Joint Power Allocation and User Scheduling for Device-to-Device-Enabled Heterogeneous Networks With Non-Orthogonal Multiple Access

Device-to-device (D2D) communications-enabled dense heterogeneous networks (HetNets) with non-orthogonal multiple access (NOMA) are promising solutions to meet the high-throughput requirement and support massive connectivity. In this paper, we propose a novel framework on the D2D-enabled HetNets with NOMA, where small cells underlay the uplink spectrum of macrocells to make full use of spectrum resources, NOMA technique is invoked to serve more downlink users simultaneously, and D2D-enabled multi-hop transmission is established to enhance signal reception of the far users (FUs) on cell edge. We investigate joint power allocation and user scheduling to maximize the ergodic sum rate of the near users (NUs) in the small cells while guaranteeing the quality-of-service requirements of the FUs and the macro-cell users. The optimal solution to this problem is complexity-prohibitive especially with large numbers of users and small base stations (SBSs) because it requires an exhaustive search over all possible combinations of SBSs, NUs, and FUs. To simplify the solution, we develop a two-step approach by decomposing the original problem into a power allocation problem and a user scheduling problem. We derive the closed-form solution of the power allocation problem via analyzing the objective function and constraints. The user scheduling problem is a joint user pairing and access point assignment problem. To solve it, we propose an SBS-NU-FU matching algorithm to obtain a near-optimal one-to-one three-sided matching of SBSs, NUs, and FUs. The simulation results show that the two-step method gets around 95% of the system throughput of the optimal one and can significantly improve the spectral efficiency of the D2D-enabled HetNets.