Detection and angle estimation of coherent plane waves
Abstract
The problem of detection and angleofarrival estimation of multiple coherent plane waves is addressed. A unified approach based on the singular value decomposition (SVD) is proposed to solve both the detection and angle estimation problems. A set of detection criteria for determining the number of coherent plane waves present in the field is developed and their performance of the case of two closely spaced coherent sources is analyzed. The probability of error for each criterion is derived as a function of the signaltonoise ratio, the angle separation, the intersignal correlation, and the number of snapshots. Using plots of these probability of error functions, thresholds of the above parameters for the correct detection are identified. For the problem of angleofarrival estimation, the SVDbased modified ForwardBackward Linear Prediction (FBLP) method of spectral estimation is applied and its performance under a multipath environment is analyzed. It is shown that, this method is capable of yielding high resolution estimates at low signaltonoise ratios and with few number of snapshots. The advantage of this approach, besides its ability to solve the detection and angle estimation problem for the coherent signals, is its high computational efficiency and stability which is inherent to the SVD technique.
 Publication:

Ph.D. Thesis
 Pub Date:
 1987
 Bibcode:
 1987PhDT........62S
 Keywords:

 Estimates;
 Numerical Analysis;
 Plane Waves;
 Prediction Analysis Techniques;
 Signal Detection;
 Signal To Noise Ratios;
 Algorithms;
 Errors;
 Probability Theory;
 Communications and Radar