Accession Number : ADA195497

Title :   Statistical Analysis of the LMS (Last Mean Squares) and Modified Stochastic Gradient.

Descriptive Note : Final rept. 15 Mar 87-14 Apr 88,

Corporate Author : CALIFORNIA UNIV IRVINE DEPT OF ELECTRICAL ENGINEERING

Personal Author(s) : Bershad,

PDF Url : ADA195497

Report Date : 07 Apr 1988

Pagination or Media Count : 6

Abstract : Research on the stochastic behavior of the last mean squares (LMS) and related algorithms has yielded results in four majors. Significant progress has been made in determining the joint transient and steady state probability density functions of the time domain LMS weights for gaussian jammers. The effects of non-linearities on the time domain LMS algorithm have been analyzed. Keywords: Transient behavior of arbitrary linear filters; Joint probability density functions; Weighting functions; Spread spectrum communications; Adaptive filtering.

Descriptors :   *STATISTICAL ANALYSIS, *ELECTRICAL ENGINEERING, ADAPTIVE FILTERS, ALGORITHMS, COMMUNICATION AND RADIO SYSTEMS, GRADIENTS, LINEAR FILTERING, PROBABILITY DENSITY FUNCTIONS, SPREAD SPECTRUM, STOCHASTIC PROCESSES, TIME DOMAIN, TRANSIENTS, WEIGHTING FUNCTIONS, APPLIED MATHEMATICS

Subject Categories : Statistics and Probability
      Electrical and Electronic Equipment

Distribution Statement : APPROVED FOR PUBLIC RELEASE