Cost design for opportunistic multi-hop routing in Cognitive Radio networks

Cognitive radio (CR) is a revolutionary technology with promising applications in military areas since it enables CR users in the field to dynamically access to the vacant licensed frequency bands if no primary users are present. In practice, multi-hop routing in CR networks presents a great challenge due to unreliable traditional links and time varying unlicensed CR links. To improve the performance of multi-hop routing, opportunistic routing (OR) has been proposed and investigated extensively. Instead of using a single next hop, OR forwards a packet to an ordered set of candidate nodes and one node is chosen to relay the packet towards the destination. Most OR protocols prioritize the candidates and make the selection based on the cost defined as expected transmission times (ETX). Actually, ETX, as well as other existing criteria, does not always lead to the best forwarder choice for OR in CR networks since it ignores numerous potential CR links. In this paper, we propose a novel cost criterion for opportunistic multi-hop routing in CR networks, which leverages the unlicensed CR links to prioritize the candidate nodes and optimally selecting the forwarder. Simulation results show that our design efficiently decreases the number of transmissions, and effectively increases the throughput for most node pairs when compared with OR and traditional single-path routing.