A Comparison of Two Noise Canceling System Algorithms.
Abstract
The purpose of this study is to compare two noise canceling system algorithms when the number of inputs to the system is small. The expected value algorithm is an estimate of the optimum solution for minimum mean square error analysis. The LMS (least mean square) algorithm was developed to avoid the matrix inversion inherent in the optimum solution given by the Wiener-Hopf equation. The comparison between the two algorithms is measured by the rate of convergence of the mean square error and the minimum error after five hundred iterations. A computer simulation of a dc signal in Gaussian noise provided the comparison between the two algorithms. The tradeoff in terms of performance for simplicity in the LMS algorithm was discovered to be great. The expected value algorithm consistently produced better results in terms of faster rates of convergence and smaller minimum errors after five hundred iterations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1977
- Accession Number
- ADA055822
Entities
People
- Rachel Patricia Messina
Organizations
- Duke University