Adaptive filtering in telecommunications
Abstract
A stochastic gradient algorithm is developed for optimizing digital filtering of random observation with statistical unknowns. The mean square deviation between the output of the adaptive filter and the useful signal. The algorithm is applicable for the equalization of high rate data transmission and for echo cancellation. Consideration is also given to the definition of limits for the stochastic iteration, to the convergence of the iteration to an optimum filter which is a solution of the WienerHopf equation, and the use of fixed or decreasing increments in the steps of the iteration to fit the stochastic nature of the algorithm.
 Publication:

Annales des Telecommunications
 Pub Date:
 December 1981
 Bibcode:
 1981AnTel..36..615M
 Keywords:

 Adaptive Filters;
 Data Transmission;
 Digital Filters;
 Iterative Solution;
 Signal Processing;
 Stochastic Processes;
 Convergence;
 Echo Suppressors;
 Equalizers (Circuits);
 Telecommunication;
 Wiener Hopf Equations;
 Communications and Radar