Iterative joint channel and noise variance estimation and primary user signal detection for cognitive radios

In this paper, we introduce a joint channel and noise variance estimation, and primary user (PU) signal detection scheme using the expectation-maximization (EM) algorithm for cognitive radios. In our investigation, we consider two scenarios: In the first scenario, the channel and noise variance are assumed to be perfectly known by the secondary user (SU). Here, we propose a maximum-likelihood (ML) solution of the PU signal detection as an upper bound on the performance of the proposed joint estimation and detection (JED) scheme. We also provide an iterative implementation of the ML-based detector using the EM algorithm. In the second case, we extend our work to the problem of channel and noise variance estimation in cognitive radios, where we propose an iterative JED scheme based on the EM algorithm. The simulation results show that the proposed JED scheme can iteratively attain a reliable performance with few iterations and modest computational complexity.