Junge Xu, Bohan Xuan, An-Ting Liu, Mo Sun, Fan Zhang, Zeke Wang, Kui Ren
{"title":"天网上的终结者:用于无人机目标检测的DNN硬件IP的实际DVFS攻击","authors":"Junge Xu, Bohan Xuan, An-Ting Liu, Mo Sun, Fan Zhang, Zeke Wang, Kui Ren","doi":"10.1145/3489517.3530516","DOIUrl":null,"url":null,"abstract":"With increasing computation of various applications, dynamic voltage and frequency scaling (DVFS) is gradually deployed on FPGAs. However, its reliability and security haven't been sufficiently evaluated. In this paper, we present a practical DVFS fault attack targeting at the SkyNet accelerator IP and successfully destroy the detection accuracy. With no knowledge about the internal accelerator structure, our attack can achieve more than 98% detection accuracy loss under ten vulnerable operating point pairs (OPPs). Meanwhile, we explore the local injection with 1 ms duration and next double the intensity which can achieve more than 50% and 74% average accuracy loss respectively.","PeriodicalId":373005,"journal":{"name":"Proceedings of the 59th ACM/IEEE Design Automation Conference","volume":"70 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Terminator on SkyNet: a practical DVFS attack on DNN hardware IP for UAV object detection\",\"authors\":\"Junge Xu, Bohan Xuan, An-Ting Liu, Mo Sun, Fan Zhang, Zeke Wang, Kui Ren\",\"doi\":\"10.1145/3489517.3530516\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"With increasing computation of various applications, dynamic voltage and frequency scaling (DVFS) is gradually deployed on FPGAs. However, its reliability and security haven't been sufficiently evaluated. In this paper, we present a practical DVFS fault attack targeting at the SkyNet accelerator IP and successfully destroy the detection accuracy. With no knowledge about the internal accelerator structure, our attack can achieve more than 98% detection accuracy loss under ten vulnerable operating point pairs (OPPs). Meanwhile, we explore the local injection with 1 ms duration and next double the intensity which can achieve more than 50% and 74% average accuracy loss respectively.\",\"PeriodicalId\":373005,\"journal\":{\"name\":\"Proceedings of the 59th ACM/IEEE Design Automation Conference\",\"volume\":\"70 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-07-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 59th ACM/IEEE Design Automation Conference\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3489517.3530516\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 59th ACM/IEEE Design Automation Conference","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3489517.3530516","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Terminator on SkyNet: a practical DVFS attack on DNN hardware IP for UAV object detection
With increasing computation of various applications, dynamic voltage and frequency scaling (DVFS) is gradually deployed on FPGAs. However, its reliability and security haven't been sufficiently evaluated. In this paper, we present a practical DVFS fault attack targeting at the SkyNet accelerator IP and successfully destroy the detection accuracy. With no knowledge about the internal accelerator structure, our attack can achieve more than 98% detection accuracy loss under ten vulnerable operating point pairs (OPPs). Meanwhile, we explore the local injection with 1 ms duration and next double the intensity which can achieve more than 50% and 74% average accuracy loss respectively.