Three Steps Toward Low-Complexity: Practical Interference Management in NOMA-Based mmWave Networks

Abstract

Beamforming, user scheduling and transmit power on existing interference management schemes in multi-cell mmWave networks have been independently controlled due to the high computational complexity of the problem. In this paper, we formulate a long-term utility maximization problem where beam activation, user scheduling and transmit power are incorporated in a single framework. To develop a low-complex algorithm, we first leverage the Lyapunov optimization framework to transform the original long-term problem into a series of slot-by-slot problems. Since the computational complexity to optimally solve the slot-by-slot problem is even significantly high like existing schemes, we decompose the problem into two different time scales: ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i</i> ) a subproblem to find beam activation probability with a long time-scale and ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ii</i> ) a subproblem to find user scheduling and power allocation with a short time-scale. Moreover, we introduce two additional gimmicks to more simplify the problem: ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i</i> ) sequentially making decisions of beam activation, user scheduling, and power allocation, and ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ii</i> ) considering a critical user for power allocation. Finally, via extensive simulations, we find that the proposed CRIM algorithm outperforms existing algorithms by up to 47.4% in terms of utility.