Abstract: In this study, we propose a new TCP based multimedia congestion control protocol called MCCP. MCCP implements a novel window based congestion algorithm on end-to-end available bandwidth estimation. MCCP effectively estimates the bottleneck bandwidth share of a connection. The estimate is based on information in the acknowledgements (ACKs) and the rate at which the ACKs are received. After a packet loss indication, which could be due to either congestion or link errors, the sender uses the estimated bandwidth to properly set the congestion window and the slow start threshold, thus maintaining the reasonable window size in case of random losses. The goal of MCCP is to estimate the connection available bandwidth to achieve high utilization of bandwidth, without starving other connections. MCCP deals well with highly dynamic bandwidth, large propagation time/bandwidth and random losses in the current and future heterogeneous Internet. MCCP uses fast probing, a mechanism that repeatedly resets slow start threshold based on available bandwidth share estimate. In congestion avoidance, MCCP invokes fast probing upon detection of extra available bandwidth via a scheme we call non-congestion detection (NCD). Fast probing is actually invoked under the following conditions: A large amount of bandwidth that suddenly becomes available due to change in network conditions. Random loss during slow-start that causes the connection to prematurely exit the slow-start phase. MCCP sender gets an accurate estimate of the connectionís fair share of the bottleneck bandwidth and effectively adjusts the sending rate to the changing estimate. As a result fast probing enhances probing during slow start and whenever non-congested conditions are detected. Experimental results in ns-2 simulation show that MCCP can significantly improve link utilization over a wide range of bandwidth, propagation delay and dynamic network loading.
S. Arumugam , T. Jebarajan and D.C. Joy Winnie Wise , 2008. Congestion Control Algorithm in Computer Communication Network. Journal of Mobile Communication, 2: 93-98.