Joint Base Station Clustering and Beamforming for Non-Orthogonal Multicast and Unicast Transmission With Backhaul Constraints

• Published in: IEEE transactions on wireless communications Vol. 17; no. 9; pp. 6265 - 6279
• Main Authors: Chen, Erkai, Tao, Meixia, Liu, Ya-Feng
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.09.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Array signal processing; Beamforming; Branch & bound algorithms; branch-and-bound (BB); Cellular communication; Cellular networks; Clustering; Clustering algorithms; Computer simulation; convex-concave procedure (CCP); Division; Layered-division multiplexing (LDM); MIMO communication; Multicast; Multiplexing; non-orthogonal multicast and unicast transmission; Simulation; sparse beamforming; Unicast
• ISSN: 1536-1276; 1558-2248
• DOI: 10.1109/TWC.2018.2858223

The demand for providing multicast services in cellular networks is continuously and fastly increasing. In this paper, we propose a non-orthogonal transmission framework based on layered-division multiplexing (LDM) to support multicast and unicast services concurrently in cooperative multi-cell cellular networks with a limited backhaul capacity. We adopt a two-layer LDM structure where the first layer is intended for multicast services, the second layer is for unicast services, and the two layers are superposed with different beamformers. Each user decodes the multicast message first, subtracts it, and then decodes its dedicated unicast message. We formulate a joint multicast and unicast beamforming problem with adaptive base station clustering that aims to maximize the weighted sum of the multicast rate and the unicast rate under per-BS power and backhaul constraints. To solve the problem, we first develop a branch-and-bound algorithm to find its global optimum. We then reformulate the problem as a sparse beamforming problem and propose a low-complexity algorithm based on convex-concave procedure. Simulation results demonstrate the significant superiority of the proposed LDM-based non-orthogonal scheme over orthogonal schemes in terms of the achievable multicast-unicast rate region.