Energy Detection Performance Enhancement Using RLS and Wavelet De-noising Filters

The fast development in wireless communications and frequency bands assignments for every communication system limits the spectrum resources. Various techniques, for example, cognitive radio have occurred to tackle this issue by allowing unlicensed users to utilize the licensed bands. The most important component for establishing a reliable cognitive radio system is spectrum sensing. One of the ordinarily used spectrum sensing techniques is energy detection. It has low computational and usage complexities. But, for low signal-to-noise ratio (SNR) values it has a poor performance as it will not be able to differentiate the interference from noise and primary users. In this paper, a new energy detection technique for spectrum sensing is introduced. The proposed technique is based on utilization of de-noising filters such as recursive least square (RLS), 1-D wavelet de-noising filter, and 2-D wavelet de-noising filter. This technique is intended to achieve SNR gain, noise variance reduction, and enhance the detection threshold estimation. Furthermore, it exhibits noticeable increase in the throughput rather than that of the traditional detector. Simulation results revealed that the RLS de-noising filter exhibits much better performance than wavelet de-noising filters.