OPTIMUM PROCESSING OF UNEQUALLY SPACED RADAR PULSE TRAINS FOR CLUTTER REJECTION,
Abstract
Coherent variably spaced pulse trains can be processed using a time domain representation, i.e., describing the signal processor by a set of complex weights applied to the successive pulses rather than by a filter transfer function. The variations in clutter return during the pulse train are expressed in terms of a covariance matrix rather than a frequency spectrum. Using this approach, methods of selecting the optimum set of complex weights for a pulse train are derived. A test based on the likelihood ratio is discussed and an optimum set of complex weights is obtained for detecting targets of known doppler frequency. A more tractable criterion, based on the maximum signal-to-clutter ratio, is developed for selection of the pulse weights. The analogous problem for equally spaced pulse trains has been discussed by Rummler. With equally spaced trains, the design problem is one of selecting an optimum set of complex weights for the pulses and of specifying the interpulse spacing. With variable spacings there is the additional problem of selecting a good interpulse spacing code. No direct method for optimizing the spacing code was found. It is shown that for a particular form of the clutter autocorrelation function, the covariance matrix can be inverted and an analytic expression obtained for system performance as a function of target doppler frequency. This should be useful in testing different spacing codes so that one can be selected for a particular system application.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1967
- Accession Number
- AD0657755
Entities
People
- I. S. Reed
- L. E. Brennan
Organizations
- RAND Corporation