Blind adaptive antenna arrays for mobile communications

Abstract

Adaptive antenna arrays have tremendous potential for increasing the capacity of mobile communications, by reducing co-channel interference, multipath, and noise. Blind adaptive algorithms, that is, algorithms which do not require a training sequence, are investigated and compared in this study. These algorithms are tested for common cellular signals. The performances of three blind adaptive algorithms: the Constant Modulus Algorithm (CMA), the Spectral self-COherence Restoral Algorithm (SCORE), and the spectral correlation predictor using a Time-Dependent Adaptive Array (TDAA), are studied. The TDAA is introduced as a new blind algorithm that exploits the cyclostationary property of the signal. Results show that the TDAA is able to out-perform the other blind algorithms for most of the test conditions and provides the optimal MSE solution.