Performance Investigations of FLRED and AGRED Active Queue Management Methods
Congestion is a major problem in computer networks. Congestion is considered one of the most essential problems in computer networks. Congestion occurs if the load in the network is greater than the capacity of router buffers. Congestion has many drawbacks. Such as, low throughput and high queuing delay. Many active queue management (AQM) methods have been developed to improve and control the congestion before the router buffer accumulated and overflowed. Many researchers have proposed (AQM) methods, such as Adaptive Gentle Random Early Detection (AGRED) and Fuzzy Logic Random Early Detection (FLRED). This paper presents a comparison between two AQM methods, i.e. FLRED, and the Adaptive GRED regarding for two performance measures throughput (T) and average queueing delay (D). In this paper, AGRED and FLRED methods have been evaluated and measured. In order to decide, which method presents more satisfactory performance results. After the simulation implemented, the result displays that the AGRED method offers the most satisfactory D results than FLRED method when congestion occurs. Also, in congestion and non-congestion status; AGRED and FLRED methods present the similarity performance results either congestion or non-congestion.
Keywords - Congestion networks, Fuzzy logic, Adaptive congestion control, Fuzzy logic Random Early Detection (FLRED), Adaptive GRED.