Performance measure of Drop tail and RED algorithm

Abstract

This Transmission control protocol (TCP) detects congestion only after a packet has been dropped and it would be undesirable to have large queue because it would significantly increase the average delay in the network. In the current high speed networks, it is increasingly important to have mechanisms that keep throughput high but average queue sizes low. This paper presents comparison between two queuing management system Random Early Detection (RED) and Drop tail for avoiding the congestion in high speed packet switched networks. In this paper we are trying to detect superfluous congestion by averaging the queue size. If average queue size exceeds with respect to required threshold value, router drops the packet with some drop probability determined by some probabilistic distribution, it can be further modified to give a better result. Drop probability of the packet is calculated by knowing the packet loss and this packet loss is minimized by modifying the congestion window size. In this paper we consider the own networks and we show the results of different protocols for particular network topology.