Resource Allocation in Heterogeneous Cognitive Radio Network With Non-Orthogonal Multiple Access

• Main Authors: Woping Xu, Runhe Qiu, Xue-Qin Jiang
• Format: Article
• Language: English
• Published: IEEE 2019-01-01
• Series: IEEE Access
• Subjects: Cognitive heterogeneous network; interweave spectrum sharing mode; kmeans clustering; non-orthogonal multiple access (NOMA); resource allocation
• Online Access: https://ieeexplore.ieee.org/document/8703419/

In this paper, we study resource allocation problems for a two-tier cognitive heterogeneous network in interweave spectrum sharing mode. Secondary users (SUs) in small cells (SCs) opportunistically access the licensed spectrum resources. Non-orthogonal multiple access (NOMA) is used to boost the number of accessible SUs sharing the limited and dynamic licensed spectrum holes. Practically, there exists a tradeoff: an SC can increase its instantaneous sum throughput by accessing more idle bandwidth, which creates higher liability due to the dynamics of licensed spectrum and contention among the multiple SCs. Aiming to maximize the sum throughput of second-tier SCs network, we formulate a mixed integer non-linear programming problem with the constraints of the available idle bandwidth, the successive interference cancellation complexity, the transmission power budget, and the minimum data requirements. To efficiently solve this problem, we decompose the original optimization problem into bandwidth resource allocation subproblem, SUs clustering subproblem, and power allocation subproblem. Based on the scale of SCs network and the activities of licensed spectrum, we introduce an optimal bandwidth configuration to maximize the average sum throughput of SCs. By analyzing the derivation of the achievable rate expression of a NOMA-enabled SU, we develop a novel SUs clustering algorithm which can improve the throughput of a cluster by grouping SUs with more distinctive channel conditions. With the results of SUs clustering, we propose power allocation within a NOMA cluster by using Karush-Kuhn-Tucker optimality conditions. Furthermore, we perform power allocation across NOMA clusters by using the difference of convex programming. The simulation results validate the performance of the proposed resource allocation algorithms.