Detection of spectrum congestion in cognitive radio ad hoc networks

Spectrum congestion refers to the situation where there are not enough frequencies to accommodate multiple radios or radio systems operating simultaneously. For instance, spectrum congestion is not uncommon in the unlicensed band, where multiple WiFi, Bluetooth, and Zigbee radios may compete for spectrum. Although there are many studies on spectrum congestion in the unlicensed band, it is not well studied for a cognitive radio ad hoc network employing dynamic spectrum access. Since spectrum congestion in a cognitive radio ad hoc network incurs severe degradation of overall network performance, a novel spectrum congestion detection method is proposed in this paper to detect and locate congestions in such networks. Specifically, we propose a Spectrum-Aware Split Multipath Routing (SA-SMR) with dynamic channel assignment as a baseline multipath routing protocol, and investigate the feasibility of detecting spectrum congestion based on the statistics of the resulted paths from multipath routing. Furthermore, a cross-layer approach is proposed to deduce the cause of the spectrum congestion by comparing information from multiple layers, so that appropriate mitigation procedures can be carried out. Simulation results show the effectiveness of the proposed scheme, and the effects of the primary users' activities and the cognitive users' strategies.