A new transmission strategy for MIMO-WCDMA spatial multiplexing networks based on principal component analysis

The goal of the study presented in this paper is to investigate the performance of a new transmission strategy for MIMO-WCDMA networks operating in spatial multiplexing transmission mode, based on Principal Component Analysis (PCA) at the reception. In this context, all individual multipath components are first processed by a space - time RAKE receiver, and then PCA is employed on the received data matrix. The maximum eigenvalue along with the corresponding eigenvector are fed back to the transmitter, who formulates a modified correlation matrix. This matrix is used in order to determine the significant terms that are taken into account during transmission matrix formulation. As results indicate, for a 4×2 MIMO orientation (i.e. four transmit and two receive antennas) and six resolvable multipath components, the deployment of the proposed strategy can reduce up to 10% the mean number of repeats that are required for convergence of the transmission matrix calculation.