Pattern representations and syntactic classification of radar measurements of commercial aircraft

A syntactic pattern recognition system is evaluated for applications to radar signal identification. Three different level-crossing-based pattern representation algorithms are considered. The utility of the resulting symbolic pattern representations is assessed by evaluating the performance of a maximum-likelihood classifier when the observed symbol strings are used as inputs to the decision algorithm. A syntax analysis algorithm is derived from the likelihood function classifier. Performance results of simulated classification experiments for both maximum-likelihood and language-theoretic classifiers are presented.< >