Y. Richard Yang and Simon S. Lam, "General AIMD congestion control," in Proceedings of ICNP, November 2000.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....probes available bandwidth by increasing the congestion window size linearly, and responds to increased congestion (indicated by packet losses) by decreasing the window size multiplicatively. Recently proposed congestion control mechanisms include generalizations of TCP like window based schemes [1, 2, 3, 4], and equation based schemes [5, 6, 7] A common objective of these schemes is to reduce the high variability of TCP s transmission rate. Such high variability may limit network utilization. In addition, it is not desirable for emerging applications such as real time streaming applications on the ....

....characterizes the steady state behavior of congestion control protocols, whereas both aggressiveness and responsiveness characterize transient behavior. An important observation is that there are tradeo s among smoothness, aggressiveness, and responsiveness [1, 9] Comparisons of TCP, general AIMD [1, 3], TFRC [5] and TEAR [2] have shown that typically higher smoothness means less aggressiveness and responsiveness. Several questions remain unanswered. First, both window based and equation based congestion control schemes have been studied recently. Window based schemes do not use history ....

[Article contains additional citation context not shown here]

Y. Richard Yang and Simon S. Lam, \General AIMD congestion control," in Proceedings of ICNP, November 2000.

....modifying the OMQN structure. 3 TCP Protocol Modification The basic congestion avoidance algorithm of TCP NewReno (presently the most diffused TCP version) is AIMD (Additive Increase Multiplicative Decrease) The basic properties of AIMD algorithms, generalizing the TCP protocol, were studied in [22, 23]. Several approximations are introduced for analytical tractability, the main of which is the incorrelation of the loss process with the TCP protocol, which does not allow to fully appreciate any property of the protocol that leads to a smaller loss probability. Recently, studies like [12, 15] ....

....and performance. Obviously, the estimation of the fair share window fs is a real value, while the window size is an integer: the rounded or truncated value can be used instead. We point out that this modification is not easily compared with the responsiveness of an AIMD protocol as defined in [23], since the relationship between the current window size and is not known, and for this reason we will avoid the use of this term. From the protocol point of view, this modification is easy, since it only implies to assign a different value to after a fast recovery. The OMQN model can take ....

Y. Yang, S. Lam, "General AIMD Congestion Control," Tech. Rep. TR-2000-09, University of Texas at Austin, May 2000. Available at http://www.cs.utexas.edu/users/lam/NRL/TechReports/

....UDP) 13] RTP does not specify any explicit congestion control mechanism like TCP. In order to be fair to protocols that use congestion control, such as TCP and to avoid any possible congestion collapse, several TCP friendly congestion control mechanisms were proposed for unreliable unicast ows [6, 12, 3, 15]. But all these mechanisms, like TCP, rely on implicit feedback and interpret any loss as a congestion loss. The implicit policy works accurately in wired networks where there are almost no losses on wired links. However, with the presence of wireless links, a large percentage of packets are lost ....

Y. R. Yang and S. S. Lam. General AIMD congestion control. In Proceedings of 8th ICNP, November 2000.

....rate adjustments can cause significant delays for applications. In recent years, researchers have proposed various TCP friendly congestion control protocols, such as equation based congestion control [2] and general additive increase and multiplicative decrease (AIMD) based congestion control [3]. These have significantly improved the performance of multimedia applications over the Internet [4] and flows of these protocols interact well with other TCP traffic. However, using TCP friendly congestion control reduces but does not remove the oscillations in the transmission rate. The rate ....

....for Two Continuous Back offs Time Rate d e f R W RTT Similar derivations can be applied to the buffering requirement that is used to smooth more than 2 continuous back offs. 3. 3 Buffering Requirement for AIMD based TCP friendly Congestion Control Protocols Early research [13,3] has studied how to make AIMD based congestion control friendly to TCP traffic in the Internet. A simplified result from the TCP friendliness study can be expressed as a constraint on its a and b parameters: b = 1 ( 3 . The derivation is available in [12] With this a and b relationship, ....

Yang Yang, and Simon Lam. "General AIMD Congestion Control" In Proceedings of ICNP 2000, Osaka, Japan, Nov 2000.

....it is a distinctly subordinate goal in our algorithm. Our algorithm is always less aggressive than AIMD TCP: it reacts the same way to losses and in addition, it reacts to increasing delays. Therefore, the work to ensure network stability under AIMD TCP applies to Nice as well. The GAIMD [52] and binomial [4] frameworks provide generalized families of AIMD congestion control algorithms to allow protocols to trade smoothness for responsiveness in a TCP friendly manner. The parameters can also be tuned to 19 make a protocol less aggressive than TCP. We considered using these frameworks ....

Y. Yang and S. Lam. General AIMD Congestion Control. In ICNP, 2000.

....It probes available bandwidth by increasing the congestion window size linearly, and responds to increased congestion (indicated by packet losses) by decreasing the window size multiplicatively. Recently proposed congestion control mechanisms include generalizations of TCPlike window based schemes [1, 2, 3, 4], and equation based schemes [5, 6, 7] A common objective of these schemes is to reduce the high variability of TCP s transmission rate. Such high variability may limit network utilization. In addition, it is not desirable for emerging applications such as real time streaming applications on the ....

....the steady state behavior of congestion control protocols, whereas both aggressiveness and responsiveness characterize transient behavior. An important observation is that there are tradeoffs among smoothness, aggressiveness, and responsiveness [1, 9] Comparisons of TCP, general AIMD [1, 3], TFRC [5] and TEAR [2] have shown that typically higher smoothness means less aggressiveness and responsiveness. 1.1 Motivation Our work is motivated by the need for new controls that have high smoothness in steady state and high aggressiveness responsiveness when network conditions change ....

[Article contains additional citation context not shown here]

Y. Richard Yang and Simon S. Lam, "General AIMD congestion control," in Proceedings of ICNP, November 2000.

....it is a distinctly subordinate goal in our algorithm. Our algorithm is always less aggressive than AIMD TCP: it reacts the same way to losses and in addition, it reacts to increasing delays. Therefore, the work to ensure network stability under AIMD TCP applies to Nice as well. The GAIMD [52] and binomial [4] frameworks provide generalized families of AIMD congestion control algorithms to allow protocols to trade smoothness for responsiveness in a TCP friendly manner. The parameters can also be tuned to make a protocol less aggressive than TCP. We considered using these frameworks for ....

Y. Yang and S. Lam. General AIMD Congestion Control. In ICNP, 2000.

....modifying the OMQN structure. 3 TCP Protocol Modification The basic congestion avoidance algorithm of TCP NewReno (presently the most diffused TCP version) is AIMD (Additive Increase Multiplicative Decrease) The basic properties of AIMD algorithms, generalizing the TCP protocol, were studied in [22, 23]. Several approximations are introduced for analytical tractability, the main of which is the incorrelation of the loss process with the TCP protocol, which does not allow to fully appreciate any property of the protocol that leads to a smaller loss probability. Recently, studies like [12, 15] ....

....and performance. Obviously, the estimation of the fair share window is a real value, while the window size is an integer: the rounded or truncated value can be used instead. We point out that this modification is not easily compared with the responsiveness of an AIMD protocol as defined in [23], since the relationship between the current window size and fs is not known, and for this reason we will avoid the use of this term. From the protocol point of view, this modification is easy, since it only implies to assign a different value to , after a fast recovery. The OMQN model ....

Y. Yang, S. Lam, "General AIMD Congestion Control," Tech. Rep. TR-2000-09, University of Texas at Austin, May 2000. Available at http://www.cs.utexas.edu/users/lam/NRL/TechReports/

....the flow pattern in cycle c repeats indefinitely, this value for the efficiency holds forever after. IV. RELATED WORK The impact of AIMD has been recently discussed mainly from two perspectives. First, from the perspective of the improvements of the original AIMD. Towards this end, Yang and Lam [9] discuss a control system which extends the system of Chiu further towards asynchronous feedback. Lahanas and Tsaoussidis in [10] propose a modification to increase both fairness and efficiency, namely AIMD FC. They prove the properties of the modified algorithm algebraically, and show ....

....the utilized bandwidth; at first, the system dynamics suggest that the higher the oscillation the less the efficiency. It also appears that the higher the oscillation, the faster we approach fairness. Some recent versions of AIMD based algorithms that take advantage of this property are [11] [9], 12] 13] More precisely, it has been observed that streaming applications could benefit from modest oscillations since these reflect the smoothness of adjusting the transmission rate backwards. Such protocols are characterized as TCP Friendly because they consume the same amount of bandwidth ....

[Article contains additional citation context not shown here]

Y. Yang and S. Lam, "General AIMD Congestion Control," in Proceedings of the IEEE International Conference on Network Protocols, November 2000.

....streams, the number of streams to use, and the size of the buffers for each stream. We also will see if our framework can be used to distinguish between random loss and congestive loss. We also plan to investigate additional algorithms for tuning the congestion control parameters (AIMD) for a flow [7, 13, 50] and to add Floyd s AIMD modifications [22] as a tunable option. We will continue to try and avoid packet loss through slow start modifications and Vegas like controls [10] and various burst reduction and rate smoothing techniques. We will look at WAD to WAD communication between sender and ....

Y. Yand and S. Lam. General AIMD Congestion Control. Technical report, University of Texas,

....from two different perspectives, namely the application requirements and the characteristics of the underlying networks. The former expounds the impact of the transmission gaps caused by halving the transmission rate during congestion on the quality of delay sensitive applications; authors in [4, 5, 8, 9] propose TCP friendly protocols that satisfy two fundamental goals: i) to achieve smooth window adjustments; this is done by reducing the window decrease ratio during congestion, and (ii) to compete fairly with TCP flows; this is approached by reducing the window increase factor according to a ....

....detected (congestion in wired networks versus transient random errors in wireless networks) 7] Implementation of such strategy requires a more responsive TCP. In this paper, we investigate the interrelation of TCP smoothness and responsiveness by studying the behaviors of TCP(#,#) protocols [8]. TCP(#,#) protocols parameterize the congestion window increase value # and decrease ratio # , where the sender s window size is increased by # if there is no packet loss in a round trip time, and the window is decreased to # times the current value if there is a loss indication. We ....

[Article contains additional citation context not shown here]

Y.R. Yang and S.S. Lam, "General AIMD Congestion Control", Proceedings of the 8th International Conference on Network Protocols", Osaka, Japan, November 2000.

....of TCP Vegas, but also as a step in incorporating learning algorithms in network protocols. In our case, one can cast the flow control problem in decision theoretic terms, however, as we discuss in Section II, setting a good optimization criteria may be tricky. Related work: Yang and Lam [4] studied AIMD flow control and derived the relationship between the parameters so that the flow would be TCP friendly [5] Bansal and Balakrishnan [6] evaluate a class of non linear congestion control algorithms. Both suggest different update parameters for the flow control, however, the ....

Yang Richard Yang and Simon S. Lam., "General aimd congestion control, " Tech. Rep. TR-2000-09, University of Texas at Austin, may 2000.

....and transition probabilities, rather than modifying the OMQN structure. 3 TCP Protocol Modification The basic congestion avoidance algorithm of TCP NewReno (presently the most diffused TCP version) is AIMD. The basic properties of AIMD algorithms, generalizing the TCP protocol, were studied in [22, 23]. Several approximations are introduced for analytical tractability, the main of which is the incorrelation of the loss process with the TCP protocol, which does not allow to fully appreciate any property of the protocol that leads to a smaller loss probability. Recently, studies like [12, 15] ....

....and performance. Obviously the estimation of the fair share window is a real value, while the window size is an integer: the rounded or truncated value can be used instead. We point out that this modification is not easily compared with the responsiveness of an AIMD protocol as defined in [23], since the relationship between the current window size and , is not known, and for this reason we will avoid the use of this term. From the protocol point of view, this modification is trivial, since it only implies to assign a different value to after a fast recovery. The OMQN model can ....

Y. Yang, S. Lam, "General AIMD Congestion Control," Tech. Rep. TR2000 -09, University of Texas at Austin, May 2000. Available from http://www.cs.utexas.edu/users/lam/NRL/TechReports/.

....convergence to the optimal point Extensive research on alternative congestion mechanisms brought the notion of TCP compatibility. The idea is that when TCP connections and TCP compatible algorithms share the line, each should get fair allocation (this is further described in [13] Yang and Lam [18] studied AIMD ow control and derived the relationship between the increase parameter and the decrease parameter so that the ow would be TCP compatible. The motivation was to enable each application to pick its version of AIMD while maintaining compatibility with TCP. By nding the general ....

....the 14 sum of k and l determines the throughput in any loss rate and speci cally when k l = 1 the mechanism is TCP compatible. In addition, it is shown that all binomial algorithms converge to fairness under synchronized feedback assumption (an assumption that was also made by Chiu and Jain [18]) provided that k l 0; k; l 0. To show this intuitively, the technique of Chiu and Jain [4] was used. Once again the simple case of two users was demonstrated. Figure 3.2 focuses on the increase strategy. When k = 0, the increase is additive, parallel to the 45 degrees line. When k 0 the ....

Yang Richard Yang and Simon S. Lam. General aimd congestion control. Technical Report TR-2000-09, University of Texas at Austin, may 2000.

....of efficiency is associated with the utilized bandwidth; at a first glance, the system dynamics suggest that the higher the oscillation the less the efficiency. It also appears that the higher the oscillation, the faster we approach fairness. Some recent versions of AIMD based algorithms [3] [4] [5] 6] attempt to take advantage of this property. More precisely, it has been observed that streaming applications could benefit from modest oscillations since these reflect the smoothness of adjusting the transmission rate backwards. Such protocols are characterized as TCPFriendly because they ....

....as TCPFriendly because they consume the same amount of bandwidth as TCP(1, does [11] Theoretically, and in the context of Figure 2, these algorithms try to push the Efficiency Line closer to the Bandwidth Limit Line at the expense of convergence speed to fairness. For example, the GAIMD [4] algorithm with = b D = converges to fairness in O(B log 1:14 B) steps, where B is the link bandwidth; the SIMD [6] with b D = 15=16 converges in O(B log 1:06 B) steps; the SQRT [5] algorithm with = p 1=B , converges in O(B ) so does the IIAD [5] The performance of AIMD ....

Y. Yang and S. Lam, "General AIMD Congestion Control," in Proceedings of the IEEE International Conference on Network Protocols, November 2000.

....[24] 3 Multiplicative decrease causes transmission gaps that hurt the performance of real time applications that experience jitter and degraded goodput. Furthermore, multiplicative decrease with a factor of or a window adjustment to 2 packets characterizes a rather conservative strategy [9, 13, 16, 24, 30, 31]. 4 Error detection lacks an appropriate classification module that would permit a responsive strategy, oriented by the nature of potential errors. That is, when errors appear to be transient due to short lived flows or random wireless interference, congestion control mechanisms (i.e. timeout ....

....However, from the research perspective of the present work it is important to consider that the authors of Vegas demonstrated effectively that measurement based window adjustment is a viable mechanism. A new family of protocols has recently emerged, namely the TCP friendly protocols [10, 13, 16, 22, 24, 30]. Congestion control was designed based on two fundamental requirements: i) to achieve smooth window adjustments; this is done by reducing the window decrease factor during congestion, and (ii) to compete fairly with TCP flows, which is achieved by adjusting the increase rate, calculated from a ....

[Article contains additional citation context not shown here]

Y.R. Yang and S.S. Lam, "General AIMD Congestion Control", In Proceedings of the 8th International Conference on Network Protocols, November 2000.

....sends its packets. Protocols with ACK based flow control utilize one or another version of TCP friendly congestion control, which includes Jacobson s modifications to TCP [1] 9] TCP like congestion control (e.g. 19] increase decrease algorithms (e.g. 2] 3] 5] 8] 10] 12] 14] [21]) and equationbased methods (e.g. 7] 15] These algorithms are shown to work well in the environment where the sender relies on self clocking, which refers to the use of positive acknowledgements in congestion control. However, current real time streaming applications in the Internet ....

....from the network, I D congestion control is executed on discrete timescales of R time units long. Typically, R is a multiple of the round trip delay (RTT) and in many cases, simply equals the RTT. Many papers study congestion control in the context of window based flow control [2] 8] 12] [21] and apply I D formulas to the size of congestion window cwnd. In such notation, assuming that the size of congestion window cwnd during interval i for a particular flow is given by w , I D congestion control can be summarized as: # # # = f w W w w D i i I i i . 1) where f is ....

[Article contains additional citation context not shown here]

Y.R. Yang and S.S. Lam, "General AIMD Congestion Control," University of Texas at Austin Technical Report ftp://ftp.cs.utexas.edu/pub/lam/gaimd.ps.gz, May 2000.

....(AIMD) It probes available bandwidth by increasing its congestion window size linearly, and responds to congestion (indicated by packet losses) by decreasing the window size multiplicatively. Recently proposed congestion control mechanisms include generalizations of TCP like window based schemes [1, 8, 19, 22] and equation based schemes [9, 17, 20] One common objective of these new schemes is to reduce variations in transmission rate. Such high variations may limit network utilization. In addition, they are not desirable for emerging applications such as real time streaming applications on the ....

....when there are drastic changes in network conditions. On the contrary, if #### is a linear or sub linear function of #, then the connections are unable to acquire bandwidth quickly. For example, TCP friendly AIMD algorithm needs to parameterize its control rules by defining # as a function of # [8, 22]. In particular, without considering the effect of TCP s timeout mechanisms, # # ##### # ##. Although smoothness is possible by using moderate decrease, AIMD becomes insensitive to sudden increases in available bandwidth. For SIMD, a remaining question is, how can we define # in the increase ....

[Article contains additional citation context not shown here]

Y. R. Yang and S. S. Lam. General AIMD congestion control. In Proceedings of ICNP, November 2000.

No context found.

Y. R. Yang and S. S. Lam. General AIMD congestion control. In Proceedings of the 8th IEEE International Conference on Network Protocols, Osaka, Japan, November 2000.

No context found.

Y.R. Yang, S.S. Lam, General AIMD congestion control, in: Proceedings of the 8th International Conference on Network Protocols #00, Osaka, Japan, November 2000, pp. 187--198.

No context found.

Yang Richard Yang and Simon S. Lam. General AIMD congestion control. In Proceedings of the 8th International Conference on Network Protocols, Osaka, Japan, November 2000. 33

No context found.

Y. R. Yang and S. S. Lam. General AIMD congestion control. In Proceedings of the 8th IEEE International Conference on Network Protocols, Osaka, Japan, November 2000.

....Let b(t) denote the allowed sending rate (in bytes per second) for rekey messages at time t. In the literature, there are extensive research results on how to determine the unicast or multicast sending rate in dynamic network situation. We refer interested readers to related papers such as [9, 24, 21, 15]. In this paper, we assume that b(t) is a given system parameter. We c l a i m t hat b(t) will not sharply change with time t. The total available bandwidth shared by data and rekey tra#c is a dynamic function of time. But we can adjust the rate of data tra#c to keep b(t) smooth. From now on, we ....

Y. Richard Yang and Simon S. Lam. General AIMD congestion control. In Proceedings of the 8th International Conference on Network Protocols '00, pages 187--198, Osaka, Japan, November 2000.

....be allocated to group rekeying. Let b(t) denote the allowed sending rate for rekey messages at time t. In the literature, there are extensive research results on how to determine the unicast multicast sending rate in dynamic network situations. We refer interested readers to related papers such as [7, 22, 19, 15]. In this paper, we assume that b(t) is a given system parameter. We claim that b(t) will not sharply change with time t. The total avail able bandwidth shared by data and rekey traffic is a dynamic function of time. However we can adjust the rate of data traffic to keep b(t) smooth. From now ....

Y. Richard Yang and Simon S. Lam. General AIMD congestion control. In Proceedings of the 8th International Conference on Network Protocols '00, Osaka, Japan, November 2000.

....will converge asymptotically to a fair state, ignoring such operational factors as randomness of the loss process and timeouts. The second method is to show that the long term mean sending rate of a protocol is approximately the same as that of TCP. However, it has been observed in experiments [27, 10, 9] that flows with TCP friendly long term mean sending rates can still have large rate variations when loss rate is high. Furthermore, fairness is only one of several desirable properties of a TCPfriendly congestion control protocol. We identify four desired properties: 1) fairness: small ....

....bandwidth. The objective of this paper is to evaluate these properties by analytically and experimentally studying the transient behaviors of several TCP friendly congestion control protocols. Proposed congestion control schemes in the literature fall into two major categories: AIMD based [13, 21, 6, 19, 27, 20] and formula based [16, 24, 22, 17, 10] For our study, we select TCP [14] and GAIMD [27] as representatives of the AIMD based schemes. GAIMD generalizes TCP by parameterizing the congestion window increase value and decrease ratio. That is, in the congestion avoidance state, the window size is ....

[Article contains additional citation context not shown here]

Yang Richard Yang and Simon S. Lam. General AIMD congestion control. In Proceedings of ICNP 2000.

....is appropriate for applications with elastic bandwidth requirements, it results in degraded performance for applications such as streaming media which need smoother increase decrease policies. Consequently, several new congestion control protocols such as RAP [22] TFRC [10] General AIMD or GAIMD [30] and Binomial Congestion Control [2] have been designed targeting the needs of speci c applications, especially streaming media. Because legacy TCP ows still dominate the Internet, it is desirable that these protocols be TCP friendly [16, 17] which roughly means that all ows must send at the ....

.... congestion control protocols in classical networking literature include Clark et al. s NETBLT [7] Cheriton and Williamson s VMTP [5] Ramakrishnan and Jain s DECBit [20] and Wang and Crowcroft s Tri S [27] The notion of TCP friendliness [16, 17] has given rise to a number of new proposals [2, 10, 22, 23, 26, 30] for the transport of streaming multimedia. The closest in approach to CYRF are GAIMD and Binomial congestion control, which are both shown to be special cases of CYRF. GAIMD [13, 30] generalizes TCP to an Additive Increase Multiplicative Decrease (AIMD) policy with di erent increase and decrease ....

[Article contains additional citation context not shown here]

Y. R. Yang and S. S. Lam. General AIMD congestion control. In Proceedings of the 8th International Conference on Network Protocols, November 2000.

....the TCP sending rate in congestion avoidance state is controlled by a congestion window which is halved for every window of data containing a packet drop, and increased by one packet per window of data acknowledged. Recently, many new congestion control protocols were proposed and investigated [4, 6, 13, 11, 1, 14, 7, 10, 12, 8, 15]. The objective of these new congestion protocols is to address the needs of new multimedia applications. We notice that, like TCP, many of these proposals are also based on the AIMD principle. Further, there is even a common belief that AIMD is optimal and is a necessary condition for a ....

Yang Richard Yang and Simon S. Lam. General AIMD congestion control. In Proceedings of ICNP 2000.

....only if it ensures convergence to fairness and eciency, atleast in ideal conditions. In this work, we look at one important subclass of end to end congestion control 1 protocols called Window based Binary Feedback Congestion Control Protocols (WBF Protocols, from here on) TCP [28, 37] GAIMD [41], and Binomial congestion control [2] are some of the well known WBF protocols. These protocols allow the application to simultaneously send a bunch of packets, called a window. The size of the window represents the protocol s estimate of its share of the bottleneck link. At the heart of every WBF ....

....several ideas from this famous paper including the metrics of fairness and eciency, and the Chiu Jain fairness index (section 4.1) 2. 2 Recent Work on Congestion Control The notion of TCP friendliness [20, 21] has given rise to a number of new proposals for unicast end to end congestion control [2, 12, 32, 33, 36, 41]. The primary motivation for most of these new protocols is the transport of streaming multimedia which requires smooth changes in sending rate. RAP [32] essentially proposes Additive Increase Multiplicative Decrease (AIMD) where each data packet is acknowledged. Redundant information in the ACKs ....

[Article contains additional citation context not shown here]

Y. R. Yang and S. S. Lam. General AIMD congestion control. In Proceedings of the 8th International Conference on Network Protocols, November 2000.

No context found.

Y. Richard Yang and Simon S. Lam, "General AIMD congestion control," in Proceedings of ICNP, November 2000.

No context found.

Y. Richard Yang and Simon S. Lam, "General AIMD Congestion Control",Technical Report TR-2000-09, Department of Computer Sciences, The University of Texas at Austin, May 9, 2000.

No context found.

Y. R. Yang and S. S. Lam, "General AIMD congestion control," University of Texas, Tech. Rep. TR-2000-09, May 2000, available http://www.cs.utexas.edu/users/lam/NRL/TechReports/.

No context found.

Y. R. Yang and S. S. Lam, "General AIMD congestion control," University of Texas, Tech. Rep. TR-2000-09, May 9, 2000.

No context found.

Y.R. Yang and S.S. Lam, "General AIMD Congestion Control," University of Texas at Austin Technical Report, May 2000.

No context found.

R. Yang and S. Lam. General AIMD Congestion Control. In Proceedings 8th International Conference on Network Protocols (ICNP'00), pages 187--198. IEEE, November 2000.

No context found.

Y. R. Yang and S. S. Lam, "General AIMD congestion control," University of Texas, Tech. Rep. TR-2000-09, May 9, 2000.

No context found.

Y. Yang and S. Lam, General AIMD Congestion Control, Proceedings of ICNP 2000.

No context found.

Y. Yang and S. Lam, "General aimd congestion control," in Proc. IEEE Intl. Conf. on Network Protocols (ICNP 2000.

No context found.

Y.R. Yang and S.S. Lam, General AIMD Congestion Control, the 8th IEEE Intl. Conf. on Network Protocols (ICNP 2000.

No context found.

Y. R. Yang and S. S. Lam. General AIMD congestion control. Technical Report UTCS TR-2000-09, University of Texas, Austin, Texas, May 2000.

No context found.

Y. Yang and S. Lam. General AIMD Congestion Control. In Proceedings of the IEEE International Conference on Network Protocols, November 2000.

No context found.

Y.R. Yang and S.S. Lam, "General AIMD Congestion Control", Proceedings of the 8th International Conference on Network Protocols", Osaka, Japan, November 2000.

No context found.

Y. Yang and S. Lam, \General aimd congestion control," 2000.

No context found.

Y. R. Yang and S. S. Lam. General AIMD Congestion Control. In Proceedings IEEE ICNP 2000.

No context found.

Y. Yang and S. Lam, "General AIMD Congestion Control," In Proceedings of ICNP 2000, Osaka, Japan, Nov. 2000.

No context found.

Y.R. Yang and S.S. Lam, "General AIMD Congestion Control," University of Texas at Austin Technical Report, May 2000.

No context found.

Y. Yang and S. Lam. General aimd congestion control. In Proc. IEEE Intl. Conf. on Network Protocols (ICNP 2000.

No context found.

Y. Yang and S. Lam. General AIMD congestion control. In Proc. of the IEEE ICNP, November 2000. http://www.cs.utexas.edu/users/lam/Vita/IEEE/YangLam00.pdf.

No context found.

Y. R. Yang and S. Lam. General aimd congestion control. In International Conference on Network Protocols, Nov. 2000.

No context found.

R. Yang and S. Lam. General AIMD Congestion Control. In Proceedings 8th International Conference on Network Protocols (ICNP'00), pages 187--198. IEEE, November 2000.

No context found.

Y. R. Yang and S. S. Lam, "General AIMD Congestion Control," IEEE ICNP, November 2000.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC