Zhanpeng Wang, Ziyuan Zhu, Limin Wang, Zhongkai Tong, Chao Yan
{"title":"SCPORAM: a hardware support scheme for protecting memory access patterns","authors":"Zhanpeng Wang, Ziyuan Zhu, Limin Wang, Zhongkai Tong, Chao Yan","doi":"10.1109/TrustCom50675.2020.00131","DOIUrl":null,"url":null,"abstract":"Oblivious RAM is a cryptographic protocol for hiding memory access patterns in the un-trusted external storage environment. Recently, Path ORAM has received wide attentions due to its simplicity and efficiency in secure processor design. Unfortunately, the memory bandwidth demand for Path ORAM is still too high, which restricts Path ORAM from further development. But overlapping paths can be cached to reduce memory bandwidth without the loss of security. In this paper, we present SCPORAM (Self Clustering Path ORAM), a clustering method for path merging, which transform memory requests into different categories. For each set, the intersection can be cached for every member sharing. Therefore, SCPORAM has the ability to effectively reduce memory access bandwidth. Based on this algorithm, a new ORAM controller is proposed. Compared to baseline Path ORAM, SCPORAM can reduce system overhead by 53%.","PeriodicalId":173549,"journal":{"name":"International Conference on Trust, Security and Privacy in Computing and Communications","volume":"24 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Conference on Trust, Security and Privacy in Computing and Communications","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/TrustCom50675.2020.00131","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Oblivious RAM is a cryptographic protocol for hiding memory access patterns in the un-trusted external storage environment. Recently, Path ORAM has received wide attentions due to its simplicity and efficiency in secure processor design. Unfortunately, the memory bandwidth demand for Path ORAM is still too high, which restricts Path ORAM from further development. But overlapping paths can be cached to reduce memory bandwidth without the loss of security. In this paper, we present SCPORAM (Self Clustering Path ORAM), a clustering method for path merging, which transform memory requests into different categories. For each set, the intersection can be cached for every member sharing. Therefore, SCPORAM has the ability to effectively reduce memory access bandwidth. Based on this algorithm, a new ORAM controller is proposed. Compared to baseline Path ORAM, SCPORAM can reduce system overhead by 53%.