BROWSE
Detail
Conference Paper
GPU based Signal Processing for Sea Surveillance Radar
Cited - time in 
Share 
Share
- Authors
- Byung-Doo Kim, Do-Hyeung Kim, Dae-Hee Seo, Byung-Gil Lee
- Issue Date
- 2013-11
- Citation
- International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) 2013, pp.632-635
- Language
- English
- Type
- Conference Paper
- Abstract
- In this paper, the signal processing system of sea surveillance radar is developed for vessel traffic system. The proposed signal processing system uses CPU and GPU together to reduce noise, clutters and detect targets. The proposed scheme is implemented on 3.2GHz Intel Xeon W3670 Processor and GTX 560 Ti GPU with 384 CUDA cores. The developed signal processing system is examined via connection with real sea surveillance radar transceiver. In these cases that CPU is used standalone and CPU and GPU are used together, the computation time is compared, respectively, when the size of raw radar image is 12006×4096. From the experimental results, the usage of CPU and GPU together provides reduced computation time about 33% in whole signal processing time.
- KSP Keywords
- Intel XEON, Radar image, Radar transceiver, Surveillance radar, Traffic system, Whole signal, computation time, processing time, sea surveillance, signal processing system