Joint Long-Term Admission Control and Beamforming in Downlink MISO Networks

Jingran Lin,Qiang Li,Mengyuan Ma

DOI: https://doi.org/10.23919/EUSIPCO.2018.8553613

2018-01-01

Abstract:Admission control has been widely utilized to alleviate network congestion. However, most current studies choose the admissible users based on instantaneous channel information. Due to the time-varying characteristics of wireless channels, the admissible user set changes quickly, which complicates network management and renders heavy operational costs. This motivates us to take the stability of the admissible user set into account in admission control, thus leading to a long-term admission control problem. In this paper, we consider a joint long-term admission control and beamforming problem in a downlink network consisting of one multi-antenna base station (BS) and multiple single-antenna users. To maintain a relatively stable admissible user set and minimize the power cost for the admissible users to achieve their quality-of-service levels, we jointly optimize the admissible users, the BS transmit beamformers, and the switching frequency of each user's admissible status in a given time period. To handle this challenging non-convex problem, we first design a sequential convex approximation (SCA) algorithm to iteratively compute a stationary solution. To facilitate algorithm's implementation, we further employ the alternating direction method of multipliers to come up with an efficient, semi-closed-form update of each SCA problem.