Maximum likelihood filters in signal detection and signal analysis systems
Abstract
Maximum likelihood estimators specified as power functions of q parameters are studied. The family of filters is generated using Pisarenko's ideas on the spectral estimation on quadratic forms of powers larger than unity. The obtention of optimal order parameters is discussed. It is shown that the characteristics of this kind of filter are comparable and in some cases better than autoregressive filters. The signal power level is more accurately determined and spectral estimations are easier.
- Publication:
-
NASA STI/Recon Technical Report N
- Pub Date:
- 1986
- Bibcode:
- 1986STIN...8719585F
- Keywords:
-
- Digital Filters;
- Maximum Likelihood Estimates;
- Signal Analysis;
- Signal Detection;
- Autoregressive Processes;
- Maximum Entropy Method;
- Optimization;
- Power Spectra;
- Communications and Radar