Threat Estimation by Electronic Surveillance of Multifunction Radars: A Stochastic Context Free Grammar Approach

Multi-function radars (MFRs) are sophisticated sensors that are widely used in military systems. It is shown that the stochastic context free grammar (SCFG) efficiently captures the essential features of the MFR dynamics compared to more traditional finite Markov models (regular grammars). The dynamics of the MFR are formulated as a mixture of two SCFGs - the mixture parameter determining the threat level. We then present a maximum likelihood threat estimation algorithm by capturing the noisy radar signals represented as strings from the MFR language. The relative simplicity of the SCFG model facilitates development of a systematic design procedure for electronic warfare (EW) surveillance algorithms