Steady-state performance analysis for adaptive filters with error nonlinearities

摘要

A unified approach to the steady-state mean square error (MSE) and tracking performance analyses for real and complex adaptive filtes with error nonlinearities is developed. Some general clofied-form analytical expressions for the steady-state performances are given. Our analyses are based on Taylor series expansion and and so-called complex Brandwood-form series expansion (BSE). Under these general explicit expressions, some well-known adaptive filters can be viewed as special cases. In addition, the closed-form analytical expressions for the steady-state performance for real and complex least-mean p-power (LMP) algorithm with different choices of parameter p are also given. A mass of simulations show the accuration of our analyses.