A least-squares system has an adjustable “bias weight,” which causes the constant c to be added to the filter output, as shown in Figure 8.28. The bias weight is used in cases where one of the signals...

A least-squares system has an adjustable “bias weight,” which causes the constant c to be added to the filter output, as shown in Figure 8.28. The bias weight is used in cases where one of the signals has a nonzero mean value and the other does not. Suppose E[f_k] = µ_f, E[d_k
] = 0, B(z) is a highpass filter, and the weight vector, b, is optimized. What is the optimal value of c in terms of μ_f
?