Spectrum Sensing Based on Dynamic Primary User with Additive Laplacian Noise in Cognitive Radio

Abstract

In this paper, spectrum sensing techniques with dynamic primary user (PU) are considered in the environment of Laplacian noise. It means the PU may not be present or absent during the whole sensing time. However, PU arrives or departs randomly in the sensing time interval. We consider three different detection schemes such as energy detection, absolute value cumulation detection (AVCD) and improved AVCD (i-AVCD). We present the performance in terms of receiver operating characteristic (ROC) and detection probability versus average signal-to-noise ratio (SNR) using simulations. We conclude that the detection performance in the dynamic scenario is better than the performance in the static scenario, when the arrival/departure parameter (\(\theta _{A}T\)/\(\theta _{D}T\)) is beyond one, where \(\theta _A\) and \(\theta _D\) are corresponding to arrival rate and departure rate of the PU respectively, and T is the sampling interval. Furthermore, the i-AVCD scheme outperforms AVCD and energy detection in the considered scenario.