Design & implementation of space-time adaptive processing for target detection on FPGAs

Space Time Adaptive Processing (STAP) has an enormous computational complexity which confined its practical applications. STAP has incremented the processing requirements to a point where implementation on CPUs is not a suitable option. Meantime, the highly iterative algorithm poses precision requirements that are arduous and incompetent to address by conventional fixed-point implementations on FPGAs. This paper focuses the design and implementation of STAP Algorithm for target detection in airborne radar systems that include the QR decomposition, backward and forward substitutions and steering vector generation blocks. For QR decomposition, a variation of the modified gram-schmidt algorithm is developed that enhancement precision when implemented on FPGAs, also it is implemented as a parametric structure, which can be configured for each the matrix size easily. Thus STAP algorithm has been implemented efficiently in floating point FPGAs. Resource utilization, performance (TFLOPs), operating frequency and power consumption are measured and reported. Results show that the proposed structure act in target detection accuracy.

Keywords:

Arria FPGAs , Gram-Schmidt algorithm , QR decomposition , STAP , ULA