Multichannel System Identification and Detection Using Output Data Techniques. Volume 1
Abstract
The multichannel innovations-based detection algorithm (MIBDA) methodology formulated in Phase I was developed in full and evaluated in the context of two applications: (a) airborne surveillance phased array radar systems, and (b) computer-based electrocardiogram (ECG) diagnosis. A software simulation was developed to generate simulated multichannel phased array radar data. This software was exercised to represent ground clutter as the output of a state variable model (SVM), and to assess the performance of such models in the MIBDA. The MIBDA was modified and extended to the automated diagnosis of ECG traces. MIBDA performance for ECG diagnosis was assessed using the Common Standards for Quantitative Electrocardiography (CSE) database. Analysis and simulation results indicate that the MIBDA offers a valid alternative to conventional systems in both of the applications considered. This suggests also that the MIBDA can be configured to apply to a wide range of problem areas. Volume I presents the MIBDA methodology, SVM identification algorithms, MIBDA design and performance evaluation issues, and simulation results for both applications. Volume II presents the analytic and software model for airborne surveillance phased array radar systems.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1997
- Accession Number
- ADB226867
Entities
People
- Dennis W. Davis
- Jaime R. Roman