摘要
访问控制在安全领域起着越来越重要的作用,个人的一些生物特征具有唯一性而且难以改变,利用这些特征识别用户身份,从而进行系统访问控制具有很强的优势。论文首先通过对人体生物击键特性的分析,阐述了基于生物击键特性的访问控制过程;然后对已有的击键序列识别算法进行分析,提出了一种新的自适应击键特性识别算法。该算法能够适应用户敲击键盘熟练程度的变化,并可防止输入过程中的个别“奇点”影响整体的识别效果。算法改进的讨论中,通过改变特征数据的存储结构使得用户修改密码后,算法仍然能够从历史数据中挖掘击键特性,提高学习效率。最后的测试结果和理论分析都表明该方法对于因密钥丢失导致的系统失控具有很好的防护作用。
Access control is playing a more and more important role in the secure systems.The biologic information of human is unique and unchangeable.So it has many advantages to indicate users in system.This paper firstly presents some details of human keystroke characteristics,then gives a basic framework of access control based on biologic keystroke characteristics.Following that,it analyzes algorithms on keystroke sequence,suggests a new self-adaptive algorithm,by which system can adjust parameters automatically according to the variety of the user＇s skill and get rid of the odd point in input.To improve the algorithm,a new structure is applied to record any useful data so that system can easily find the user＇s characteristics when he modifies his password.In the end,theory analysis and test results are both given to explain the validity of this method ：even if password is stolen ,the attacker can hardly find how the owner types because of user＇s biometrics for keystroke.Hence this way can overcome the weak point of password checking schemes.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第31期1-4,85,共5页
Computer Engineering and Applications
基金
国家自然科学基金(编号:90104025)
关键词
生物击键特性
访问控制
自适应
身份识别
信息安全
biologic keystroke characteristics, access control, adapt, personal identification, information security
