A novel approach for radar network’s detection power analysis based on GPU

摘要

The detection power analysis of radar network is an essential but time-consuming problem in the domain of situation awareness. In this paper, we propose a novel GPU based method more efficient than existing solutions. The proposed approach is based on the field model, and the calculation of detection power is regarded as the construction and information handling process of the field model. Furthermore, it's a computation and visualization combined method. It directly produces the geotagged image, which is needed in the final display terminal. Finally, it fully utilizes the GPU hardware through various techniques, such as compressing solution space, optimizing memory usage, and instruction usage optimization. The result of the comparative experiment has shown that this novel GPU based algorithm is scalable to various conditions such as different cell resolution, detection possibility, and height. The optimized GPU version can reap the processing efficiency to about 93 times than the serial version.