The universal implementation of Space-Time Adaptive Processing

Space-Time Adaptive Processing (STAP) is the key technology of new generation airborne digital signal processing system. The primary differences between STAP methods are the order of the operations performed and the choice of transformation matrix. The operations performed in different methods are similar to each other, such as FFT and QR decomposition. This similarity makes the idea of universal platform be very attractive. Based on a parallel computational model the model, a universal real-time platform for STAP is developed. Scalable cluster-organized multi-processors structure is adopted as hardware architecture. Common operations are built into software libraries. And a heuristic methodology is used for task mapping. Two STAP methods are applied. Results show that the universal platform can meet the real-time requirement, and its scalability is good.