The analysis obtained originally in 1968 by the author is extended to provide precise simple formulas for the solutions of weight covariance, mean norm-squared error, misadjustment, etc. In addition, new simple necessary and sufficient conditions for convergence are displayed. These results yield simple relationships governing the convergence rate of the weight covariance of the adaptive linear processor as well as a variety of good bounds on misadjustment. (Author).

Project 7904. Extension of the author's thesis, Adaptive linear discrete-time estimation, Stanford University, 1968. Also available as AD712101.