Dynamic programming training period for an MSE adaptive equalizer.A steepest descent variable step-size algorithm has been designed using dynamic programming for a mean-square-error adaptive equalizer. Additive noise and a constraint have been included. It is found that the new algorithm converges faster than the common fixed step-size algorithm.
Document ID
19720040128
Acquisition Source
Legacy CDMS
Document Type
Conference Proceedings
Authors
Richman, S. H. (Signatron, Inc. Lexington, Mass., United States)
Schwartz, M. (Brooklyn, Polytechnic Institute, Brooklyn, N.Y., United States)