Robust Analytical ZF Optimization for MISO IFC

Abstract

This letter proposes a fixed-step learning algorithm for the beamforming vectors of the K-user multiple-input single-output (MISO) interference channel (IFC) when only local channel state information (CSI) is available at every transmitter. The algorithm determines the optimal zero-forcing (ZF) beamforming solution to maximize the rate, after K(K-1)(6M-3K+2)/2 steps, in which M(≥ K) is the number of antennas per transmitter. As it requires no crosstalk CSI, it is highly robust to CSI errors.