Parallel Implementation of Space-Time Adaptive Processing Algorithms

Space-time adaptive processing(STAP) technique that is well-known in the area of airborne radars requires very high computing power.In order to meet the real-time requirements in STAP applications,Multi-DSP system is used.In this paper,a method is presented to design parallel algorithms of partially adaptive STAP approaches.Data remapping is used between successive computation steps of STAP.Task level parallelism is used for these steps.As an example,the method is applied to Element space post-Doppler STAP approaches.We also implemented the parallel STAP algorithm on a Multi-DSP system.The effectiveness of our implementation is demonstrated with experimental results.The STAP algorithm is highly signal processing oriented and hence executes efficiently on DSP-based systems.