Quantum-inspired Tabu Search algorithm for antenna selection in massive MIMO systems

Abstract

Massive Multiple-Input Multiple-Output (MIMO) systems can significantly improve the system performance and capacity by using a large number of antenna elements at the base station (BS). However, having a massive number of radio-frequency (RF) chains at the BS can be costly and energy inefficient. One way to achieve the diversity gain of massive MIMO systems is to employ a massive number of antennas with limited number of RF chains. Thus, antenna selection techniques can be applied to reduce the system complexity and hardware cost. In this paper, a Quantum-inspired Tabu Search (QTS) algorithm is applied to antenna selection in Massive MIMO systems and compared with two well known algorithms; namely, a Classical Tabu Search (CTS) algorithm and a Genetic Algorithm (GA). The QTS algorithm has a great advantage over CTS, since it only requires finding the optimum rotation angle to evolve the system towards a better solution. In contrast, in CTS, the dimensions of the tabu matrix are dynamic and need to be optimized. Moreover, to achieve maximum performance, these dimensions need to be reconfigured when changing the number of antennas or the number of iterations, while no such a problem occurs in the QTS. The QTS algorithm also shows better results in terms of the system capacity compared to CTS and GA. Furthermore, the classical and quantum inspired TS algorithms require much lower complexity than the GA.