An expert-control-based multicast congestion control mechanism for wireless networks

Abstract

With the increasing popularity of various multicast applications, the application of existing multicast congestion control mechanisms to wireless networks, which were designed for wired networks, is a challenging work due to high bit error rate, fading and handover. In this paper, we proposed an expert-control-based multicast congestion control mechanism for wireless networks, termed ECBMCC. In this novel mechanism, multicast receivers sent their feedback information to the expert controller rather than the sender, and the expert controller made sure the state of TCP connection by inferring according to the feedback information. With the help of expert inference, ECBMCC easily differentiated between wireless link error and network congestion, and chose the accurate congestion control policy to control the sending rate. And ECBMCC avoided the problem of feedback implosion, thus enhanced the scalability of multicast congestion control mechanism. The simulation results show that this mechanism can strengthen the congestion and random errors processing capability in real time and improve the network throughput, and this control scheme is more robust than existing mechanisms and more scalable.