Title :
Broadband measures of performance for active sonars
Author :
Huster, Michael E. ; Fish, David
Author_Institution :
AETC Inc., San Diego, CA, USA
Abstract :
The conventional measures of detection performance make assumptions that are invalid for broadband systems (>1 octave). This paper discusses a figure of merit (FOM) derived from the maximum likelihood quadratic detection statistic that makes neither narrowband nor small input signal to noise assumptions. This FOM predicts the performance of the general energy detector with sub-optimal processing parameters near a probability of detection of 0.5. It is also extended to predict the signal excess. The FOM is used as the basis for predicting the performance in the presence of unknown parameter values
Keywords :
maximum likelihood detection; sonar detection; FOM; active sonar; broadband measures; detection performance; figure of merit; general energy detector; maximum likelihood quadratic detection statistic; performance; sub-optimal processing parameters; Background noise; Detectors; Frequency; Maximum likelihood detection; Power system modeling; Signal processing; Signal to noise ratio; Sonar measurements; Statistics; System performance;
Conference_Titel :
OCEANS 2000 MTS/IEEE Conference and Exhibition
Conference_Location :
Providence, RI
Print_ISBN :
0-7803-6551-8
DOI :
10.1109/OCEANS.2000.881785
