Cooperative Spectrum Sensing with Censoring of Cognitive Radios in Fading Channel Under Majority Logic Fusion

Abstract

In a cooperative spectrum sensing (CSS) scheme, the detection of the presence of activity of a primary user (PU) is improved by the fact that several cognitive radio (CR) users send, through reporting channels (R-channels), their sensed information on the activity of this PU to a common base station (BS). The benefits are particularly relevant in scenarios where the sensing channels (S-channels) towards the PU of interest of CR users are affected by severe fading or shadowing. However, in a CSS scheme with R channels affected by fading or shadowing as well, there may be erroneous reception, at the BS, of decisions from CR users: this can be counter-acted by using censoring of CR users. In this chapter, we discuss the performance of CSS with censoring of CR users based on their R-channels’ statuses. Two schemes of censoring are considered: (i) rank-based censoring, where a pre-defined number of CR users, associated with the best R-channels, are selected; and (ii) threshold-based censoring, where CR users, whose R-channel fading coefficients exceed a pre-determined threshold, are selected. The performance of both censoring schemes is evaluated considering two different R-channel fading conditions: (i) Rayleigh fading and (ii) Nakagami-\(m\) fading. In both cases, majority logic fusion is considered at the BS (also denoted re-interpreted as fusion center, FC). The impact of various network parameters—such as censoring threshold, number of CR users, average S-and R-channels’ SNRs, channel estimation (CE) quality, and fading severity—on the performance of the considered CSS schemes will be evaluated in terms of missed detection and total error probabilities.