Target model pitfalls (Illness, diagnosis, and prescription)

Abstract

It is a very common custom to use extremely simple target models in specification for radar, radar system design, performance analysis, comparison, and evaluation. While the use of rudimentary models simplifies the work, their use can only result in ultimate disappointment of the radar user, severe illness in the course of the radar development and testing, followed by major surgery, and sometimes a premature death of the program. Consideration of the purposes of radar indicates that its functions are to detect/identify/track real-world targets, not paper targets. Problems in relating simple target models to real-world targets are treated for both aircraft and ballistic targets. Based on proper radar representation, simulation results and data from field performance tests, the inadequacy of simple target models is shown. A number of situations in the design and comparison of radars are given to further illustrate the problem. Special treatment is accorded to "average" radar cross section (RCS). Finally, recommendations are made for target models which more nearly approach the real world and permit an improvement in the design, evaluation, and analytical comparison of radars. It is shown that these models, while giving more valid results, are not necessarily complex and are achievable.