New adaptive Kalman filters using filter bank

The adaptive algorithms that update adaptive filter coefficients are important in adaptive signal processing. For the adaptive algorithms, the following are required: the reduction of the computational complexity, the simple hardware implementation and so on. For example, the systolic array implementations are known as a method to improve the data throughput. While unlike the above, methods based on the division of the impulse response length are also utilized. In this paper, we propose a new algorithm using filter bank. This method has better properties than those of the traditional method in the convergence property and the computational complexity. The performances of the proposed algorithm are shown by computer simulation.