Another method of implementing the inverse filter is to use an iterative procedure where an estimate of the signal is undated in each iteration. Let denote the signal estimate after k iterations. If is a good estimate of will be close to . The signal estimate after k+1 iterations , is obtained by adding to a correction term which consists of a scaling constant multiplied to the difference between and . Using as the initial estimate , the iterative procedure is

	(6.4.5a)

	(6.4.5b)

where is a parameter which is positive and which can be used to control the convergence behavior of the iterative procedure.

To see how (6.4.5) is related to the inverse filter, we express (6.4.5) in the frequency domain:

	(6.4.6a)

	(6.4.6b)

Solving for recursively, using (6.4.6), we obtain

		(6.4.7)