| ISLAM SYFUL | M, 2回目発表 | ディペンダブルシステム学 | 井上 美智子, 中島 康彦, 大下 福仁, 新谷 道広 |
|
Title:Random Variation Aware Hardware Trojan Detection Through Power based Side-Channel Analysis*
Abstract: Hardware Security has become a growing concern in the design and test of chips since its design and fabrication processes are becoming increasingly vulnerable to malicious activities and alterations. These malicious addition and modification of circuits done by intruders are commonly referred to as Hardware Trojans (HTs). Power based side channel analysis is one of the most promising techniques to detect HTs. But due to elevated process variations (Specially random process variation) as process technology nodes are gradually scaled down; obtaining high detection sensitivity using power based side channel analysis is becoming a really challenging task. Therefore, the purpose of our research is to improve detection sensitivity of HTs in presence of random process variation using power based side channel analysis.Here, we propose a random variation aware hardware Trojan detection method named as APN (Arbitrary Pair Neighboring) which shows higher detection sensitivity than existing EPN (Equal Power Neighboring) method. language of the presentation: *** English *** | |||
| DAWADI RESEARCH | D, 中間発表 | ユビキタスコンピューティングシステム | 安本 慶一, 松本 健一, 荒川 豊, 水本 旭洋 |
|
title: MutualMonitor: Elderly people monitoring each other anonymously using a smartphone application
abstract: There has been a subsequent increase in the number of elderly people living alone, contributed by the advancement in medicine and technology. However, hospitals and nursing homes are crowded, expensive and uncomfortable, while personal caretakers are expensive and few in number. Home monitoring and activity recognition technologies provide a solution for elderly monitoring but they often have issues related to privacy and security. Hence, we propose an anonymous monitoring system through which it is possible to check every day basic activity of the elderly such as taking medicine, shower, and food intake, and determine if the elderly is in normal state or requires some assistance. We do not disclose any personal information about the elderly getting monitored, to protect their privacy. We also aim to make the monitoring mutual i.e. use an elderly to monitor another elderly, through the use of a smartphone application. The application sends notifications periodically, to encourage frequent use of the application. Moreover, if the system detects some abnormality in the activity pattern of the elderly, an emergency notification is sent to the monitors, who can then analyse the situation and take necessary actions. We have conducted preliminary experiments to assess the perception of participants towards recurring notifications, and also to determine if it is possible to detect abnormal situations in the activity of the elderly through the use of our application. language of the presentation: English 発表題目: *** この部分を発表題目に *** 発表概要: *** この部分を発表概要に *** | |||
| 芹野 武 | D, 中間発表 | 計算システムズ生物学 | 金谷 重彦, 安本 慶一, MD. ALTAF-UL-AMIN, 小野 直亮, 黄 銘 |
|
title: Research on the prediction of recovery rate of pesticide in vegetables by GC/MS based on the chemical structure
abstract: GC/MS is widely used for redicual pesticide analysis in vegetables, but the recovery rate of pesticide is affected by the Matrix Effect, the interaction between pesticide and vegetables' residual matrix, due to the variety of chemical property of pesticides and redicual matrix of vegetables. The skilled food scientist can predict the rough recovery rate by the chemical structure and experiences in the past, but that kind of human-dependent skill is not always availabe in the food testing laboratory. The regression model to predict the pesticide recovery will help the food loboratory to predict the new pesticides in advance and optimize the analytical condition easily. In the present study, the regression models using the machine learning and molecular descriptor for prediction of pesticide recovery are built. These machine learning models are evaluated the performance for the data set of 7 different crops, 248 pesticides. language of the presentation: Japanese 発表題目: 化学構造に基づくGC/MSの農薬添加回収率の予測性の開発研究 発表概要: GC/MSは食品中の残留農薬分析で広く用いられているが、農薬の化学特性や、前処理後でも残る食品試料（野菜や果物、肉類など。以下試料と表記）由来の「マトリクス」などによる化学的な相互作用（マトリクス効果）などで、回収率が変化することが報告されている。 農薬の化学構造の多様性により、作物ごとにより変動する農薬の回収率の予測は容易でなく、経験を積んだ分析者の「勘」で判断されることが多い。 このような、経験者の勘に頼っている回収率の回帰モデルを、機械学習を活用して開発することで、新規農薬について作物ごとの回収率を推定することが可能となり、実験戦略を立てることが可能となる。 7種類の248農薬の回収率データに対して、分子記述子(Molecular Descriptor)による機械学習の回帰モデルを構築し、結果を考察した。 | |||