Precoding and spreading for coded MIMO-OFDM

Abstract

In this paper, precoding in combination with spreading for coded transmission using MIMO-OFDM is investigated. Although MIMO-OFDM systems provide both spatial and frequency diversities that can be exploited by the transmitter, e. g. through spreading, they often suffer from the presence of antenna correlations which can degrade the system performance. Full channel knowledge is usually assumed at the receiver and none at the transmitter. In this paper, partial channel knowledge (transmit correlations) is assumed to be available at the transmitter and is exploited by the precoder. Spreading is also applied at the transmitter to extract the MIMO-OFDM diversity either fully or partially. In addition, coded transmission using convolutional codes is assumed. At the receiver, iterative equalization and decoding is employed assuming full channel knowledge. Two sub-optimum equalizers, soft Cholesky equalizer (SCE) and recurrent neural network equalizer (RNN), are used for equalization. Through the used of EXIT-charts, the performance of the MIMO-OFDM system with both precoding and spreading is analyzed.