Collaborative Spectrum Sensing Based on Hidden Bivariate Markov Models

The purpose of spectrum sensing is to determine idle portions of a licensed spectrum band that could be used by unlicensed or secondary users without causing harmful interference to primary users. Collaborative spectrum sensing involves multiple secondary users to make joint decisions about spectrum occupancy. By exploiting multiuser diversity, collaborative sensing can alleviate the effects of hidden terminals and severely shadowed radio environments. In this paper, we investigate and compare two schemes for collaborative spectrum sensing of a narrowband channel based on online parameter estimation of a hidden bivariate Markov model: a hard decision scheme and a soft decision scheme. Relative to prior collaborative sensing approaches that do not incorporate a model of the state of the primary user, the proposed schemes improve the accuracy and reliability of collaborative spectrum sensing, especially in low signal-to-noise ratio environments. Numerical results are presented to demonstrate the performance of the proposed model-based collaborative spectrum sensing schemes.