Performance Analysis of Subspace Methods
Abstract
The main focus of this research was determining the accuracy of subspace based methods for estimating the Direction of Arrival (DoA) of multiple sources from measurements obtained at the output of a sensor array. Subspace methods like MUSIC (MUltiple SIgnal Classification), ESPRIT (Estimation of Signal Parameters via Rotational Invariant Techniques), the Minimum-Norm methods have recently received much attention, and their estimation accuracy as well as a rigorous comparative study is of much interest. This was the goal of this research. Of particular interest was the affect of spatial smoothing on the performance of the subspace methods. Spatial smoothing is useful in dealing with coherent sources and for the possible enhancement of the performance of the methods. Also, examined were the implementation issues associated with these methods. As opposed to implementing a single algorithm, implementing a signal processing task which consists of several stages on special purpose hardware gives prominence to the interesting issues of partitioning, and composite tasking, which are examined in this research. We believe our results have significantly improved the understanding of the performance of subspace methods, and have lead to interesting insights into the implementation issues.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 31, 1993
- Accession Number
- ADA263984
Entities
People
- Bhaskar D. Rao
Organizations
- University of California, San Diego