Burst traffic: Congestion management and performance optimization strategies in heterogeneous cognitive radio networks

Abstract

In order to solve the network congestion problem caused by burst service flows in heterogeneous cognitive radio networks (Het-CRN), a two-stage dual-threshold random early detection (TD-RED) control mechanism is proposed here. Meanwhile, a three-state Markov modulated Poisson process (MMPP-3) is used to model the burst service to analyze its arrival in depth. Moreover, a performance evaluation system (PES) is proposed to quantify the interference of burst services to different classes of services and their quality of service, which can comprehensively evaluate the performance enhancement of the TD-RED control mechanism for each service. The system examines performance metrics such as packet loss rate, average delay, average captain etc. PES simulation results show that TD-RED significantly reduces the packet rejection rate and transmission delay compared to the no congestion control (NCC) and random early detection (RED) methods, thus providing superior quality of service to all services. thereby providing a superior quality of service to cognitive users. Its lower average captain fully proves that TD-RED can effectively alleviate congestion.