Firefly Algorithm Based Proportional Integral Derivative Controller for AQM in TCP Networks

Abstract

Due to the rapid growth of internetworking traffic, the management of congestion in the computer network is considered an important problem. In general, congestion control mechanisms aim to utilize the network resources effectively by reducing delay and packet loss. Active queue management (AQM) techniques have been developed as an efficient Internet router mechanism for controlling network congestion and achieving stability. In this context, a feedback controller based on the Firefly Algorithm and a Proportional Integral Derivative Controller, which is called FA-PID, is designed to control the problem of network congestion and achieve stability and robustness against various network parameters. The proposed method stabilizes the queue length by increasing link utilization, minimizing queuing delays, and minimizing packet drop rates. The proposed method has been implemented using the MATLAB/Simulink model, and the experimental results have been discussed in TCP network environments. The simulation results showed the capability of the proposed method to achieve stability and robustness. In addition, it shows that the FA-PID reduced the packet drop rate and the average delay, as well as achieved better link utilization. The comparison results revealed demonstrate the superior performance of the proposed FA-PID method in comparison to alternative methods.