Toward of a GPU Acclerated Software Navagation Radar

Qian Zhang,Yangdong Deng
DOI: https://doi.org/10.1109/icsess.2013.6615277
2013-01-01
Abstract:Software radars are attracting an increasing interest in recent years due to their advantages in cost efficiency, flexibility, and ease of updating. However, the processing and rendering of radar signal have to be done in a real-time fashion. Such a requirement pose challenge to software radar design because radar signal processing can be extremely time-consuming on large computation load such as the persistence effect and targets detection. This paper proposes a software solution by using graphics processing units (GPUs) to accelerate the radar signal processing and rendering process. A key feature of our solution is to take advantage of both the graphics and the general purpose computing capabilities of GPUs. When implemented on the NVIDIA GTS 250 GPU, our software radar outperforms its CPU equivalent by a factor of over 100. Our work demonstrates that GPU based software radars have strong potential in future applications.
What problem does this paper attempt to address?