Modeling and Performance Analysis of Dynamic Random Early Detection (DRED) Gateway for Congestion Avoidance

Introduction Computer networks have experienced an explosive growth over the past few years and with that growth have come severe congestion problems. For example, it is now common to see Internet gateways to drop 10% of the incoming packets because of local buffer overflows. Investigations of some of these problems have shown that much of the course lies in transport protocol implementation (not in the protocols themselves). The obvious ways to implement a window-based transport protocol can result in exactly the wrong behaviour in response to network congestion (Jacobson, 1998). In 1993, Floyds and Jacobson presented a very interesting work about how to detect congestion using routers provided by a random early detection mechanism. In high speed networks with connection with are likely to be designed with correspondingly large maximum queues to accommodate transient congestion. In the current Internet, the TCP (Transmission Control Protocol) transport protocol detects congestion only after a packet has been dropped at the gateway. However, according to Floyds and Jacobson (1993), it would be clearly undesirable to have large queues (possibly on the order of a delay-bandwidth product) that were near full much of the time; this would significantly increase the average delay in he network. Therefore, with increasingly high-speed networks, it is increasingly important to have mechanisms that keep throughput high but average queue sizes low. In this paper, the authors are interested in the concept of Dynamic Random Early Detection (DRED) Gateway for congestion avoidance. In the absence of explicit feedback from the gateway, transport-layer protocols could infer congestion from the estimated bottleneck service time, from changes in throughput, from changes in end-to-end delay, as well as from packet drops or other methods. Only the gateway has a unified view of the queueing behaviour over time; the perspective of individual connections is limited by the packet arrival patterns for those connections. In addition, a gateway is shared by many active connections with a wide range of roundtrip times, tolerances of delay, throughput requirements, etc; decisions about the duration and magnitude of transient congestion to be allowed at the gateway are best made by the gateway itself. Congestion Control and Resource Allocation Mechanisms for handling congestion i.e. congestion control may be divided into congestion prevention, congestion avoidance and congestion recovery. Congestion prevention guides against congestion at all times while congestion avoidance disallows the possibility of the occurrence of congestion, and congestion recovery tries to restore an operating state when demand has already exceeded capacity. One of the congestion avoidance mechanisms developed is the Random Early Detection (RED) gateway for congestion avoidance with somewhat different method for detecting congestion. To avoid congestion, resources are pre-allocated a in the case of Asynchronous Transfer Mode Network (ATM) of are allocated on-demand whether they are sufficient for the negotiated traffic or not as in he case of TCP/UDP traffics. In some Cases congestions are usually controlled if (and when) they occur. Some of the underlying service models include best-effort and multiple qualities of service. Figures 1 and 2 show the transmissions of packets from source to destination through single path and multiple paths respectively. In this research effort, we will consider single path model. While the principle behind RED gateways are fairly general, and RED gateways can be useful in controlling the average queue size even in a network where the transport layer protocol cannot be trusted to be cooperative, RED gateways are intended for a network where the transport protocol responds to congestion indications from the network. According to Jacobson (1998), the Internet is an arbitrary mesh-connected network. …