Carl L. Colena, Michael J. Russell, Stephen A. Braun
{"title":"扫雷:一种新的快速有序统计CFAR算法","authors":"Carl L. Colena, Michael J. Russell, Stephen A. Braun","doi":"10.1109/HPEC43674.2020.9286140","DOIUrl":null,"url":null,"abstract":"A novel algorithm named ‘Minesweeper’ was developed for computing the Ordered Statistic Constant False Alarm Rate (CFAR) in a computationally efficient and novel way. OS-CFAR processing chains are used in radar applications for noise-floor estimation and target discrimination. Unlike other approaches, this algorithm aims to minimize data reuse by using training cell geometry and an accumulation matrix to compute the noise estimate. Computing the OS-CFAR in this manner affords some unique efficiencies that are novel for this application. This includes runtime invariance by bit-depth of the input data and by training geometry. Three implementations of Minesweeper were developed and benchmarked. The Optimized GPU Implementation (GPU-OPT) performed the best in both throughput and latency for large inputs. This algorithm has potential for use in real-time GPU-accelerated SDR applications.","PeriodicalId":168544,"journal":{"name":"2020 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Minesweeper: A Novel and Fast Ordered-Statistic CFAR Algorithm\",\"authors\":\"Carl L. Colena, Michael J. Russell, Stephen A. Braun\",\"doi\":\"10.1109/HPEC43674.2020.9286140\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"A novel algorithm named ‘Minesweeper’ was developed for computing the Ordered Statistic Constant False Alarm Rate (CFAR) in a computationally efficient and novel way. OS-CFAR processing chains are used in radar applications for noise-floor estimation and target discrimination. Unlike other approaches, this algorithm aims to minimize data reuse by using training cell geometry and an accumulation matrix to compute the noise estimate. Computing the OS-CFAR in this manner affords some unique efficiencies that are novel for this application. This includes runtime invariance by bit-depth of the input data and by training geometry. Three implementations of Minesweeper were developed and benchmarked. The Optimized GPU Implementation (GPU-OPT) performed the best in both throughput and latency for large inputs. This algorithm has potential for use in real-time GPU-accelerated SDR applications.\",\"PeriodicalId\":168544,\"journal\":{\"name\":\"2020 IEEE High Performance Extreme Computing Conference (HPEC)\",\"volume\":\"13 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-09-22\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2020 IEEE High Performance Extreme Computing Conference (HPEC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/HPEC43674.2020.9286140\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 IEEE High Performance Extreme Computing Conference (HPEC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/HPEC43674.2020.9286140","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Minesweeper: A Novel and Fast Ordered-Statistic CFAR Algorithm
A novel algorithm named ‘Minesweeper’ was developed for computing the Ordered Statistic Constant False Alarm Rate (CFAR) in a computationally efficient and novel way. OS-CFAR processing chains are used in radar applications for noise-floor estimation and target discrimination. Unlike other approaches, this algorithm aims to minimize data reuse by using training cell geometry and an accumulation matrix to compute the noise estimate. Computing the OS-CFAR in this manner affords some unique efficiencies that are novel for this application. This includes runtime invariance by bit-depth of the input data and by training geometry. Three implementations of Minesweeper were developed and benchmarked. The Optimized GPU Implementation (GPU-OPT) performed the best in both throughput and latency for large inputs. This algorithm has potential for use in real-time GPU-accelerated SDR applications.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
