Performance of Spectrum Sensing and Optimization Based on User Selection in Cognitive Radio
The paper first proposes a fast novel spectrum sensing algorithm for cognitive radios based on cyclic autocorrelation. When only the existence of primary users in noise is detected, special cyclic frequency can be choose to sense, which will significantly reduce the computational cost in applying the cyclo stationarity detection. The paper also proposes to select the users with good detection performance for cooperative sensing so as to improve sensing sensitivity. It demonstrates that the throughput of CR system is also improved by user selection.