Sensing and Transmission in Probabilistically Interference-Limited Cognitive Radio Systems.

In this paper, we provide a fundamental channel model to characterize the interference effect inherent in cognitive radio systems. Mutual information rates of our proposed probabilistic block interference channels for both primary and secondary users are derived without assuming that receivers have knowledge on channel interference states. Novel constrained optimization problems are then put forward with a constraint on the performance loss margin tolerated by the primary user. Furthermore, we investigate some special cases where conditions are provided to justify the optimality of adopting Neyman-Pearson rule. Also presented are some scenarios in which randomized decision without using sensing measurement is needed to balance the rateloss for PU and throughput gain for SD.