GPU ACCELERATED COMPUTING IN RADAR INFORMATION PROCESSING

89 views

Abstract

Position synchronization on maps is a critical problem in displaying radar information. Previously, this process was done in a sequential manner in the central processing (CPU). This often causes some bottlenecks leading to a long processing time because of the limited CPU architecture. To mitigate this problem, we proposed an algorithm of parallel processing to take advantage of the multi-core architecture of GPU to quickly reproject coordinates in radar information processing, improving the information display performance. Practical tests were done on different data sets to demonstrate the performance of this method, and the result showed that the improvement speeds are between 10 to 100 times better in each test.