Gross spatial structure of land clutter

Abstract

The performance of a ground-based surveillance radar system against a particular target depends critically on whether land clutter is important within the region covered, and on the path of the aircraft relative to that clutter. Often the land clutter is very strong over parts of the region surveyed, but essentially nonexistent elsewhere. In such cases, the detailed clutter characteristics can have a much smaller effect on detectability than the gross spatial distribution of the clutter. For any given site, it is possible to produce a binary clutter map and examine several different aircraft paths to find the system performance for that site. However, often the performance for a given class of sites rather than one particular site is required. Repeating the sitespecific calculation for a reasonable sample of different sites is very time-consuming. A more powerful approach is to use a statistical characterisation of the gross structure of clutter visibility (e.g. as a function of terrain type). The paper describes the development of such a statistical representation for use in radar performance modelling, using the theory of binary Markov random fields.