RFID认证系统的匿名隐私模型和技术研究

发布时间：2018-01-14 01:22

本文关键词：RFID认证系统的匿名隐私模型和技术研究　出处：《广西师范大学》2015年硕士论文　论文类型：学位论文

【摘要】：RFID技术是新型无线感知技术,其利用无线电波对目标进行自动识别,具有识别速度快、环境适应性好、可复用等优点,主要应用在识别和认证领域,如智慧交通、身份识别、物流管理等。由于阅读器与标签之间无线通信信道的暴露性、交互协议的自身缺陷以及各组件内部的安全风险等因素,导致攻击者可以对阅读器和标签之间的通信进行窃听、重放,也可以对标签进行追踪。因此在RFID应用中,存在安全和隐私问题。随着系统规模的扩大,标签的数量越来越多,由此引发了以下问题。一方面,标签的庞大数量迫使制造商不断压低标签的生产成本,导致标签的计算能力和存储资源不足,因此标签只能执行一些简单的操作,如异或、生成伪随机数、Hash运算等,而不能支持复杂的加密技术；另一方面,应用系统要求阅读器能快速的完成对标签的识别和认证,减少时延,保证较高的认证效率。显然在资源受限的条件下,如何实现隐私保护强度和认证效率之间的平衡,成为关键问题。为了解决上述问题,已经提出了很多隐私模型和轻量级认证协议。如Sarma等人提出的Hash-Lock协议,Weis提出的树形结构协议,Li Lu等人提出的刷新隐私模型等。但是,这些协议都在安全性或者认证效率方面存在一些问题。如Hash-Lock协议存在信息泄露的风险,并且其认证效率较低；树形结构协议对存储资源需求高,且容易遭受关联性攻击；刷新隐私模型容易遭受追踪攻击,导致标签的位置隐私泄露。本文主要研究如何在低成本的前提下解决RFID系统中的安全和隐私问题。通过对系统各组件的深入分析,了解其特点和面临的威胁；研究和分析当前的隐私模型和交互协议,针对其中的缺陷,给出优化和解决方案；面对新型攻击方式,提出新的隐私保护模型,同时给出可用的认证协议,更加注重隐私保护强度和认证效率之间的平衡。本文的研究成果包括以下方面：(1)结合RFID认证系统隐私保护的需求,对RFID认证系统中的隐私威胁进行分类：分析RFID认证系统的隐私模型中存在的问题,并给出解决方案。其中强隐私模型要求标签的输出信息具有严格的随机性和不可预测性,导致系统的认证效率大大降低。而刷新隐私模型是一种弱隐私模型,其主要基于刷新机制。之所以称之为弱模型,是因为其对标签输出信息的随机性与不可预测性没有严格的要求。刷新隐私模型允许标签输出信息中包含一个不变的临时信息,并且利用此信息提高认证效率。但是攻击者通过窃听截获此信息,并且将此信息与标签进行绑定,可以对标签进行追踪,即标签携带者的位置信息将被泄露。虽然,攻击者并不知道标签的密钥信息,但是这依然严重威胁到用户的个人隐私安全。(2)提出了基于k-匿名的匿名隐私保护模型。k-匿名是数据发布时用于隐私保护的经典方法,其通过在数据中构造k条相似的记录,使得攻击者无法区分这k条记录。具体来说,k-匿名的方法是通过概念层次泛化,对数据的某一属性值进行泛化处理,保证每个属性值都至少对应k条记录。即使攻击者通过其他途径获得某一个属性值,也无法区分这k条记录,更无法定位到具体的某一条记录,因此用户的隐私得以保护。在新的隐私模型中,利用k-匿名的思想对标签进行处理。将系统中所有的标签随机的划分为若干组,每个组包含至少k个标签,这k个标签使用相同的准标识符,组与组之间的准标识符不相同并且相互独立。准标识符以明文的形式直接包含在标签的响应信息中,但不存在刷新隐私模型中的安全问题。因为,即使攻击者获得准标识符,也无法分辨同一组的k个标签,无法对标签进行追踪,所以,新模型可以保护标签的位置隐私不被泄露。同时,阅读器在认证标签时,可以利用明文的准标识符加快认证效率。(3)提出一个基于k-匿名隐私模型的新的轻量级隐私保护协议k-PPA,其可以保护隐私安全,并且具有较高的认证效率。k-PPA协议可以抵抗窃听攻击、重传攻击、假冒攻击和追踪攻击等。与同类协议相比,该协议在隐私保护强度和认证效率之间取得了较好的平衡。
[Abstract]:RFID technology is a new wireless sensing technology, the use of radio waves to identify the target, high recognition speed, good environmental adaptability, reusable advantages, mainly used in the field of identification and authentication, such as identity recognition, intelligent transportation, logistics management and so on. Because of the exposure of the wireless communication channel between the reader and tag the defects, interaction protocols and components of the internal security risks and other factors, an attacker can eavesdrop on the communication between the reader and the tag, the replay can also tag tracking. Therefore in the RFID application, there are security and privacy problems. With the expansion of the scale of the system, the increasing number of tags, which triggered the following problems. On the one hand, a large number of label forced manufacturers continue to depress the production cost of the resulting label label, computing and storage resources, so the label Can only perform some simple operations, such as XOR, pseudo random number generation, Hash operation, and can not support complex encryption technology; on the other hand, the application requirements of the system can quickly complete the reader to tag identification and authentication, reduce the delay, ensure the high efficiency of authentication. Obviously in resource constrained conditions. How to realize the strength of privacy protection and authentication efficiency balance, become a key problem. In order to solve the above problems, a lot of privacy model and lightweight authentication protocols have been proposed as proposed by Sarma et al. Hash-Lock protocol, protocol tree structure proposed by Weis, Li proposed by Lu et al. Refresh the privacy model but these agreements exist some problems in the security or authentication efficiency. Such as risk Hash-Lock protocol for information disclosure, and the authentication efficiency is low; the tree structure of storage resource demand high agreement, And the relevance of vulnerable to attack; refresh the privacy model vulnerable to tracking attacks, resulting in tag location privacy. This paper mainly studies how to solve the security and privacy issues in RFID system on the premise of low cost. Based on the in-depth analysis of the various components of the system, understand its characteristics and threats; research and analysis of privacy model and the current interaction protocol, aiming at the defects, and proposed the optimal solution; in the face of new attack methods, put forward a new privacy protection model, and gives the available authentication protocol, pay more attention to the strength of privacy protection and authentication between the balance of efficiency. The research results of this paper include the following aspects: (1) combined with privacy protection RFID certification the demand of the system, the classification of RFID authentication system in privacy threats: analysis of the existing privacy model RFID authentication system in the problem, and gives the solution The strong privacy case. The output information model label has strict randomness and unpredictability, cause the system authentication efficiency greatly reduced. And refresh the privacy model is a kind of weak privacy model, which is mainly based on the refresh mechanism. It is called a weak model, because of the randomness of tag output information and unpredictability is not a strict requirement. Refresh privacy model allows the inclusion of a temporary information the same tag output information, and use this information to improve the authentication efficiency. But the attacker by eavesdropping intercepted this information, and this information and tag binding, the tag can be tracked, the position information label carriers will be leaked. Although, the attacker does not know the key information labels, but it is still a serious threat to the user's personal privacy. (2) proposed anonymous anonymous privacy based on k- .k- model is a classic for anonymous protection method for privacy preserving data publishing, by constructing K similar records in the data, the attacker can not distinguish the K records. Specifically, k- anonymous is through the concept of a hierarchical generalization, attribute data values that each attribute value learning process. At least the corresponding K records. Even if the attacker obtained through other means a certain attribute value, also cannot separate the K records, more unable to locate a record of a specific, so the user privacy can be protected. In the privacy model, to deal with the label using k- anonymous system. All the tags randomly divided into several groups, each group contains at least k tags, K tags that use the same quasi identifier, quasi identifiers between the groups were not identical and independent of each other. The quasi identifier to The plaintext form directly included in the response information label, but there is no refresh security issues privacy model. Because, even if an attacker to obtain quasi identifiers, also cannot distinguish the same set of K tags, to track on the label so that the new model can location privacy protection label is not leaked. At the same time in the authentication tag, reader, can speed up the authentication efficiency using quasi identifiers. The plaintext (3) proposed a k- based model of the new lightweight anonymous privacy privacy protection protocol k-PPA, which can protect the privacy and security authentication,.K-PPA protocol has high efficiency and can resist the attack, replay attack, impersonation attack and tracking attack. Compared with similar protocols, the proposed protocol achieves good balance between privacy protection strength and authentication efficiency.

【学位授予单位】：广西师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP309;TP391.44

【相似文献】