Pseudo Bounds for Deterministic Maximum Likelihood (DML) Direction Finding Errors in Correlated Noise Environments
Model based array parametric direction nding algorithms are among the most accurate techniques known of determining the direction to a source. The deterministic 1 maximum likelihood (DML) algorithm is one such technique that determines the direction of a set of narrow-band sources by tting the recei...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Text |
| Language: | English |
| Subjects: | |
| Online Access: | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.33.8578 http://www-i.mitre.org/centers/wc3/asto/docs/Papers/Technical/DFerrors.pdf |
| Summary: | Model based array parametric direction nding algorithms are among the most accurate techniques known of determining the direction to a source. The deterministic 1 maximum likelihood (DML) algorithm is one such technique that determines the direction of a set of narrow-band sources by tting the received sensor data to a statistical model. For many algorithms such as DML, the sensor data is modeled as a uniform and continuous spatial background with a set of discrete point sources. When the assumed model is a good characterization of the received data, the model is said to be matched. In practice there is often, if not always, some degree of model mismatch encountered with real data because models are inherently idealized constructions. The eects of mismatch on parameter estimation algorithms is an important issue to system designers. In this paper we specically examine model mismatch in DML when the background noise is spatially correlated and unknown. In correlated background noi. |
|---|