A Novel Decision Fusion Periodogram-based Algorithm for Centralized Cooperative Spectrum Sensing Under Errors at the Report Channel

Spectrum sensing is the main task of cognitive radios (CRs) to detect idle bands for opportunistic use. In cooperative and centralized mode, CRs sense a target band and send data to a fusion center for a final decision upon its occupancy. It can be raw data (soft decision) or local decision (hard decision) on the channel occupation state in each CR. Soft decision schemes typically achieve better performances but decision fusion is unbeatable in terms of traffic. This paper proposes a novel periodogram power spectral density estimate-based centralized cooperative spectrum sensing algorithm for decision fusion and makes a performance-traffic-trade-off with the circular folding cooperative power spectral density split cancellation soft decision algorithm, both under errors at the report channel. Results show that the proposed scheme is sensitive to errors but encoding can easily overcome it, still keeping the data traffic smaller than the quantized soft decision scheme with no coding. It can be concluded that a case-by-case analysis, performance versus data traffic, must be made in order to elect the best scheme for a given scenario.