Throughput-Sensing Tradeoff of Cognitive Radio Networks Based on Quickest Sensing

In this paper, we maximize the throughput of a cognitive radio (CR) network with respect to the frame length when quickest sensing is used. The amount of interference to the primary network, measured by the probability of collision with primary users (PUs), is constrained below a certain maximum. The corresponding problem when CRs use block sensing is also solved, assuming that collision with PUs causes a drop in throughput for the CR - this important assumption was missing in prior work. We then compare the maximum achievable throughput with block and quickest sensing schemes and show that for a practical range of protection levels for the primary network, the quickest sensing approach results in significantly higher average throughput.