Application of deterministic techniques to STAP

In this chapter, the authors have presented four deterministic direct data domain least-squares techniques for nulling interferers and extracting the signal of interest. This approach generates the adaptive weights on a snapshot-by-snapshot basis that eliminates the need for auxiliary training data sets. These techniques are capable of handling both coherent and non-coherent interferers in a stationary or non-stationary environment. The four processors are the eigenvalue processor, and the three least-squares methods: forward, backward and forward-backward procedures which may be implemented in real time on a signal processing chip. After describing the radar environment used to demonstrate these techniques, the authors developed the one-dimensional (space domain) implementation for each of the methods. The chapter has shown the generation of system matrices using the spatial samples, solves for the adaptive weights, and estimates the amplitude of the desired signal. The performance of the forward, backward and forward-backward methods has been evaluated using the output SINR and adaptive weight pattern for each of the methods.