Improving recital of transport protocol in cognitive radio mobile ad hoc networks

Cognitive Radio (CR) networks has the potential to solve the scarcity problem of wireless resources were unlicensed users can opportunistically perform transmission without impacting the operations of licensed users. In CogMANET, the channel switching is inherently necessary when primary users appears on the channel. During the channel switching, the communication characteristics such as bottleneck bandwidth and RTT will change. Due to this change, TCP has to adaptively update its congestion window (cwnd) to make an efficient use of the available resources. TCP CRAHN which incorporate spectrum awareness by combining explicit feedback from intermediate nodes and destination was proposed. In this paper, TCP CRAHN is evaluated in cases where bottleneck bandwidth and RTT changes. Based on the results, TCP CoBA (Collabarative Bottleneck Analysis) is proposed to further improve the throughput. This protocol updates the cwnd based upon the available buffer space in the relay node upon channel switch. Comparison of TCP CRAHN and TCP CoBA throughput will be analyzed and implemented using the NS2 Simulator.