Interference Management and User Association for Nested Array-Based Massive MIMO HetNets

• Published in: IEEE transactions on vehicular technology Vol. 67; no. 1; pp. 454 - 466
• Main Authors: Feng, Mingjie, Mao, Shiwen
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.01.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: 5G wireless; Antenna arrays; Antennas; Array signal processing; Computer architecture; Degrees of freedom; Direction-of-arrival estimation; heterogeneous networks (HetNet); Integer programming; Interference; interference nulling; Linear programming; massive MIMO; MIMO; nested array; Radio equipment; Upper bounds
• ISSN: 0018-9545; 1939-9359
• DOI: 10.1109/TVT.2017.2741900

The nested array, implemented by nonuniform antenna placement, is an effective approach to achieve O(N 2 ) degrees of freedom (DOF) with an antenna array of N antennas. Such DOF refers to the number of directions of incoming signals that can be resolved. With the increased number of DOF, an important application of nested array is to nullify interference signals from multiple directions. In this paper, we apply nested array in a massive multiple input multiple output (MIMO) heterogeneous network (HetNet) for interference management. With a nested array, a base station (BS) can nullify a certain number of interference signals based on their directions. Then, a key design issue is how to select the set of interference sources to be nullified at each BS. As the DOF of each BS is used to resolve both desired signals and interference, the number of interference signals that can be nullified depends on the number of users served by the BS. Thus, user association is another factor that impacts the system performance and should be jointly considered with interference nulling. We formulate the joint interference nulling scheduling and user association problem as an integer programming problem, aiming to maximize the sum rate of all users subject to BS DOF constraints. We first investigate the case of interference nulling with a given user association, and propose a scheme to solve a relaxed problem as well as derive a performance upper bound. Then, we propose a distributed joint interference nulling and user association scheme based on a poly matching between users and BSs. Simulation results show that the proposed schemes effectively improve the sum rate and achieves a near optimal performance.