Adaptive methods employing optimal convergence factors for processing complex signals and systems

Abstract

Complex adaptive methods for complex information processing employ optimal individual convergence factors for real and imaginary components of the weight vector. For wireless receivers operating on QPSK, a Complex IA-ICA performs better than existing Complex Fast-ICA methods in terms of accuracy and convergence speed, can process such complex signals in time-varying channels, and employs time-varying and time-invariant convergence factors, independent for the real and imaginary components of the system parameters, and provide individual or group system parameter adjustments. Such systems employ the within complex adaptive ICA with individual element adaptation (Complex IA-ICA). In adaptive beamforming, system identification and other adaptive systems based on the Least Squares method, complex least mean square methods, with optimally and automatically derived convergence factors, are employed and which perform much better in terms of convergence speed and accuracy, when compared to the traditional Complex LMS and Block Complex LMS methods.