Joint Optimization of Sensing Time and Fusion Rules for Cognitive Radios

Abstract

 In Cognitive Radios, sensing-time and fusion rules affect the performance of spectrum sensing when cooperative sensing is applied. Besides, the more unlicensed users are involved in cooperative sensing, the higher spectrum utilization the channel can achieve, while from the unlicensed users’ perspective, the lower average throughput the unlicensed users can obtain. In this paper, we explore the issue on the sensing-time and the fusion rules to optimize the average throughput of the unlicensed users under the constraint that the licensed users are sufficiently protected. At first, we formulate this issue as an optimization problem, and showed the unimodal characteristics of the unlicensed users’ average throughput as a function of the sensing-time and the fusion parameter. Then a numerical optimization algorithm is proposed to obtain the optimal solution. At last, by theoretical analysis and performance comparisons we derived the optimal fusion rule selection scheme under different scenarios. Computer simulations show that significant improvement in the average throughput of unlicensed users is achieved when the sensing-time and the fusion rule are jointly optimized.

DOI : http://dx.doi.org/10.11591/telkomnika.v12i4.4793