Cyclostationarity-based spectrum sensing with subspace projection

Many communication signals exhibit cyclostation-arity with some cycle frequencies and the cycle frequencies are related to signal parameters such as symbol rates or carrier frequencies. Dandawate proposed an asymptotically optimal chi-square statistical test to check the presence of cyclostationarity. In the last few years, the statistical test has been applied to spectrum sensing for cognitive radios. We notice that, for most modulated communication signals like PSK, QAM, OFDM and GMSK, the cyclic autocorrelations for different delays at the same cycle frequency are deterministically related in average sense. And this property is not considered in Dandawate´s test. So we propose a novel spectrum sensing algorithm in this paper, which improves the performance of cyclostationary feature detection by using the relationship among the cyclic autocorrelations for different delays. The simulation results demonstrate that the improved method has a better performance in an additive white gaussian noise (AWGN) channel and a multipath fading environment.