Exploiting cognitive cyclostationary noise subspace for noncognitive spectrum sensing

This paper addresses the problem of spectrum sensing performed by a silent non intended cognitive receiver in the presence of an active cognitive transmitter. Because the silent receiver does not decode the information, the transmitted signal acts as strong interference to the spectrum sensing process. We propose a method that exploits the statistical features of cyclostationarity manifested in the cognitive signal. The silent cognitive user is able to identify a noise subspace where the noncognitive spectrum sensing is performed free of interference. We employ deflection as a theoretical performance indicator, while numerical results are employed as benchmarking and to corroborate the theoretical results.