Simulation Based Performance Evaluation of Several Active Queue Management Algorithms for Computer Network

Congestion crumbles the network performance, therefore. This paper evaluates several Active Queue Management (AQM) algorithms which used to control congestion. The main Objective of the paper is to find optimal algorithm that can be use to avoid congestion. AQM are router based mechanism which can detect congestion in early stage in the network ask the transmitter to decrease its transmitting rate, in this way the network can control the congestion for incoming packets. So some of AQM algorithms were evaluated by analyzed their performance, the selected AQM algorithms are Gentle BLUE (GB), Dynamic Gentle Random Early Detection (DGRED), Effective Random Early Detection (ERED), BLUE and Adaptive Max Threshold algorithms. Performance evaluation is carried by using JAVA simulation environments. Evaluation results show that GB compared with DGRED, ERED, BLUE and Adaptive Max Threshold outperformed in terms of mean queue length, delay and packet loss. GB had maximum dropping probability as compare with DGRED, ERED, BLUE and Adaptive Max Threshold. In term of throughput all tested algorithms all most give same throughput. The results prove that the GB is can be appropriate algorithm to handle congestion as compare to DGRED, ERED, BLUE and Adaptive Max Threshold.