Recursive ALE using gradient-based algorithm

Addresses the application of recursive filters in adaptive signal filtering. Simulations are performed, of the adaptive enhancement of sinusoidal signals embedded in white Gaussian noise, based on LMS algorithm. A digital version of Tellengen´s theorem is used to evaluate the exact instantaneous gradient for parameter updates. The feasibility of implementing IIR-LMS filtering is being studied, and instability is found to be the major problem. The linear gradient approximation is shown to be a satisfactory method, offering stability and simplicity in computation. It is also decided that a second order recursive filter is the suitable form for this application