Simulation-Based Performance Analysis of Several Types of Normalizers Under Correlated Stationary Noise Conditions.
Abstract
The results of a simulation-based performance analysis of several types of normalization algorithms commonly considered for use in active sonar applications are presented. The techniques investigated are the mean value (MV), two pass peak shear (2PP), greater-of (GO) mean values and the ordered statistic (OS). Each algorithm incorporates a sliding, split window which encompasses the data to be used to compute the normalizing parameter and is centered on the data point to be normalized. A gap, immediately adjacent to and centered on the data point to be normalized, is used to avoid using any data that may be part of the signal in the estimate. The simulations consist of the generation of numerous independent sequences of correlated, Rayleight distributed data samples, representing stationary, zero-mean Gaussian noise passed through a matched filter then envelope detected, which are then normalized and the resulting normalized data analyzed. Normalizer, Mean Value(MV), Two Pass Peak Shear(2PP), Greater-Of(GO), Order Statistic(OS), Rayleigh, Correlation, Constant False Alarm Rate(CFAR).
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 21, 1993
- Accession Number
- ADA271488
Entities
People
- Fyzodeen Khan
Organizations
- Naval Undersea Warfare Center