Development and Analysis of ARMA Parameter Estimation Schemes in the Presence ofNoise

摘要

The work done and the number of subjects studied goes beyond what was promised inthe original two-year proposal. The proposal focused on single-sensor parameter estimation for ARMA signals in noise. We solved several open problems in this area and added sine wave noise. We extended the work to sensor array estimation algorithms for localization (tracking applications). The music estimator was developed and compared to the maximum-likelihood estimator. Closed form expressions for the Cramer-Rao bound weve discovered for certain cases of significance also is the thesis of David Starer written with AFOSR support, that provides up-dating of the roots of an nth order polynomial in O(u-sq) time, and numerous applications.