Kalman filter for adaptive antennas

Abstract

Adaptive control algorithm is a key technique for an adaptive array. It is necessary to find a fast and efficient algorithm for gaining rapid interference suppression. Convergence speed of conventional gradient-based algorithms is extremely slow and very sensitive to eigenvalue spread of autocorrelation matrix. This paper presents the Kalman filtering method to adaptive array processing. It has good transient response and achieves faster convergence speed, so it is most suitable for complicated adaptive systems. We analyzed its convergence performance which proved to be related to initial value forP(0). Computer simulation by applying four elements array to null steering was carried out. It is shown that fast convergence needs only 2M numbers of iteration. The less the value was selected forP(0), the slower convergence was found. Further, ifP(0) were too small, the method would not be able to converge. On the other hand, bigger values ofP(0) can not achieve more improvement in convergence performance. Simulation also discovered steady response denoted with SNIR dependent on the power of reference signal. These results demonstrate the validity and effectiveness of the Kalman type adaptive antenna.