New congestion control algorithm to improve computer networks performance

摘要

Rapid advances in Internet and network technology affect dramatically on Congestion control methods. Congestion generally occurs when the router buffer cannot accommodate all arriving packets. This paper proposed an Adaptive Gentle Adaptive Random Early Detection (AGARED) method based on Adaptive Gentle Random Early Detection (AGRED) method and Adaptive Random Early Detection (ARED). AGARED tries to detect the congestion in early stage before the router buffer overflows. The proposed algorithm combine advantages of both AGRED and ARED, and enhance the parameter setting of average queue length (aql). AGARED is simulated using NS-2 by varying the variable of average queue length, minthreshold and maxthreshold to create different scenarios. The simulation results reveals that AGARED offers marginally better performance results than previous methods. Therefore, the results prove that AGARED is an effective method in controlling congestion router buffers of networks. Whereby, improve networks performance.