TOWARDS INTELLIGENT RADAR SYSTEMS

Based on the author's previously published work that relates to Bayes' decision making in a fuzzy and random environment [Fuzzy Sets and Systems 35 (1990) 197-212] and the work addressing the problem of fuzzy set ordering over the real line [Fuzzy Sets and Systems 50 (1992) 237-2461, the theory of fuzzy sets is applied here to radar detection problems. This has the purpose of providing the possibility for accommodating the operation of radar systems according to the particular situation and thus, making these systems, in some sense, intelligent. First, the prior probabilities and cost assignments that are involved in the Bayes risk expression are fuzzified. Second, in addition to fuzzy prior probabilities and costs, an unknown fuzzy signal amplitude is considered. This leads to a restoration of the usefulness of the Bayes criterion in the area of radar, one that has lately been neglected to the advantage of the Neyman-Pearson criterion. Also, a parallelism between the fuzzy approach and the minimax Bayes risk approach is drawn.