Online tracking of coherent signals over multiple-input multiple-output code division multiple access communication channel using JADE–Capon method in conjunction with iterative space–time lifting preprocessing scheme

Abstract

The main motivation for this work is how to handle the circumvent problem encountered in decorrelating perfect coherent multipath signals over MIMO-CDMA channel with small number of sensors and few data snapshots. The authors propose a new iterative space–time lifting preprocessing (ISTLP) decorrelating scheme that guarantees a perfect decorrelating of a great number of impinging signals even when they are fully correlated and overlaying in space and time domains. It is based on exploiting iteratively the spatial and temporal redundancies arising from multiple antennas and multiple slots. The temporal lifting scheme is introduced first, for efficient tracking and cancelling of ISI, MAI and noise effects within the bank of the N-tapped-delay-lines associated with the user of interest. Secondly, this iterative multi-resolution approach will be coupled to a preprocessing scheme to allow easy transformation of resulting temporal-lifted covariance matrix from time domain to an equivalent Vandermonde structure in frequency domain to be exploited in spatial lifting step. Two detection criteria are therefore established to attempt the convergence state in ISTLP algorithm. Finally, a JADE–Capon method based on ST-MVDR beamformer is applied to the spatial–temporal-lifted preprocessed covariance matrix to compute the joint azimuth-direction-of-arrival and delay of desired Ki multipath signals. The efficacy and high resolution capability of the proposed estimation approach is supported by numerical simulation results.