Application of OpenMP to compression/decompression and decoding for weather radar base data
Article
Figures
Metrics
Preview PDF
Reference
Related
Cited by
Materials
Abstract:
In this paper, a parallel chunking processing scheme is designed for the compression/decompression process by adopting OpenMP technology based on a conventional PC workstation as a benchmark, and the compression and decompression time of the data can be reduced to about 1/5 and 1/8 of the single-threaded one when the number of chunks reaches 16 under a 12-core/24-threaded CPU. For the decoding process, the results of parallel processing by PPI, radial and distance banks are compared, and it is found that the parallel processing scheme on radial is optimal, which can reduce the decoding time to about 1/8 of the single-threaded one. Through the application of the above two parallel optimization techniques, the pressure of radar base data in data transmission and preprocessing can be significantly reduced, and the data loading performance of the radar software can also be enhanced to improve the interactive experience of the radar analysis software based on base data.