Impact of Heterogeneous Fading Channels in Power Limited Cognitive Radio Networks

• Published in: IEEE transactions on cognitive communications and networking Vol. 4; no. 1; pp. 1 - 14
• Main Authors: Shaojie Zhang, Hafid, Abdelhakim Senhaji, Haitao Zhao, Shan Wang
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.03.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Channels; cluster based particle swarm optimization; Clustering algorithms; Clusters; Cognitive radio; Cognitive radio networks; Computer simulation; Data transmission; Design optimization; Electric power distribution; Fading; Fading channels; heterogeneous fading channels; Interference; Particle swarm optimization; Power consumption; Power demand; Radio networks; Sensors; Throughput; Tradeoffs
• ISSN: 2332-7731; 2332-7731
• DOI: 10.1109/TCCN.2017.2779858

The tradeoff between decreasing the interference to primary user (PU) and increasing secondary users' (SUs') achievable throughput is an important problem in cognitive radio networks. Heterogeneous fading channels from PU to multiple SUs, PU's traffic distribution, limited SU's power and multiple SUs' access contention impact both these two conflicting objectives. In this paper, we study the joint impact of these four factors on the tradeoff. More specifically, we consider that the channels from PU to SUs are exposed to non-identically independent free space path losses, PU's traffic randomly arrives and departs from the channel, every SU's average power consumption is limited, while multiple SUs contend to transmit. We first model the impact of these factors on SUs' spectrum sensing and data transmission. Then, we formulate the tradeoff aiming at maximizing SUs' aggregated throughput under two constraints: 1) interference probability to PU and 2) SUs' average power consumption. To solve the optimization problem, we design a novel cluster based particle swarm optimization (C-PSO) algorithm. By iteratively updating the particles in a cluster based on the comparison of their fitnesses, the cluster converges to the optimal solution rapidly. Simulation results validate the feasibility of the C-PSO algorithm and the outperformance of our proposal compared against related contributions which consider the homogeneous fading channel. They also show how the optimal solution varies with path losses and PU's traffic distribution.