Cooperative Prediction-and-Sensing-Based Spectrum Sharing in Cognitive Radio Networks

• Published in: IEEE transactions on cognitive communications and networking Vol. 4; no. 1; pp. 108 - 120
• Main Authors: Nguyen, Van-Dinh, Shin, Oh-Soon
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.03.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Approximation algorithms; Array signal processing; Beamforming; Channel estimation; Cognitive radio; Computer simulation; Data transmission; Detection; Interference; nonconvex programming; opportunistic spectrum access; Optimization; prediction accuracy; Prediction algorithms; Radio networks; Sensors; spectrum sensing; spectrum sharing; spectrum underlay; sum rate; transmit beamforming
• ISSN: 2332-7731; 2332-7731
• DOI: 10.1109/TCCN.2017.2776138

This paper proposes prediction-and-sensing-based spectrum sharing, a new spectrum-sharing model for cognitive radio networks, with a time structure for each resource block divided into a spectrum prediction-and-sensing phase and a data transmission phase. Cooperative spectrum prediction is incorporated as a sub-phase of spectrum sensing in the first phase. We investigate a joint design of transmit beamforming at the secondary base station (BS) and sensing time. The primary design goal is to maximize the sum rate of all secondary users (SUs) subject to the minimum rate requirement for all SUs, the transmit power constraint at the secondary BS, and the interference power constraints at all primary users. The original problem is difficult to solve since it is highly nonconvex. We first convert the problem into a more tractable form, then arrive at a convex program based on an inner approximation framework, and finally propose a new algorithm to successively solve this convex program. We prove that the proposed algorithm iteratively improves the objective while guaranteeing convergence at least to local optima. Simulation results demonstrate that the proposed algorithm reaches a stationary point after only a few iterations with a substantial performance improvement over existing approaches.