快速輕量級分類器，用於智能手機上基於行為生物識別的身份驗證基於;Fast and Lightweight Classifier for Behavioral Biometric-Based Authentication on Smartphone

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/84011

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/84011

Title:	快速輕量級分類器，用於智能手機上基於行為生物識別的身份驗證基於;Fast and Lightweight Classifier for Behavioral Biometric-Based Authentication on Smartphone
Authors:	羅賓曼;Robihamanto
Contributors:	資訊工程學系
Keywords:	輕分類器;快速行為生物識別;能耗;移動設備;電池消耗;light classifier;fast behavioral biometric;energy consumption;mobile device;battery consumption
Date:	2020-07-28
Issue Date:	2020-09-02 17:54:39 (UTC+8)
Publisher:	國立中央大學
Abstract:	傳統的智能手機身份驗證機制當前使用PIN，密碼和基於生物特徵的方法。問題是，未鎖定的智能手機將保持打開狀態，直到再次將其主動鎖定。攻擊者總是有一個時間範圍可以竊取未鎖定的手機並竊取設備上的所有數據。連續認證的方法需要在智能手機上提供多種安全性。基於生物特徵的隱式身份驗證方法是更方便的行為，因為用戶不會意識到身份驗證階段。但是，在對手機上執行的過程進行身份驗證時，我們還必須考慮電池的使用，以便使用身份驗證不會消耗過多的能量。我們致力於構建快速，輕量級但仍具有競爭準確性的分類器。在進行了許多實驗並與常用的分類器（常用的分類器）進行比較之後，我們認為我們的分類器通過提供良好的準確性和低能耗而最有效。;The traditional smartphone authentication mechanism currently uses PINs, passwords and biometric-based methods. The problem is, an unlocked smartphone will remain open until it is actively locked again. There is always a time frame when an attacker can steal an unlocked cell phone and steal all data on the device. The method of continuous authentication needs to provide multiple security on smartphones. Biometric based implicit authentication methods are more convenient behavior because users will not be aware of the authentication phase. However, when authenticating the process carried out on a cell phone we must also take into account the use of batteries so that the use of authentication does not consume too much energy. We approach the building of a classifier that is fast and lightweight but still provides competitive EER. After conducting a number of experiments and comparing with the popular classifier, which is a classifier that is often used, we consider that our classifier is the most effective by providing good EER and low energy usage.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	164	View/Open

社群 sharing

Loading...