Applications of space-time decision and estimation theory to antenna processing system design

A mathematical solution to the problem of optimum radar target detection and parameter estimation in receiver noise and heavy clutter has been achieved by means of space-time decision theory. The theory leads to a conceptual design for an antenna processing system that is optimum in the sense that it makes least risk parameter estimates and least probability of error decisions. The system can be instrumented by separately demodulating the individual radiator outputs and feeding them simultaneously and in parallel to a digital computer, Hypothesis testing problems such as target detection are formulated by means of a generalized likelihood ratio test. Optimum mean-square estimation is carried out by instrumenting the mean of the parameter in question conditioned on the observed signal. With the aid of a priori statistics available or assumed, the a posteriori likelihood function is derived. From this function, the required generalized likelihood ratio hypothesis tests and parameter estimators are synthesized. Specific illustrations include systems for detection and angular location estimation for one or two targets in a clutter environment. Optimum tests as well as optimum and suboptimum estimators are realized as flow diagrams for computation by the special-purpose digital processor.