A region-varying and step size-varying LMS adaptive algorithm based on computational verb rules

This paper presents a region-varying and step size-varying LMS adaptive algorithm based on computational verb rules. This method allows the step-size μ to be adjusted by the error signal e based on the predefined computational verb rules. What´s more, the variable region of μ is adjusted according to the variation of e, and the factor expand and contract is a function with two variables, error signal e and it´s variation ec. Compared with the conventional LMS adaptive filtering algorithm, this algorithm gives attention to the three performance indices below, initial convergent speed, time-varying track ability and constringent precision.