Reduced-complexity methods for throughput maximization in MIMO channels

This paper focuses on how computational complexity associated to an algorithm for the maximisation of throughput in a multiple-input multiple-output system can be reduced. Throughput optimisation is achieved by selecting the appropriate subset of antennas on the transmit side. Given the highly non-linear nature of the throughput expression and the finite set of combinations, a straightforward approach consists in checking every single subset. However, this results in a prohibitive complexity. This paper presents two methods, named top-down and bottom-up approaches, that lower computational burden by recursively obtaining throughput expressions and, thus, avoiding unnecessary re-computations. The resulting schemes are compared in terms of number of required floating-point operations.