Adaptive spatial multiplexing for time‐varying correlated MIMO channels

Abstract

AbstractThis work presents an adaptive spatial multiplexing scheme for time‐varying correlated multiple‐input multiple‐output (MIMO) channels. Unlike the conventional spatially uncorrelated block fading channel model, both temporal channel variation in each data block and spatial correlation are considered in the design. The proposed system continuously transmits packets with identical structure. Each packet consists of a training phase followed by a data transmission phase. The training signal is used to estimate instantaneous channel state at the receiver. In each symbol period of data transmission phase, the transmitter simultaneously transmits multiple data streams, which are decoded at the receiver based on the estimated channel state. The transmitter is assumed to know only the channel statistics, with which the power and rate for each data stream as well as the number of data streams are adjusted in each symbol period to achieve a target bit error rate (BER). The results reveal that the throughput‐maximising transmission strategy makes a judicious diversity‐multiplexing tradeoff by allocating power to an optimum number of data streams. Moreover, the packet length is also optimised based on the temporal channel variation rate. The simulation results demonstrate that the throughput can be significantly enhanced by the optimum packet length. Copyright © 2009 John Wiley & Sons, Ltd.