CONSTRAINT ADAPTIVE NATURAL GRADIENT ALGORITHM (CANA) FOR ADAPTIVE ARRAY PROCESSING

摘要

A novel Constraint Adaptive Natural Gradient Algorithm (CANA) is proposed based on the Natural Gradient technique, which is capable of rapidly adjusting the response of an array of sensors to a signal coming from Direction of Interest (Dol) and suppressing noises coming from other directions. Constrained optimization techniques have been extended to enhance the look direction signal in the presence of interference using adaptive natural gradient techniques. Mathematical analysis and realistic MATLAB simulations are developed to confirm the fast convergence of the adaptive weights, and the suitability of algorithm for operations in adverse environments.