing in the network. Finally we provide the throughputs different TCP CUBIC and TCP NewReno connections obtain while sharing a channel when they may have different round trip delays and packet loss probabilities. Index Terms TCP CUBIC, Internet window flow control, High speed Internet. I. INTRODUCTION TCP along with UDP has been the dominant transport layer protocol for the Internet. Traditionally TCP (TCP Reno [1] and TCP NewReno [2]) has been using an AIMD (Additive Increase Multiplicative Decrease) algorithm for congestion control. In the AI phase of the algorithm, the window sizes are increased linearly at the rate of one packet per RTT (round trip time). On detection of congestion, the window size is reduced to half of its current value, this is the MD phase of the algorithm. The specifics of the AIMD algorithm can be found in [3], [4]. The AIMD algorithms have performed well and have prevented severe congestion for more than two decades [5]. However as the Internet evolves to higher speeds, it has been found that the congestion control mechanism of AIMD TCP prevents it from using the link capacity efficiently. For example, in [5] it is shown that with a 10Gbps link, the maximum data rate attained by using an AIMD TCP is 7.55Gbps. It turns out that the slow linear window growth function of the AIMD TCP is a bottleneck in the high speed, large delay environment. Thus there has been a significant amount of research on improving the TCP performance by using aggressive congestion control mechanisms. H-TCP [6], Fast TCP [7], BIC [8], CUBIC [9] are a few examples. The H-TCP algorithm changes the rate of additive increase based on time elapsed since last congestion. It behaves like TCP Reno when the time elapsed since last congestion is low but as it increases, the window growth becomes more aggressive. FAST TCP measures RTT to update its window size and uses queuing delay as a congestion signal. BIC TCP uses packet drops as a sign of congestion and then searches for the optimum sending rate using a binary search mechanism. CUBIC, like H-TCP, uses time elapsed since last congestion to update its window size via a cubic function. The high speed congestion control mechanisms are expected to utilize the link efficiently but should not do this at the expense of standard TCP, i.e., they should be fair to standard TCP. Also, RTT fairness, which is a measure of fairness in data rate of flows with different RTTs, is another measure of performance. In [9], the authors show through experiments that TCP CUBIC scores high on the above performance measures compared to other congestion control mechanisms. TCP CUBIC has also been the default congestion control mechanism on Linux since 2006 (Linux kernel ). Thus, there has been considerable work on performance evaluation of TCP CUBIC through experiments and simulations [10], [11]. However [12] seems to be the only theoretical study on CUBIC. It considers a single TCP source. A Markov model of the window size is formed. Mean window size is obtained when the packet loss process is Poisson and RTT is constant. There are a number of analytical models for AIMD TCP. In [13] and [14], the authors make deterministic models of the TCP window size evolution and obtain expressions for TCP throughput. In [15], the authors model the window size evolution as a Markov regenerative process. They also consider the effect of retransmission timeouts on the data rate of TCP. [16] considers the effect of connection establishment and slow start on TCP data transfers. They extend the steady state model of [15] to compute latencies of short as well as long lived TCP flows. [17] analytically evaluates performance of TCP Tahoe and TCP Reno sharing a link subject to random losses. In [18], the authors calculate the throughput for multiple TCP connections which share the bottleneck link with UDP connections. [19] analyzes performance of multiple TCP connections going through multiple routers with exogenous traffic. The stability of the system is proved and throughput for each TCP flow is obtained when the routers employ RED [20]. In [21], the authors develop a model for determining drop probabilities of TCP flows in a TCP/IP network. The above models hold for AIMD TCP where the window growth function is linear. However, as the window growth function for CUBIC is nonlinear, we require a different model for its analysis. In this paper, we build an analytical model for TCP CUBIC with random packet drops and constant RTT. The model is further extended to the case where there is queuing, the RTT is variable and there are multiple TCP CUBIC and NewReno connections sharing a bottleneck link. Through comparison with ns-2 simulations we show that our model has good accuracy in capturing the dynamics of TCP-CUBIC /11/$ IEEE
2 TCP 1 Δ 1 ACKs III. WINDOW DYNAMICS OF TCP CUBIC TCP CUBIC, unlike TCP Reno has a cubic window growth function given by TCP N Data :. Data Fig. 1. Router Δ N C ACKs System with N TCP connections sharing a link The paper is organised as follows. Section II describes the system model. Section III makes a simple model of TCP CUBIC window flow control. It is then extended in Section IV to the case where there may be queuing and we obtain an approximation of the mean window size and the throughput obtained. In Section V we use these approximations to obtain throughputs and mean window sizes when multiple TCP CUBIC and NewReno connections share a link. Section VI concludes the paper. II. SYSTEM MODEL We consider the system shown in Fig 1. It has N TCP connections using TCP CUBIC (some perhaps TCP NewReno). They pass through a router/base station with link speed C bps. Each connection is transmitting a long file, i.e., it always has a packet to send. Each packet of connection i gets lost in transmission with probability p i 0 independently of others. (This is a likely scenario if the link is wireless). The connection i also has other propagation delay Δ i which may be a combination of other (non-bottleneck) link propagation and/or processing delays in the network. We will take Δ i sto be constant. This is a standard building block for a possibly larger network and is commonly considered in literature ([12], [15]). The main performance measure for this system is the throughputs obtained by the different TCP connections. To obtain the throughputs of different connections in this setup, we proceed as follows. First we consider a single TCP CUBIC connection and obtain its throughput when there is no queuing delay (i.e., the RTT is constant). To obtain its throughput, we study the window dynamics of TCP CUBIC which is quite different from AIMD TCP connections. In particular, we obtain the mean window size E[W ] of the connection when the packets are dropped with probability p and there is a fixed RTT Δ. Then the throughput is E[W ]/RT T. Next we generalize the setup to include queuing delays. Finally we consider the multiple TCP scenario and also integrate the dynamics of TCP NewReno. W cubic (W 0,t)=C(t K) 3 + W 0 (1) where W 0 is the window size at the last congestion epoch (the last instant there was a drop), t is the time elapsed since last congestion, K = 3 (W 0 β/c), C is a constant called the CUBIC parameter and β is another constant called the multiplicative drop factor. When there is a packet drop, the window is reduced to (1 β)w 0. The CUBIC function is chosen such that just after congestion the window size grows steeply towards W 0 which is the window size at which last congestion happened and then increasing slowly as the window size is near to W 0.The window size W 0 is deemed to be the steady state and hence the increase is slow near W 0. If the window size stays for a fair amount of time near W 0 without congestion, the window size further increases steeply to identify a new steady state. As the window size evolution in CUBIC is independent of RTT, CUBIC is RTT-fair. TCP CUBIC also incorporates TCP-friendly behaviour into its operation. A high speed protocol is called TCP-friendly if it is fair to standard TCP (AIMD TCP), i.e., does not steal bandwidth from standard TCP. To achieve both link efficiency and TCP friendliness, TCP CUBIC uses two window growth functions. When it is operating in TCP friendly region it evolves as W reno (W 0,t)=W 0 (1 β)+3 β t 2 β RT T. (2) The congestion window is updated when an ACK is received. On receipt of an ACK, equations (1) and (2) are evaluated to identify the mode of operation. If W reno (W 0,t) is larger than W cubic (W 0,t), then CUBIC operates in the TCP friendly region and uses eqn. (2), else it uses the cubic window growth function in eqn. (1). The details of TCP CUBIC operation can be found in [9]. We construct a model for CUBIC based on equations (1) and (2). We consider the case where the packet losses are random. Such a scenario would be common when the end systems use wireless links which are subject to random packet losses. We first consider the case where the link capacity is sufficiently large such that there is no queuing. Hence the RTT is constant in this case. In this setup it is convenient to assume that all the ACKs reach the TCP source at the same time (at the end of the RTT). Then if a packet is lost, its effect on the window size is also seen at the end of the RTT. This happens if the RTT is much larger than the transmission time. Let the window size at the beginning of the k th RTT period be W k. The window size at the congestion epoch preceding the k th RTT is denoted by W k. The time elapsed since the preceding congestion epoch (in multiples of RTT) is denoted by D k. We see that {(W k,d k)} forms a discrete time countable state (since D k is a multiple of RTT) Markov process.
3 The TCP sending rate is also constrained by the buffer space available at the receiver. The receiver notifies this to the sender during the initial handshake. The sender s congestion window size is then restricted to be less than (say) W max. Hence, W k is given as W k = W k (W k = w,d k = d) = min(w max, max(w cubic (w,d),w reno (w,d))) where, the functions W cubic (w,d) and W reno (w,d) are specified in equations (1) and (2) respectively. The process {(W k,d k)} evolves as (3) W k+1 = W k,d k+1 = D k +1 (4) if there is no packet loss in the k th RTT and as W k+1 = W k (1 β), D k+1 =0 (5) if some of the packets in the k th RTT are lost. (We ignore the effect of multiple packet drops in an RTT. We will see that it will have negligible effect on window dynamics for packet loss probabilities of practical interest). The probability of atleast one drop in the k th RTT period, when the window size is W k packets, is given by 1 (1 p) W k. This together with (4), (5) provides the transition matrix P of the Markov chain {(W k,d k)}. Since W k 1, the minimum probability of loss is p and maximum window size is W max <. Thus the state (1, 0) is reachable from any state in {(W k,d k)} within N steps with probability p N where N is the smallest integer with W max (1 β) N 1. This implies that the mean inter-arrival time of state (1, 0) is finite. Thus {(W k,d k} is an irreducible positive recurrent Markov chain with a unique stationary distribution π. Then, E[W ]= (1 β)w max (d= w =1 d=0 ) W k (w,d)π(w,d). (6) The average window sizes that we obtain using the above analysis for different values of RT T and packet error rate p are shown in Fig. 2. We also include the results via ns-2 simulation. For the ns-2 simulations, we set the bottleneck link capacity to be 1 Gbps which causes negligible queuing. We see a close match between the theoretical values and the ns-2 simulations. Once steady state E[W ] is obtained, the throughput in this setup (since RTT is fixed) is E[W ]/RT T packets/sec. We observe that for TCP Reno in this setup, E[W ] does not change with RTT; only with p. But for TCP CUBIC it is increasing with RTT for any given p. This is what gives it better throughput than TCP Reno for large RTT. In the next section we extend our model to the case where there is queuing and hence the RTT is not constant. IV. WINDOW DYNAMICS AND THROUGHPUT WITH VARIABLE RTT When the queuing delays at the router/base station are not negligible, computing TCP throughput is more complicated. In EW E[W] vs RTT for TCP CUBIC with random packet losses with packet error rate = p 70 p = (model) p = (model) 60 p = (model) p = (model) p = (ns 2) 50 p = (ns 2) p = (ns 2) 40 p = (ns 2) Fig RTT Average window sizes obtained using the Analytical model [22], the following M/GI/1 approximation was used for TCP Reno and was found to provide a good approximation. When RTT is constant, E[W ] for TCP Reno (at the end of RTT) depends on packet loss probability p but not on the RTT. Then the throughput obtained is E[W ]/RT T. If there is random RTT due to queuing delay then we approximate the TCP throughput by E[W ]/E[RT T ] λ packets/sec where E[RT T ] is the steady state mean RTT. However, we do not know E[RT T ]. This is approximated by E[RT T ]=Δ+E[S], (7) where E[S] is the mean sojourn time in an M/GI/1 queue with rate λ and the service time distribution same as the transmission time distribution in our system. Thus if the packet lengths in the TCP connection are iid with a generic length s, and its mean and second moment given by E[s] and E[s 2 ], from M/GI/1 results ([23]), we obtain E[S] = λe[s2 ] 2C 2 (1 ρ) + E[s] C where ρ = λe[s] C. Solving the system of equations (8) λ = E[W ],E[RT T ]=Δ+E[S] (9) E[RT T ] with E[S] given in (8) we obtain λ and E[RT T ] (and hence E[S]). It is worth remarking that in general an M/GI/1 queue will not give a good approximation to a TCP connection due to the window flow control of TCP. But (8) and (9) together capture that dynamics. Since the above approximation worked well for TCP Reno, we would like to use it for TCP CUBIC also. The difference in this case, as mentioned above, is that E[W ] depends on RTT. In the last section we computed E[W ] as a function of RTT and p when RTT was fixed. We use the same relationship now but replace RTT by E[RT T ] (this is an approximation which will be validated via ns-2 simulations). We write this relationship as E[W ]=g p (E[RT T ]) (10)
4 EW E[W] vs Δ for TCP CUBIC, link capacity = 10Mbps, link packet error rate = p NS2, p = 0.01 MG1 model, p = 0.01 NS2, p = MG1 model, p = Δ (Propagation delay) Fig. 3. Comparison of average window sizes obtained through the model and ns-2 for variable RTT Throughput Throughput vs Δ for TCP CUBIC, link capacity = 10Mbps, link packet error rate = p NS2, p = 0.01 MG1 model, p = 0.01 NS2, p = MG1 model, p = Δ (Propagation delay) Fig. 4. Comparison of throughput obtained through the model and ns-2 for variable RTT when we keep the packet loss probability p fixed. Although we do not have a closed form expression for g p, we compute g p (E[RT T ]) numerically as in the last section. We solve equations (8), (9) and (10) simultaneously to find λ, E[RT T ] and E[W ]. The results thus obtained are compared with ns-2 simulations with the same parameters. We use the Linux TCP CUBIC code incorporated into ns-2 ([24]) for our simulations. We compare the average window sizes obtained using the model and using ns-2 simulations in Fig. 3 for packet error probabilities p = 0.01 and The throughputs are compared in Fig. 4. The link capacity C is 10 Mbps and the packet size is 1050 bytes. (C is kept 10Mbps to have non-negligible queuing delay.) We see a close match between theory and simulations. In the next section, we use the model and the M/GI/1 approximation to get results for multiple TCP connections with different RTT and different probabilities of packet loss. V. THROUGHPUT AND AVERAGE WINDOW SIZES FOR MULTIPLE TCP CONNECTIONS Now we consider the system shown in Fig. 1 where multiple TCP connections 1, 2,,N share a single link. The TCP connection i is subject to random packet losses with rate p i and has a total propagation delay (i.e., from the source node to the destination node and back) of Δ i outside the queue. The shared link is the bottleneck link with capacity C bps. Initially we assume that all connections use TCP CUBIC. We will include TCP NewReno later. We use the M/GI/1 approximation to evaluate the sojourn time in the queue at the router. The average RTT for connection i is given by E[RT T i ]= λe[s2 ] 2C 2 (1 ρ) + E[s] C +Δ i (11) when E[s] and E[s 2 ] are the overall mean packet length and its second moment and λ packets/sec is the overall throughput. If λ i is the throughput of TCP connection i, then E[s] = i λ i λ E[s i], E[s 2 ]= i Then the mean window size of i th connection is λ i λ E[s2 i ]. (12) E[W i ]=g pi (E[RT T i ]) (13) and its throughput is approximated by λ i = E[W i] E[RT T i ]. (14) We can solve (11), (13) and (14) to obtain λ i, E[W i ] and E[RT T i ] for i =1, 2,,N. If some of the TCP connections (say connection i) areusing TCP NewReno, then we will still use the above system of equations except that g pi in (13) becomes a constant (i.e., independent of E[RT T i ]), equal to E[W i ] corresponding to its p i. There are several approximations available for E[W i ] for TCP NewReno. We use in our computation an expression from [14]. Table I compares results obtained using our model and ns- 2 simulations for the average window sizes E[W 1 ], E[W 2 ] for two TCP CUBIC connections with loss probabilities and propagation delays as mentioned in the table. The bottleneck link capacity is 2 Mbps (which provides substantial queuing delays). The Tables II, III compare the throughputs and the average RTT respectively. The approximations for throughput are better than for E[W ] and RTT. Next, we provide results for three TCP CUBIC connections. Tables IV and V compare the average window sizes and throughputs respectively. In this case connection i (i =1, 2, 3) experiences packet drop probability p i and propagation delay Δ i. The bottleneck link capacity is set to 5 Mbps. The approximations in this case are better. We have also compared the results for 10Mbps and 100Mbps link speeds. The approximations improve with link speed. These results are omitted due to lack of space. Tables VI and VII compare the average window sizes and throughputs for two TCP connections, one CUBIC and the other NewReno, sharing a link. Both the connections are subject to the same loss rate p =0.005 and same propagation delay Δ. We see that CUBIC is fair to NewReno at low RTTs but as the propagation delay, and hence the RTT, increases the data rates for CUBIC are much higher than NewReno. VI. CONCLUSIONS AND FUTURE WORK This paper develops an analytical model for TCP CU- BIC with fixed RTT and random packet losses. Through comparison with ns-2 simulations, we show that the model captures the dynamics of TCP CUBIC fairly well. We get the
5 TABLE I AVERAGE WINDOW SIZES FOR TWO TCP CONNECTIONS WITH BOTTLENECK LINK CAPACITY = 2 MBPS p 1,p 2 Δ 1, Δ 2 EW 1,EW 2 EW 1,EW TABLE II THROUGHPUTS FOR TWO TCP CONNECTIONS WITH BOTTLENECK LINK CAPACITY = 2 MBPS p 1,p 2 Δ 1, Δ 2 λ 1,λ 2 λ 1,λ TABLE III AVERAGE RTT FOR TWO TCP CONNECTIONS WITH BOTTLENECK LINK CAPACITY = 2 MBPS p 1,p 2 Δ 1, Δ 2 RT T 1,RTT 2 RT T 1,RTT TABLE IV AVERAGE WINDOW SIZES FOR THREE TCP CONNECTIONS WITH BOTTLENECK LINK CAPACITY = 5 MBPS Connection i p i Δ i EW i EW i (NS2) (Theory) TABLE V THROUGHPUT FOR THREE TCP CONNECTIONS WITH BOTTLENECK LINK CAPACITY = 5 MBPS Connection i p i Δ i λ i λ i (NS2) (Theory) TABLE VI AVERAGE WINDOW SIZE FOR CUBIC AND NEWRENO CONNECTIONS WITH BOTTLENECK LINK CAPACITY = 2 MBPS, 1 GBPS Link Speed Δ EW 1,EW 2 EW 1,EW 2 CUBIC, NewReno CUBIC, NewReno 2Mbps Gbps average window size, throughput and average RTT for TCP connections with variable RTT, i.e., connections subject to non-negligible queueing by approximation of the router with a M/GI/1 queue. We also obtain the above quantities for multiple TCP connections sharing a common bottleneck link. Finally we model the scenario where the TCP CUBIC connections share a link with TCP NewReno connections. The results using the model closely match the results obtained through ns-2 simulations. Further work would consider TCP connections going through multiple bottleneck links. TABLE VII THROUGHPUT FOR CUBIC AND NEWRENO CONNECTIONS WITH PACKET ERROR RATE, P = Bottleneck Link Speed Δ λ 1,λ 2 λ 1,λ 2 CUBIC, NewReno CUBIC, NewReno 2Mbps Gbps
6 REFERENCES [1] M. Allman, V. Paxson, and W. Stevens, TCP Congestion Control, RFC 2581 (Proposed Standard), Internet Engineering Task Force, Apr [2] S. Floyd and T. Henderson, The NewReno Modification to TCP s Fast Recovery Algorithm, RFC 2582 (Experimental), Internet Engineering Task Force, Apr [3] V. Jacobson, Congestion avoidance and control, in Symposium proceedings on Communications architectures and protocols, ser. SIG- COMM 88. New York, NY, USA: ACM, 1988, pp [4] R. Jain, K. K. Ramakrishnan, and D.-M. Chiu, Innovations in internetworking. Norwood, MA, USA: Artech House, Inc., 1988, ch. Congestion avoidance in computer networks with a connectionless network layer, pp [5] G. Huston, Gigabit TCP, Internet Protocol Journal, [6] R. N. Shorten and D. J. Leith, H-TCP: TCP for high-speed and longdistance networks, in Proc. PFLDnet, Argonne, 2004., [7] C. Jin, D. Wei, S. H. Low, J. Bunn, H. D. Choe, J. C. Doyle, H. Newman, S. Ravot, and S. Singh, Fast TCP: From theory to experiments, IEEE Network, vol. 19, pp. 4 11, [8] L. Xu, K. Harfoush, and I. Rhee, Binary Increase Congestion Control (BIC) for Fast Long-Distance Networks, in INFOCOM 04, [9] S. Ha, I. Rhee, and L. Xu, CUBIC: a new TCP-friendly high-speed TCP variant, SIGOPS Oper. Syst. Rev., vol. 42, pp , July [10] M. Bateman, S. Bhatti, G. Bigwood, D. Rehunathan, C. Allison, T. Henderson, and D. Miras, A comparison of TCP behaviour at high speeds using ns-2 and linux, in Proceedings of the 11th communications and networking simulation symposium, ser. CNS 08. New York, NY, USA: ACM, 2008, pp [11] M. C. Weigle, P. Sharma, and J. Freeman IV, Performance of competing high-speed TCP flows, in Proceedings of NETWORKING, Coimbra, Portugal, may 2006, pp [12] W. Bao, V. W. S. Wong, and V. C. M. Leung, A Model for Steady State Throughput of TCP CUBIC, in GLOBECOM, 2010, pp [13] T. Bonald, Comparison of TCP Reno and TCP Vegas: efficiency and fairness, Performance Evaluation, vol , pp , [14] M. Mathis, J. Semke, J. Mahdavi, and T. Ott, The macroscopic behavior of the TCP congestion avoidance algorithm, SIGCOMM Comput. Commun. Rev., vol. 27, pp , July [15] J. Padhye, V. Firoiu, D. F. Towsley, and J. F. Kurose, Modeling TCP Reno Performance: A Simple Model and Its Empirical Validation, IEEE/ACM Transactions on Networking, vol. 8, pp , [16] N. Cardwell, S. Savage, and T. Anderson, Modeling TCP latency, in in IEEE INFOCOM, 2000, pp [17] T. V. Lakshman and U. Madhow, The performance of TCP/IP for networks with high bandwidth-delay products and random loss, IEEE/ACM Trans. Netw., vol. 5, pp , June [18] V. Sharma and A. Gupta, Performance analysis of routers with TCP and UDP connections with priority and RED control, in Proceedings of the 15th international conference on Computer communication, ser. ICCC 02. Washington, DC, USA: International Council for Computer Communication, 2002, pp [19] V. Sharma and P. Purkayastha, Stability and Analysis of TCP Connections with RED Control and Exogenous Traffic, Queueing Syst. Theory Appl., vol. 48, pp , November [20] S. Floyd and V. Jacobson, Random early detection gateways for congestion avoidance, IEEE/ACM Trans. Netw., vol. 1, pp , August [21] E. Altman, K. Avrachenkov, and C. Barakat, TCP Network Calculus: The case of large delay-bandwidth product, in in Proceedings of IEEE Infocom, [22] K. Anand and V. Sharma, Computing TCP Throughput in a UMTS Network, in IEEE WCNC, 2008, pp [23] D. Bertsekas and R. Gallager, Data Networks 2nd Edition. Prentice- Hall, Inc., [24] D. X. Wei and P. Cao, NS-2 TCP-Linux: an NS-2 TCP implementation with congestion control algorithms from Linux, in WNS2 06: Proceeding from the 2006 workshop on ns-2: the IP network simulator. New York, NY, USA: ACM Press, 2006, p. 9.
Theoretical Analysis of High-speed Multiple TCP Connections through Multiple Routers
Theoretical Analysis of High-speed Multiple TCP Connections through Multiple Routers Sudheer Poojary and Vinod Sharma Department of ECE, Indian Institute of Science, Bangalore, India Email: {sudheer,vinod}@ece.iisc.ernet.in
More informationAnalysis of High Speed TCP Variants and TCP over Cognitive Radio
Analysis of High Speed TCP Variants and TCP over Cognitive Radio Sudheer Poojary Advisor: Vinod Sharma IISc, Bangalore November 30, 2015 Sudheer Poojary Analysis of High Speed TCP Variants and TCP over
More informationarxiv: v1 [cs.ni] 22 Feb 2016
arxiv:162.6653v1 [cs.ni] 22 Feb 216 Analysis of Multiple Flows using Different High Speed TCP protocols on a General Network Sudheer Poojary Vinod Sharma Department of ECE, Indian Institute of Science,
More informationBuffer Requirements for Zero Loss Flow Control with Explicit Congestion Notification. Chunlei Liu Raj Jain
Buffer Requirements for Zero Loss Flow Control with Explicit Congestion Notification Chunlei Liu Raj Jain Department of Computer and Information Science The Ohio State University, Columbus, OH 432-277
More informationPerformance Analysis of Loss-Based High-Speed TCP Congestion Control Algorithms
Performance Analysis of Loss-Based High-Speed TCP Congestion Control Algorithms HABIBULLAH JAMAL, KIRAN SULTAN Electrical Engineering Department University Of Engineering and Technology Taxila PAKISTAN
More informationTransmission Control Protocol (TCP)
TETCOS Transmission Control Protocol (TCP) Comparison of TCP Congestion Control Algorithms using NetSim @2017 Tetcos. This document is protected by copyright, all rights reserved Table of Contents 1. Abstract....
More informationImproved Model for a Non-Standard TCP Behavior
IJCSNS International Journal of Computer Science and Network Security, VOL.11 No.10, October 2011 45 Improved Model for a Non-Standard TCP Behavior Mohammed Abdullah Alnuem, King Saud University, Riyadh,
More informationA Comparison of TCP Behaviour at High Speeds Using ns-2 and Linux
A Comparison of TCP Behaviour at High Speeds Using ns-2 and Linux Martin Bateman, Saleem Bhatti, Greg Bigwood, Devan Rehunathan, Colin Allison, Tristan Henderson School of Computer Science, University
More informationCOMPARISON OF HIGH SPEED CONGESTION CONTROL PROTOCOLS
COMPARISON OF HIGH SPEED CONGESTION CONTROL PROTOCOLS Jawhar Ben Abed 1, Lâarif Sinda 2, Mohamed Ali Mani 3 and Rachid Mbarek 2 1 Polytech Sousse, 2 ISITCOM Hammam Sousse and 3 ISTLS Sousse, Tunisia ba.jawhar@gmail.com
More informationPerformance of Competing High-Speed TCP Flows
Performance of Competing High-Speed TCP Flows Michele C. Weigle, Pankaj Sharma, and Jesse R. Freeman, IV Department of Computer Science, Clemson University, Clemson, SC 29634 {mweigle, pankajs, jessef}@cs.clemson.edu
More informationCongestion Control. Daniel Zappala. CS 460 Computer Networking Brigham Young University
Congestion Control Daniel Zappala CS 460 Computer Networking Brigham Young University 2/25 Congestion Control how do you send as fast as possible, without overwhelming the network? challenges the fastest
More informationRED behavior with different packet sizes
RED behavior with different packet sizes Stefaan De Cnodder, Omar Elloumi *, Kenny Pauwels Traffic and Routing Technologies project Alcatel Corporate Research Center, Francis Wellesplein, 1-18 Antwerp,
More informationOn the Transition to a Low Latency TCP/IP Internet
On the Transition to a Low Latency TCP/IP Internet Bartek Wydrowski and Moshe Zukerman ARC Special Research Centre for Ultra-Broadband Information Networks, EEE Department, The University of Melbourne,
More informationImpact of Short-lived TCP Flows on TCP Link Utilization over 10Gbps High-speed Networks
Impact of Short-lived TCP Flows on TCP Link Utilization over Gbps High-speed Networks Lin Xue, Chui-hui Chiu, and Seung-Jong Park Department of Computer Science, Center for Computation & Technology, Louisiana
More informationTM ALGORITHM TO IMPROVE PERFORMANCE OF OPTICAL BURST SWITCHING (OBS) NETWORKS
INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS ISSN 232-7345 TM ALGORITHM TO IMPROVE PERFORMANCE OF OPTICAL BURST SWITCHING (OBS) NETWORKS Reza Poorzare 1 Young Researchers Club,
More informationPerformance of Competing High-Speed TCP Flows
Performance of Competing High-Speed TCP Flows Michele C. Weigle, Pankaj Sharma, and Jesse R. Freeman IV Department of Computer Science, Clemson University, Clemson, SC 29634 {mweigle, pankajs, jessef}@cs.clemson.edu
More informationSynopsis on. Thesis submitted to Dravidian University for the award of the degree of
Synopsis on AN EFFICIENT EXPLICIT CONGESTION REDUCTION IN HIGH TRAFFIC HIGH SPEED NETWORKS THROUGH AUTOMATED RATE CONTROLLING Thesis submitted to Dravidian University for the award of the degree of DOCTOR
More informationTCP based Receiver Assistant Congestion Control
International Conference on Multidisciplinary Research & Practice P a g e 219 TCP based Receiver Assistant Congestion Control Hardik K. Molia Master of Computer Engineering, Department of Computer Engineering
More information100 Mbps. 100 Mbps S1 G1 G2. 5 ms 40 ms. 5 ms
The Influence of the Large Bandwidth-Delay Product on TCP Reno, NewReno, and SACK Haewon Lee Λ, Soo-hyeoung Lee, and Yanghee Choi School of Computer Science and Engineering Seoul National University San
More informationAnalysis of Reno: A TCP Variant
International Journal of Electronics and Communication Engineering. ISSN 0974-2166 Volume 5, Number 3 (2012), pp. 267-277 International Research Publication House http://www.irphouse.com Analysis of Reno:
More informationAnalysis of Dynamic Behaviors of Many TCP Connections Sharing Tail Drop/RED Routers
Analysis of Dynamic Behaviors of Many TCP Connections Sharing Tail Drop/RED Routers Go Hasegawa and Masayuki Murata Cybermedia Center, Osaka University -3, Machikaneyama, Toyonaka, Osaka 560-853, Japan
More informationA Bottleneck and Target Bandwidth Estimates-Based Congestion Control Algorithm for High BDP Networks
A Bottleneck and Target Bandwidth Estimates-Based Congestion Control Algorithm for High BDP Networks Tuan-Anh Le 1, Choong Seon Hong 2 Department of Computer Engineering, Kyung Hee University 1 Seocheon,
More informationTransmission Control Protocol. ITS 413 Internet Technologies and Applications
Transmission Control Protocol ITS 413 Internet Technologies and Applications Contents Overview of TCP (Review) TCP and Congestion Control The Causes of Congestion Approaches to Congestion Control TCP Congestion
More informationCross-layer TCP Performance Analysis in IEEE Vehicular Environments
24 Telfor Journal, Vol. 6, No. 1, 214. Cross-layer TCP Performance Analysis in IEEE 82.11 Vehicular Environments Toni Janevski, Senior Member, IEEE, and Ivan Petrov 1 Abstract In this paper we provide
More informationcs/ee 143 Communication Networks
cs/ee 143 Communication Networks Chapter 4 Transport Text: Walrand & Parakh, 2010 Steven Low CMS, EE, Caltech Recap: Internet overview Some basic mechanisms n Packet switching n Addressing n Routing o
More informationImproving TCP Performance over Wireless Networks using Loss Predictors
Improving TCP Performance over Wireless Networks using Loss Predictors Fabio Martignon Dipartimento Elettronica e Informazione Politecnico di Milano P.zza L. Da Vinci 32, 20133 Milano Email: martignon@elet.polimi.it
More informationPerformance of high-speed TCP Protocols over NS-2 TCP Linux
Performance of high-speed TCP Protocols over NS-2 TCP Linux Masters Project Final Report Author: Sumanth Gelle Email: sgelle@cs.odu.edu Project Advisor: Dr. Michele Weigle Email: mweigle@cs.odu.edu Project
More informationOn ACK Filtering on a Slow Reverse Channel
On ACK Filtering on a Slow Reverse Channel Chadi Barakat and Eitan Altman INRIA, 2004 route des Lucioles, 06902 Sophia Antipolis, France {cbarakat,altman}@sophia.inria.fr Abstract ACK filtering has been
More informationA Generalization of a TCP Model: Multiple Source-Destination Case. with Arbitrary LAN as the Access Network
A Generalization of a TCP Model: Multiple Source-Destination Case with Arbitrary LAN as the Access Network Oleg Gusak and Tu rul Dayar Department of Computer Engineering and Information Science Bilkent
More informationTitle Problems of TCP in High Bandwidth-Delay Networks Syed Nusrat JJT University, Rajasthan, India Abstract:
Title Problems of TCP in High Bandwidth-Delay Networks Syed Nusrat JJT University, Rajasthan, India Abstract: The Transmission Control Protocol (TCP) [J88] is the most popular transport layer protocol
More informationTCP NewReno: Slow-but-Steady or Impatient?
TCP NewReno: Slow-but-Steady or Impatient? Nadim Parvez Anirban Mahanti Carey Williamson Department of Computer Science University of Calgary 5 University Drive NW Calgary, AB, Canada TN 1N Email: {parvez,
More informationCUBIC. Qian HE (Steve) CS 577 Prof. Bob Kinicki
CUBIC Qian HE (Steve) CS 577 Prof. Bob Kinicki Agenda Brief Introduction of CUBIC Prehistory of CUBIC Standard TCP BIC CUBIC Conclusion 1 Brief Introduction CUBIC is a less aggressive and more systematic
More informationEnhancing TCP Throughput over Lossy Links Using ECN-capable RED Gateways
Enhancing TCP Throughput over Lossy Links Using ECN-capable RED Gateways Haowei Bai AES Technology Centers of Excellence Honeywell Aerospace 3660 Technology Drive, Minneapolis, MN 5548 E-mail: haowei.bai@honeywell.com
More informationStudying Fairness of TCP Variants and UDP Traffic
Studying Fairness of TCP Variants and UDP Traffic Election Reddy B.Krishna Chaitanya Problem Definition: To study the fairness of TCP variants and UDP, when sharing a common link. To do so we conduct various
More informationCS321: Computer Networks Congestion Control in TCP
CS321: Computer Networks Congestion Control in TCP Dr. Manas Khatua Assistant Professor Dept. of CSE IIT Jodhpur E-mail: manaskhatua@iitj.ac.in Causes and Cost of Congestion Scenario-1: Two Senders, a
More informationThe effect of reverse traffic on the performance of new TCP congestion control algorithms
The effect of reverse traffic on the performance of new TCP congestion control algorithms Saverio Mascolo and Francesco Vacirca Dipartimento di Elettrotecnica ed Elettronica Politecnico di Bari Via Orabona
More informationEXPERIMENTAL EVALUATION OF TCP CONGESTION CONTORL MECHANISMS IN SHORT AND LONG DISTANCE NETWORKS
EXPERIMENTAL EVALUATION OF TCP CONGESTION CONTORL MECHANISMS IN SHORT AND LONG DISTANCE NETWORKS Mudassar Ahmad, Md Asri Ngadi, Mohd Murtadha Mohamad Department of Computer Science, Faculty of Computing,
More informationA Survey on Quality of Service and Congestion Control
A Survey on Quality of Service and Congestion Control Ashima Amity University Noida, U.P, India batra_ashima@yahoo.co.in Sanjeev Thakur Amity University Noida, U.P, India sthakur.ascs@amity.edu Abhishek
More informationTCP Veno: Solution to TCP over Wireless
TCP Veno: Solution to TCP over Wireless Franklin FU Presented by Franklin Fu Asst Professor School of Computer Engineering Nanyang Technological University Singapore January 31, 2004, 5:00am Singapore
More informationLecture 4: Congestion Control
Lecture 4: Congestion Control Overview Internet is a network of networks Narrow waist of IP: unreliable, best-effort datagram delivery Packet forwarding: input port to output port Routing protocols: computing
More informationPerformance Enhancement Of TCP For Wireless Network
P a g e 32 Vol. 10 Issue 12 (Ver. 1.0) October 2010 Global Journal of Computer Science and Technology Performance Enhancement Of TCP For Wireless Network 1 Pranab Kumar Dhar, 2 Mohammad Ibrahim Khan, 3
More informationTCP testing: How well does ns2 match reality?
TCP testing: How well does ns2 match reality? Martin Bateman CEPS University of Central Lancashire Preston, UK mbateman@uclan.ac.uk Saleem Bhatti School of Computer Science University of St Andrews St
More informationMean Waiting Delay for Web Object Transfer in Wireless SCTP Environment
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE ICC 009 proceedings Mean aiting Delay for eb Object Transfer in
More informationTCP Throughput Analysis with Variable Packet Loss Probability for Improving Fairness among Long/Short-lived TCP Connections
TCP Throughput Analysis with Variable Packet Loss Probability for Improving Fairness among Long/Short-lived TCP Connections Koichi Tokuda Go Hasegawa Masayuki Murata Graduate School of Information Science
More informationECE 610: Homework 4 Problems are taken from Kurose and Ross.
ECE 610: Homework 4 Problems are taken from Kurose and Ross. Problem 1: Host A and B are communicating over a TCP connection, and Host B has already received from A all bytes up through byte 248. Suppose
More informationAn Analytical Model for Short-Lived TCP Flows in Heterogeneous Wireless Networks
An Analytical Model for Short-Lived TCP Flows in Heterogeneous Wireless Networks Sameer Bataineh Jordan University of Science and Technology, Irbid, Jordan samir@just.edu.jo Jamal Al-Karaki Hashemite University,
More informationEffects of Applying High-Speed Congestion Control Algorithms in Satellite Network
Effects of Applying High-Speed Congestion Control Algorithms in Satellite Network Xiuchao Wu, Mun Choon Chan, and A. L. Ananda School of Computing, National University of Singapore Computing 1, Law Link,
More informationAn Enhanced Slow-Start Mechanism for TCP Vegas
An Enhanced Slow-Start Mechanism for TCP Vegas Cheng-Yuan Ho a, Yi-Cheng Chan b, and Yaw-Chung Chen a a Department of Computer Science and Information Engineering National Chiao Tung University b Department
More informationPerformance Analysis of TCP Variants
102 Performance Analysis of TCP Variants Abhishek Sawarkar Northeastern University, MA 02115 Himanshu Saraswat PES MCOE,Pune-411005 Abstract The widely used TCP protocol was developed to provide reliable
More informationCS 5520/ECE 5590NA: Network Architecture I Spring Lecture 13: UDP and TCP
CS 5520/ECE 5590NA: Network Architecture I Spring 2008 Lecture 13: UDP and TCP Most recent lectures discussed mechanisms to make better use of the IP address space, Internet control messages, and layering
More informationTCP/IP THROUGHPUT ENHANCEMENT FOR GLOBAL IP NETWORKS WITH TRANS-OCEANIC SUBMARINE LINK
/ THROUGHPUT ENHANCEMENT FOR GLOBAL NETWORKS WITH TRANS-OCEANIC SUBMARINE LINK Yohei Hasegawa, Masahiro Jibiki, Tatsuhiro Nakada, Yasushi Hara and Yasuhiro Aoki (NEC Corporation) Email:
More informationEVALUATING THE DIVERSE ALGORITHMS OF TRANSMISSION CONTROL PROTOCOL UNDER THE ENVIRONMENT OF NS-2
Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 4, Issue. 6, June 2015, pg.157
More informationAN IMPROVED STEP IN MULTICAST CONGESTION CONTROL OF COMPUTER NETWORKS
AN IMPROVED STEP IN MULTICAST CONGESTION CONTROL OF COMPUTER NETWORKS Shaikh Shariful Habib Assistant Professor, Computer Science & Engineering department International Islamic University Chittagong Bangladesh
More informationTHE NETWORK PERFORMANCE OVER TCP PROTOCOL USING NS2
THE NETWORK PERFORMANCE OVER TCP PROTOCOL USING NS2 Ammar Abdulateef Hadi, Raed A. Alsaqour and Syaimak Abdul Shukor School of Computer Science, Faculty of Information Science and Technology, University
More informationMarkov Model Based Congestion Control for TCP
Markov Model Based Congestion Control for TCP Shan Suthaharan University of North Carolina at Greensboro, Greensboro, NC 27402, USA ssuthaharan@uncg.edu Abstract The Random Early Detection (RED) scheme
More informationCS4700/CS5700 Fundamentals of Computer Networks
CS4700/CS5700 Fundamentals of Computer Networks Lecture 15: Congestion Control Slides used with permissions from Edward W. Knightly, T. S. Eugene Ng, Ion Stoica, Hui Zhang Alan Mislove amislove at ccs.neu.edu
More informationAn analytical model for evaluating utilization of tcp reno
An analytical model for evaluating utilization of tcp reno mohammad mehdi hassani reza berangi Abstract: This paper presents an analytical model for TCP Reno. For this model an algorithm is derived to
More informationCUBIC: A New TCP-Friendly High-Speed TCP Variant
CUBIC: A New TCP-Friendly High-Speed TCP Variant Sangtae Ha, Injong Rhee and Lisong Xu Presented by Shams Feyzabadi Introduction As Internet evolves the number of Long distance and High speed networks
More informationComparative Analysis of TCP Variants using NS and Netanim on Ubuntu Version Platform
Comparative Analysis of TCP Variants using NS- 3.25 and Netanim on Ubuntu Version 16.04 Platform Archita Saxena 1, Manish Choudhary 2 1 PG Student (Software Engg.), 2 Assistant Professor, Department of
More informationCongestion Control for High Bandwidth-delay Product Networks. Dina Katabi, Mark Handley, Charlie Rohrs
Congestion Control for High Bandwidth-delay Product Networks Dina Katabi, Mark Handley, Charlie Rohrs Outline Introduction What s wrong with TCP? Idea of Efficiency vs. Fairness XCP, what is it? Is it
More informationA New Fair Window Algorithm for ECN Capable TCP (New-ECN)
A New Fair Window Algorithm for ECN Capable TCP (New-ECN) Tilo Hamann Department of Digital Communication Systems Technical University of Hamburg-Harburg Hamburg, Germany t.hamann@tu-harburg.de Jean Walrand
More informationChapter II. Protocols for High Speed Networks. 2.1 Need for alternative Protocols
Chapter II Protocols for High Speed Networks 2.1 Need for alternative Protocols As the conventional TCP suffers from poor performance on high bandwidth delay product links [47] meant for supporting transmission
More informationCongestion Avoidance
Congestion Avoidance Richard T. B. Ma School of Computing National University of Singapore CS 5229: Advanced Compute Networks References K. K. Ramakrishnan, Raj Jain, A Binary Feedback Scheme for Congestion
More informationISSN: International Journal of Advanced Research in Computer Engineering & Technology (IJARCET) Volume 2, Issue 4, April 2013
Balanced window size Allocation Mechanism for Congestion control of Transmission Control Protocol based on improved bandwidth Estimation. Dusmant Kumar Sahu 1, S.LaKshmiNarasimman2, G.Michale 3 1 P.G Scholar,
More informationA New Modified Split-Connection Approach for Improving TCP Performance Over Wireless Networks
A New Modified Split-Connection Approach for Improving TCP Performance Over Wireless Networks Yashar Komijani University of Tehran y.komijany@ece.ut.ac.ir Arash Hooshmand University of Tehran arash@cad.ece.ut.ac.ir
More informationCongestion Collapse in the 1980s
Congestion Collapse Congestion Collapse in the 1980s Early TCP used fixed size window (e.g., 8 packets) Initially fine for reliability But something happened as the ARPANET grew Links stayed busy but transfer
More informationOpen Box Protocol (OBP)
Open Box Protocol (OBP) Paulo Loureiro 1, Saverio Mascolo 2, Edmundo Monteiro 3 1 Polytechnic Institute of Leiria, Leiria, Portugal, loureiro.pjg@gmail.pt 2 Politecnico di Bari, Bari, Italy, saverio.mascolo@gmail.com
More informationPERFORMANCE COMPARISON OF THE DIFFERENT STREAMS IN A TCP BOTTLENECK LINK IN THE PRESENCE OF BACKGROUND TRAFFIC IN A DATA CENTER
PERFORMANCE COMPARISON OF THE DIFFERENT STREAMS IN A TCP BOTTLENECK LINK IN THE PRESENCE OF BACKGROUND TRAFFIC IN A DATA CENTER Vilma Tomço, 1 Aleksandër Xhuvani 2 Abstract: The purpose of this work is
More informationTCP/IP Modeling and Validation
TCP/IP Modeling and Validation Chadi Barakat, INRIA, France Abstract We discuss in this article the different issues to be considered when modeling the TCP protocol in a real environment. The discussion
More informationModeling TCP/RED: a dynamical approach
Modeling TCP/RED: a dynamical approach Hui Zhang, Mingjian Liu, Vladimir Vukadinović, and Ljiljana Trajković Simon Fraser University, Burnaby, BC, V5A 1S6 Canada {hzhange, jliu1, vladimir, ljilja}@cs.sfu.ca
More informationComputer Networking
15-441 Computer Networking Lecture 17 TCP Performance & Future Eric Anderson Fall 2013 www.cs.cmu.edu/~prs/15-441-f13 Outline TCP modeling TCP details 2 TCP Performance Can TCP saturate a link? Congestion
More informationComparing TCP Congestion Control Algorithms Based on Passively Collected Packet Traces
Comparing TCP Congestion Control Algorithms Based on Passively Collected Packet Traces Toshihiko Kato, Atsushi Oda, Celimuge Wu, and Satoshi Ohzahata Graduate School of Information Systems University of
More informationOn the Effectiveness of CoDel for Active Queue Management
1 13 Third International Conference on Advanced Computing & Communication Technologies On the Effectiveness of CoDel for Active Queue Management Dipesh M. Raghuvanshi, Annappa B., Mohit P. Tahiliani Department
More informationCongestion Control. Principles of Congestion Control. Network assisted congestion. Asynchronous Transfer Mode. Computer Networks 10/23/2013
Congestion Control Kai Shen Principles of Congestion Control Congestion: Informally: too many sources sending too much data too fast for the network to handle Results of congestion: long delays (e.g. queueing
More informationTransport Layer (Congestion Control)
Transport Layer (Congestion Control) Where we are in the Course Moving on up to the Transport Layer! Application Transport Network Link Physical CSE 461 University of Washington 2 Congestion Collapse Congestion
More informationExploring Congestion Control Mechanism of TCP Variants over Wired & Wireless Networks
International Journal of Scientific & Engineering Research Volume 9, Issue 9, September-2018 183 Exploring Congestion Control Mechanism of TCP Variants over Wired & Wireless Networks Syful Islam, Ratnadip
More informationTelecommunication & Network Research Lab
Performance Modeling of SCTP Multihoming Shaojian Fu, Mohammed Atiquzzaman TR-OU-TNRL-- March Telecommunication & Network Research Lab School of Computer Science THE UNIVERSITY OF OKLAHOMA Felgar Street,
More informationFair TCP: A Novel TCP Congestion Control Algorithm to Tackle TCP s Challenges
Fair TCP: A Novel TCP Congestion Control Algorithm to Tackle TCP s Challenges Guodong Wang, Yongmao Ren, Jun Li Graduate University of Chinese Academy of Sciences Beijing, China Email: wangguodong@cstnet.cn
More informationTraffic Management using Multilevel Explicit Congestion Notification
Traffic Management using Multilevel Explicit Congestion Notification Arjan Durresi, Mukundan Sridharan, Chunlei Liu, Mukul Goyal Department of Computer and Information Science The Ohio State University
More informationReport on Transport Protocols over Mismatched-rate Layer-1 Circuits with 802.3x Flow Control
Report on Transport Protocols over Mismatched-rate Layer-1 Circuits with 82.3x Flow Control Helali Bhuiyan, Mark McGinley, Tao Li, Malathi Veeraraghavan University of Virginia Email: {helali, mem5qf, taoli,
More informationCongestion Control In The Internet Part 2: How it is implemented in TCP. JY Le Boudec 2015
1 Congestion Control In The Internet Part 2: How it is implemented in TCP JY Le Boudec 2015 Contents 1. Congestion control in TCP 2. The fairness of TCP 3. The loss throughput formula 4. Explicit Congestion
More informationAn End-to-End Support for Short-Lived TCP Flows in Heterogeneous Wired-cum-Wireless Networks: An Analytical Study
212 The International Arab Journal of Information Technology, Vol. 8, No. 2, April 211 An End-to-End Support for Short-Lived TCP Flows in Heterogeneous Wired-cum-Wireless Networks: An Analytical Study
More informationCongestion Control In The Internet Part 2: How it is implemented in TCP. JY Le Boudec 2014
1 Congestion Control In The Internet Part 2: How it is implemented in TCP JY Le Boudec 2014 Contents 1. Congestion control in TCP 2. The fairness of TCP 3. The loss throughput formula 4. Explicit Congestion
More informationDelayed ACK Approach for TCP Performance Improvement for Ad Hoc Networks Using Chain Topology
Delayed ACK Approach for TCP Performance Improvement for Ad Hoc Networks Using Chain Topology Prashant Kumar Gupta M.Tech. Scholar, Computer Networks, Bhilai Institute of Technology, Durg (C.G.), India
More informationA Framework For Managing Emergent Transmissions In IP Networks
A Framework For Managing Emergent Transmissions In IP Networks Yen-Hung Hu Department of Computer Science Hampton University Hampton, Virginia 23668 Email: yenhung.hu@hamptonu.edu Robert Willis Department
More informationFast Retransmit. Problem: coarsegrain. timeouts lead to idle periods Fast retransmit: use duplicate ACKs to trigger retransmission
Fast Retransmit Problem: coarsegrain TCP timeouts lead to idle periods Fast retransmit: use duplicate ACKs to trigger retransmission Packet 1 Packet 2 Packet 3 Packet 4 Packet 5 Packet 6 Sender Receiver
More informationRecap. TCP connection setup/teardown Sliding window, flow control Retransmission timeouts Fairness, max-min fairness AIMD achieves max-min fairness
Recap TCP connection setup/teardown Sliding window, flow control Retransmission timeouts Fairness, max-min fairness AIMD achieves max-min fairness 81 Feedback Signals Several possible signals, with different
More informationA Modified Congestion Control Algorithm for Evaluating High BDP Networks
84 IJCSNS International Journal of Computer Science and Network Security, VOL.12 No.11, November 2012 A Modified Congestion Control Algorithm for Evaluating High BDP Networks Ehab Aziz Khalil Dept. of
More informationInferring TCP Congestion Control Algorithms by Correlating Congestion Window Sizes and their Differences
ICSNC 24 : The Ninth International Conference on Systems and Networks Communications Inferring TCP Congestion Control Algorithms by Correlating Congestion Window Sizes and their Differences Toshihiko Kato,
More informationCS 344/444 Computer Network Fundamentals Final Exam Solutions Spring 2007
CS 344/444 Computer Network Fundamentals Final Exam Solutions Spring 2007 Question 344 Points 444 Points Score 1 10 10 2 10 10 3 20 20 4 20 10 5 20 20 6 20 10 7-20 Total: 100 100 Instructions: 1. Question
More informationDesign and Performance Evaluation of High Efficient TCP for HBDP Networks
Design and Performance Evaluation of High Efficient TCP for HBDP Networks TaeJoon Park 1, ManKyu Park 2,JaeYongLee 2,, and ByungChul Kim 2 1 Electronics and Telecommunications Research Institute 161 Gajong-Dong,
More informationCongestion Control In The Internet Part 2: How it is implemented in TCP. JY Le Boudec 2014
1 Congestion Control In The Internet Part 2: How it is implemented in TCP JY Le Boudec 2014 Contents 1. Congestion control in TCP 2. The fairness of TCP 3. The loss throughput formula 4. Explicit Congestion
More informationMEASURING PERFORMANCE OF VARIANTS OF TCP CONGESTION CONTROL PROTOCOLS
MEASURING PERFORMANCE OF VARIANTS OF TCP CONGESTION CONTROL PROTOCOLS Harjinder Kaur CSE, GZSCCET, Dabwali Road, Bathinda, Punjab, India, sidhuharryab@gmail.com Gurpreet Singh Abstract CSE, GZSCCET, Dabwali
More informationTCP START-UP BEHAVIOR UNDER THE PROPORTIONAL FAIR SCHEDULING POLICY
TCP START-UP BEHAVIOR UNDER THE PROPORTIONAL FAIR SCHEDULING POLICY J. H. CHOI,J.G.CHOI, AND C. YOO Department of Computer Science and Engineering Korea University Seoul, Korea E-mail: {jhchoi, hxy}@os.korea.ac.kr
More informationRate Based Pacing with Various TCP Variants
International OPEN ACCESS Journal ISSN: 2249-6645 Of Modern Engineering Research (IJMER) Rate Based Pacing with Various TCP Variants Mr. Sreekanth Bandi 1, Mr.K.M.Rayudu 2 1 Asst.Professor, Dept of CSE,
More informationImproving the Ramping Up Behavior of TCP Slow Start
Improving the Ramping Up Behavior of TCP Slow Start Rung-Shiang Cheng, Hui-Tang Lin, Wen-Shyang Hwang, Ce-Kuen Shieh Department of Electrical Engineering, National Cheng Kung University, Taiwan chengrs@hpds.ee.ncku.edu.tw
More informationTCP so far Computer Networking Outline. How Was TCP Able to Evolve
TCP so far 15-441 15-441 Computer Networking 15-641 Lecture 14: TCP Performance & Future Peter Steenkiste Fall 2016 www.cs.cmu.edu/~prs/15-441-f16 Reliable byte stream protocol Connection establishments
More informationAnalysis of TCP Latency over Wireless Links Supporting FEC/ARQ-SR for Error Recovery
Analysis of TCP Latency over Wireless Links Supporting FEC/ARQ-SR for Error Recovery Raja Abdelmoumen CRISTAL Laboratory, Tunisia Email: Raja.Abdelmoumen@ensi.rnu.tn Chadi Barakat Projet Planète, INRIA-Sophia
More informationDelay Performance of the New Explicit Loss Notification TCP Technique for Wireless Networks
Delay Performance of the New Explicit Loss Notification TCP Technique for Wireless Networks Wenqing Ding and Abbas Jamalipour School of Electrical and Information Engineering The University of Sydney Sydney
More informationTCP Westwood: Efficient Transport for High-speed wired/wireless Networks
TCP Westwood: Efficient Transport for High-speed wired/wireless Networks Mario Gerla, Medy Sanadidi, Ren Wang and Massimo Valla UCLA Computer Science 1 Outline 1. TCP Overview 2. Bandwidth Estimation and
More information