Distributed Optimal Relay Selection for Improving TCP Throughput over Cognitive Radio Networks: A Cross-Layer Design Approach

In cognitive radio (CR) networks, cooperative relaying is emerging as a key technology to improve the performance of secondary users (SUs), while ensuring the quality of service of primary transmissions. Most previous work considers maximizing physical layer throughput as a design criterion. However, the end-to-end Transmission Control Protocol (TCP) performance perceived by SUs is largely ignored. In this paper, we take a cross-layer design approach to jointly consider optimal relay selection strategy, physical layer adaptive modulation and coding and data-link layer frame size to maximize the TCP throughput in CR networks. Specifically, we formulate the CR relay network as a restless bandit system, where the finite-state Markov channel model is used to characterize the time-varying channel state. With this stochastic optimization formulation, the optimal policy has indexability property that dramatically reduces the computation and implementation complexity. Simulation results are presented to show the effectiveness of the proposed scheme.