Energy Efficiency Optimization for MIMO Distributed Antenna Systems With Pilot Contamination

In this paper, we study the energy-efficiency (EE) maximization problem for a multiple-input multiple-output distributed antenna system (DAS) with pilot contamination. With per-user quality of service constraints and per remote antenna unit (RAU) power requirements, we formulate the EE maximization problem as a joint optimization of sparse transmit beamforming, RAU selection, and RAU clustering. The considered problem is a non-convex multivariate optimization problem. To solve the problem, we transform it to an equivalent parametric programming problem (PPP) with a given EE parameter and design a two-layer optimization scheme to solve the original problem. The outer layer involves two kinds of algorithms to iteratively update the EE parameter based on Dinkelbach's algorithm and bi-section search, respectively. The more challenging issue lies in the inner loop, where a non-convex multivariate PPP needs to be tackled. A series of techniques, including the reweighted ℓ₁-norm, D.C. function, and semidefinite relaxation (SDR), is adopted to approximate the non-convex multivariate PPP with a convex SDR problem. Furthermore, a heuristic algorithm is proposed to reduce the complexity of a two-layer scheme. Simulation results show that the proposed algorithms significantly improve the EE and demonstrate that RAU selection and RAU clustering contribute to a higher EE.