Robust scheduling in MIMO-OFDM multi-user systems based on convex optimization

Abstract

This paper proposes a new technique in the context of multi-user scheduling in orthogonal frequency division multiplexing (OFDM) and multi-input-multi-output (MIMO) systems by considering a design that is also robust to channel uncertainties. The whole technique is divided into two parts: i) a robust design for each particular MIMO link exploiting an imperfect channel state information (CSI) and ii) a multi-user design for the scheduling policy. The link between both stages is practical, intuitive, and simple. The robust MIMO design is based on a worst-case approach that maximizes via convex optimization procedures the worst possible signal to noise ratio (SNR) given a specified channel uncertainty region. The scheduling policy, also calculated with convex optimization techniques, minimizes the total number of utilized subcarriers. Therefore, the design is scalable in the sense of having the capability to incorporate new future potential users.