Ambiguity-Region Analysis for Double Threshold Energy Detection in Cooperative Spectrum Sensing

Abstract

In this paper we investigate cooperative spectrum sensing technique in which double threshold based energy detection is used at local cognitive radio (CR) sensors. Cooperative spectrum sensing improves reliability in detection of underutilized spectrum by fusing the decisions of local CR sensors. Local CR sensors use two thresholds, λ1 and λ2 (with λ2 > λ1). If the sensed energy is greater than λ2 or less than λ1 then with complete certainty local sensors decide for the presence or absence of primary user (PU) respectively. The difference between λ2 and λ1 defines the ambiguity region, i.e., Δth = λ2 − λ1 in which local CRs are unable to decide the presence or absence of PU. In this work local sensors report the decision along with situation in which sensed energy falls in ambiguity region. In our analysis we use majority rule at the fusion center and compute an analytical expression of average probability of sensing error as a function of width of ambiguity region, i.e., Δth. Using the derived analytical expression and by simulation we show that there exists an optimal value of Δth for which the probability of sensing error is minimum