Spectrum sensing techniques in practical cognitive radio applications

Spectrum sensing is one of the core techniques in the cognitive radio network. In this paper, the cyclostationary feature detection for OFDM signals and the MAC-layer sensing-period adaptation algorithm are presented. Exact analytical expressions for spectral correlation function (SCF) are derived applying cyclostationarity fundamentals, and the simulation results demonstrate that the presence of cyclic prefix enhances inherent cyclostationary features yielding a good performance. The sensing-period adaptation algorithm is proposed to maximize the chance of discovering opportunities. The achieved opportunity ratio is analyzed when sensing-period adaptation is employed. Our simulation results indicate that the new proposed scheme has apparent performance improvements over the previously non-adaptive schemes. This improvement may become significantly greater as the initial sensing period grows.