DECbit

Abstract: : From the early days of modern congestion control, ushered in by the development of TCP's and DECbit's congestion control algorithm and by the pioneering theoretical analysis of Chiu and Jain, there has been widespread agreement that linear additive-increase-multiplicative-decrease (AIMD) control algorithms should be used. However, the early congestion control design decisions were made in a context where loss recovery was fairly primitive (e.g. TCP Reno) and often timed-out when more than a few losses occurred and routers were FIFO drop-tail. In subsequent years, there has been significant improvement in TCP's loss recovery algorithms. For instance, TCP SACK can recover from many losses without timing out. In addition, there have been many proposals for improved router queueing behavior. For example, RED active queue management and Explicit Congestion Notification (ECN) can tolerate bursty flow behavior. Per-flow packet scheduling (DRR and Fair Queueing) can provide explicit fairness. In view of these developments, we seek to answer the following fundamental question in this paper: Does AIMD remain the sole choice for congestion avoidance and control even in these modern settings? If not, can other mechanism(s) provide better performance? We evaluate the four linear congestion control styles - AIMD, AIAD, MIMD, MIAD - in the context of these various loss recovery and router algorithms. We show that while AIMD is an unambiguous choice for the traditional setting of Reno-style loss recovery and FIFO drop-tail routers, it fails to provide the best goodput performance in the more modern settings. Where AIMD fails, AIAD proves to be a reasonable alternative.

Abstract: Modern protocols designed for high-speed networks mostly adapt the rate at which packets are transmitted to control the flow of data between a sender and a receiver, and to control the load put upon the underlying network. In this article, the use of rate-adaption to realize 'congestion avoidance', a recently proposed highly sophisticated technique to control overload-conditions in networks is examined. With a simple analytical model of the system, expressions which help to tune it are derived. By directly comparing the resulting system to the well-known DECbit scheme which uses a window to adapt the load it is shown that most of the limitations of this scheme can be eliminated while preserving its advantages. More specific, because of the direct adaption of a rate fairness can be guaranteed in general topologies. This is possible without the need to put the burden of enforcing fairness on the (in a high-speed network already heavily loaded) routers. Finally the system can better adapt to varying packet-sizes, thus offering greater robustness and efficiency when used in real networks.

Abstract: In this paper a new approach has been proposed to eliminate congestion that occurs in Internet, mostly caused by multimedia traffic. The new algorithm, Priority Queue Algorithm for multimedia traffic (PQAM) has been developed and simulated. Analysis has been performed on TEXT, AUDIO and VIDEO data packets. It has been found by simulation that PQAM gives better service in terms of less delay in comparison to Adaptive Virtual Queue Algorithm, Virtual Queue Algorithm, and DECbit Algorithm.

Abstract: In this paper a new approach has been proposed to eliminate congestion that occurs in Internet, mostly caused by multimedia traffic. The new algorithm, Priority Queue Algorithm for multimedia traffic (PQAM) has been developed and simulated. Analysis has been performed on TEXT, AUDIO and VIDEO data packets. It has been found by simulation that PQAM gives better service in terms of less delay in comparison to Adaptive Virtual Queue Algorithm, Virtual Queue Algorithm, and DECbit Algorithm.