HMM based cooperative spectrum occupancy prediction using hard fusion

Abstract

Spectrum occupancy prediction provides cognitive radio secondary users with proactive ability to exploit spectrum opportunities. When feasible, spectrum sensing outcome can be further analysed to assist secondary users to pre-emptively avoid interfering with primary users. Where applicable, cooperative spectrum prediction in multi-user environment has the potential to overcome limitations of local spectrum prediction accuracy. In this paper, we provide a thorough analysis on hard fusion based spectrum occupancy prediction. We present accuracy enhancement in terms of prediction error. The analysis and simulation results showed the improvement in prediction accuracy, and the required threshold..