Space-time processing with channel knowledge at the transmitter

We investigate how knowledge of the channel at the transmitter can improve performance in array-to-array wireless communications. We explore the practical issues associated with space-time processing based on a singular-value decomposition (SVD) of the channel matrix. Compared to space-time block codes that do not require channel knowledge at the transmitter, the proposed method without channel coding requires 8.3 dB less SNR per antenna to achieve 6 bits/sec/Hz with 4 transmit and 4 receive antennas. Exploiting the properties of the SVD, we propose fixed bit and power allocations that avoid dynamic allocations. We also describe how reciprocity in a time-division duplex system may be exploited to simplify implementation of the SVD processing.