A Wideband 5G Cyclostationary Spectrum Sensing Method by Kernel Least Mean Square Algorithm for Cognitive Radio Networks

This brief proposes a new cyclostationary method on the basis of the Gaussian Kernel Least Mean Square (KLMS) algorithm for wideband spectrum sensing in the upcoming 5G mobile networks. To focus on a special fragment of the 5G network frequency spectrum, the KLMS algorithm uses a set of inaccurate cyclic frequencies of low complexity belonging to the signal of the primary user (PU). A suboptimal detector is designed on the basis of the KLMS weights and the performance of the proposed detector is measured and compared with other recent spectrum sensing techniques. The method is very easy to implement and less complex to compute than other spectrum-sensing counterparts. Analytical and simulation results have been verified by an experimental 5G communication setup which shows the effectiveness of the proposed method.