Performance of Cooperative Spectrum Sensing Based on Random Transition of the Primary User in Laplacian Noise

In this paper, cooperative spectrum sensing (CSS) of dynamic primary user (PU) is considered in Laplacian noise environment. The dynamic PU is characterized by its transitions from ON (present) state to OFF (absent) state and vice-versa. It means, during the entire sensing duration, the PU appears or disappears intermittently. We assume that each cognitive radio (CR) uses conventional test-statistics such as energy detection (ED), absolute value cumulation detection (AVCD) and improved AVCD (i-AVCD). The hard decision from each CR fuses at the fusion center (FC) according to CSS based on OR rule (CSS-OR), CSS-AND rule and CSS-majority rule to make a final decision on the appearance or disappearance of the PU. We further consider dynamic nature of the PUin terms of its arrival rate (theta(A)) and departure rate (theta(D)). We present performance of the CSS of dynamic PU using receiver operating characteristic (ROC) and detection probability (P-D) versus average signal-to-noise ratio (SNR), denoted by gamma, using Monte Carlo simulations. We conclude that the CSS-OR rule based spectrum sensing outperforms CSS-majority rule and CSS-AND rule based spectrum sensing over a wide range of average SNR, i.e., -10 < gamma < 10 dB. We further conclude that CSS-AND rule is unsuitable for enhancing the detection probability of conventional sensing schemes. Furthermore, CSS-majority rule outperforms conventional sensing schemes ED, AVCD and i-AVCD beyond gamma = - 1, - 5 and -6 dB, respectively.