Low Complexity NOMA Design for Millimeter-Wave Massive MIMO System

Abstract

Massive machine type communications (mMTC) will be a killer use case of fifth-generation (5G) wireless systems using millimeter-wave (mmWave) band Since 5G will be characterized by the beam-based air interface a critical challenge is to serve massive users using available beams Due to the sparsity of mmWave channels several users might receive the strongest power from the same beam and then introduce severe intra-beam interference (intra-BI) While intra-BI can be mitigated through full digital precoding the required number of RF chains is identical to the number of antennas resulting in high power consumption using full digital precoding By selecting a set of beams and applying non-orthogonal multiple access (NOMA) to properly allocate transmit power to each beam and user the sum rate can be maximized using a smaller number of RF chains Since the problem of joint beam selection and power allocation is NP-hard and thus difficult to solve we decompose the problem into two sub-problems i e beam selection and power allocation Both sub-problems are solved by deterministic algorithms with much lower complexity than exhaustive search Besides we exploit some celebrated meta-heuristic approaches including generic algorithm (GA) particle swarm optimization (PSO) and simulated annealing (SA) to efficiently search for good solutions Simulation results show that the performance of the proposed deterministic algorithms are comparable to that of a sub-optimal algorithm based on convex-relaxation power allocation (CRPA) and exhaustive beam selection (EBS) but with much lower complexity On the other hand meta-heuristic methods can also delivery promising and sometimes better results than CRPA and exhaustive beam selection The proposed approaches can relax the hardware burden of beam-based air interface and achieve close-to-optimal performance with low complexity and thus can be useful to support massive connectivity in 5G

Date of Award	2018 Jun 26
Original language	English
Supervisor	Kuang-Hao Liu (Supervisor)