Suppose that we have a desired signal characterized by a second-orderAR process, that is, with e(t) zero-mean, white of variance Ree. Assume that we are to design a two-weight, all-pole stochastic...

Suppose that we have a desired signal characterized by a second-orderAR process, that is,

with e(t) zero-mean, white of variance Ree. Assume that we are to design a two-weight, all-pole stochastic gradient processor characterized by

(a) In terms of the AR process parameters calculate the covariance matrix, Rxx .

(b) Determine the eigenvalue spread of Rxx and the corresponding eigen-vectors (normalized to unit length).

(c) Calculate the minimum mean-squared error and sketch a plot as a function of eigenvalue spread. What is the overall conclusion?