Online Learning Method for TCP Congestion Control in Cognitive Radio Networks
Jiang He-son · Science Technology and Engineering · 2014
Improved transport layer protocol have been widespread concerned to improve the throughput in cognitive radio networks. An online method( TCP-Learning) based on ACK interval is proposed which is used to probe the available bandwidth of TCP. The method can learn available residual bandwidth in the network and quickly adjust the congestion window of TCP. The simulation results show that the throughput of TCP-Learning performs the traditional congestion control algorithm such as Reno and Vegas in bad link conditions.