A decision directed smart antenna system with neural estimation for M-quadrature amplitude modulated signals

Abstract

Adaptive beamforming and direction of arrival (DOA) estimation are among the prime areas of research which draw the attention of research community to address the ever growing demand for increased capacity in mobile communication systems. Smart antennas, which are essentially adaptive array antennas coupled with intelligent signal processing have emerged as an important category of systems capable of providing solution to this problem. Neural networks are also being widely used for developing intelligent systems and influencing signal processing in communication systems. A decision directed approach is proposed for blind adaptation of smart antenna system using a complex neural estimation of parameters for beamforming. The paper presents a simulation study of the decision directed smart antenna system with neural estimation (DDSA-NE) for multi-quadrature amplitude modulated (M-QAM) signal environment with 4, 16, and 64 constellations. The results indicate the effectiveness of the proposed scheme.