Dynamic Markov Chain Monte Carlo-Based Spectrum Sensing
Zheng Wang, Ling Liu, Kezhi Li · IEEE Signal Processing Letters · 2020
In this letter, a random sampling strategy is proposed for the non-cooperative spectrum sensing to improve its performance and efficiency in cognitive radio (CR) networks. The proposed refined Metropolis-Hastings (RMH) algorithm generates the desired channel sequence for fine sensing by sampling from the approximated channel availability distributions in an Markov chain Monte Carlo (MCMC) way. The proposal distribution during the sampling is fully exploited and the convergence of the Markov chain is studied in detail, which theoretically demonstrate the superiorities of the proposed RMH sampling algorithm in both sensing performance and efficiency.