On the Performance of Cooperative Spectrum Sensing Under Binary Erasure Reporting Channels

Abstract

The spectrum sensing based on single detector suffers from adverse effects such as fading, shadowing and hidden node problems. Cooperative spectrum sensing (CSS) is proposed to overcome these effects by utilizing the spatial diversity among the spatially located cooperating secondary users (CSUs). In traditional centralized CSS, the CSUs share their detection results with the fusion center (FC). Based on the information received from the CSUs, the FC analyze the information and takes the final decision on the occupancy of the primary user (PU) channel. Generally, in literature, the reporting channel between CSUs and the FC is assumed to be error free. However, these reporting channels are wireless channels which undergo fading and shadowing resulting in imperfect channels. In this work, we propose a model based on binary erasure channel (BEC) which is more suitable to model the reporting channel between the CSUs and the FC. We then, analyze the performance of the CSS under imperfect reporting channels. We then analyze the performance of repetition code based algorithm to improve the performance of CSS under imperfect reporting channels. It is found that the repetition code based algorithm significantly improves the performance. Also, by properly choosing the repetition length in the repetition based algorithm, the effects of reporting channel errors can be completely removed.