VOIP Congestion Control with Adaptable Token Generation Rate

VoIP users are concerned about quality of voice communication. Voice quality depends upon many factors like delay, packet loss. Such factors consequently bring in congestion. Congestion occurs when sender do not decrease sending rate or there is delay or packet loss. A smooth flow of traffic is needed for VoIP to maintain its quality. It is also important for VoIP throughput. To maintain a smooth transfer rate of voice packets a method is presented to control traffic congestion. It reduces the chances of congestion occurrence almost to zero. It also adopts according to changing traffic rate, handles burst and regulates the traffic for a continuous flow. This paper includes a token bucket traffic shaper in which token generation rate is predicted according to data burst and hence tackles the problem of congestion control. Token bucket traffic shaper uses fuzzy logic to predict adoptable token generation rate. Token rate prediction helps dynamic traffic shaping. Consequently token bucket traffic shaper scheme achieves less cell loss, lower delay and higher throughput. A simulation study was performed for token bucket traffic shaper using OMNeT Simulator. Simulation results show that token bucket traffic shaper gives better results as compared to normal token bucket with fixed token generation rate.