Linear Precoding in the Intelligent Reflecting Surface Assisted MIMO Broadcast Channel

Dominik Semmler, Michael Joham, Wolfgang Utschick

Abstract: We propose efficient algorithms to solve the sum-rate maximization in the Intelligent Reflecting Surface (IRS) assisted Multiple-Input Multiple-Output (MIMO) Downlink (DL) scenario. The recommended methods are based on Linear Successive Allocation (LISA), a well performing linear precoding algorithm for the traditional MIMO DL. Taking LISA as a basis, we can exploit its characteristic zero-forcing structure which allows to obtain a special form of alternating optimization. This special form enables a quick convergence and we observe a reduced iteration number together with a good performance of the proposed methods in the simulations.