Fuzzy Congestion Control and Policing in ATM Networks

In this paper, usage parameter control (UPC) is presented to ensure that each source conforms to its negotiated parameters in ATM networks. To meet the requirements for the policing function, a fuzzy logic-based system is proposed to deal with the congestion control and policing problem in ATM networks. There are four types of different mechanisms, Jump Window (JW), Exponentially Weighted Moving Average (EWMA), Leaky Bucket (LB) and modified Fuzzy Leaky Bucket (FLB) techniques to be identified, analyzed and simulated by the traffic parameters. The Fuzzy Leaky Bucket (FLB) modifies the token rate in accordance with the peak rate, mean rate and the burst time, which characterize the source behavior. Simulation results show that FLB system is transparent to the sources with the parameter values which are negotiated at the call setup and effective in detecting source violation with low response time. The performance of FLB is significantly better than other mechanisms, in terms of both dynamic behaviors of responsiveness and selectivity.