Year

2018

Degree Name

Doctor of Philosophy

Department

School of Computing and Information Technology

Abstract

Nowadays the communication networks have acted as nearly the most important fundamental infrastructure in our human society. The basic service provided by the communication networks are like that provided by the ubiquitous public utilities. For example, the cable television network provides the distribution of information to its subscribers, which is much like the water or gas supply systems which distribute the commodities to citizens. The communication network also facilitates the development of many network-based applications such as industrial pipeline controlling in the industrial network, voice over long-term evolution (VoLTE) in the mobile network and mixture reality (MR) in the computer network, etc. Since the communication network plays such a vital role in almost every aspect of our life, undoubtedly, the information transmitted over it should be guarded properly. Roughly, such information can be categorized into either the communicated message or the sensitive information related to the users. Since we already got cryptographical tools, such as encryption schemes, to ensure the confidentiality of communicated messages, it is the sensitive personal information which should be paid special attentions to. Moreover, for the benefit of reducing the network burden in some instances, it may require that only communication information among legitimated users, such as streaming media service subscribers, can be stored and then relayed in the network. In this case, the network should be empowered with the capability to verify whether the transmitted message is exchanged between legitimated users without leaking the privacy of those users. Meanwhile, the intended receiver of a transmitted message should be able to identify the exact message sender for future communication. In order to cater to those requirements, we re-define a notion named conditional user privacy preservation.

In this thesis, we investigate the problem how to preserve user conditional privacy in pubic key encryption schemes, which are used to secure the transmitted information in the communication networks. In fact, even the term conditional privacy preservation has appeared in existing works before, there still have great differences between our conditional privacy preservation definition and the one proposed before. For example, in our definition, we do not need a trusted third party (TTP) to help tracing the sender of a message. Besides, the verification of a given encrypted message can be done without any secret.

In this thesis, we also introduce more desirable features to our redefined notion user conditional privacy preservation. In our second work, we consider not only the conditional privacy of the message sender but also that of the intended message receiver. This work presents a new encryption scheme which can be implemented in communication networks where there exists a blacklist containing a list of blocked communication channels, and each of them is established by a pair of sender and receiver. With this encryption scheme, a verifier can confirm whether one ciphertext is belonging to a legitimated communication channel without knowing the exact sender and receiver of that ciphertext. With our two previous works, for a given ciphertext, we ensure that no one except its intended receiver can identify the sender. However, the receiver of one message may behave dishonest when it tries to retrieve the real message sender, which incurs the problem that the receiver of a message might manipulate the origin of the message successfully for its own benefit. To tackle this problem, we present a novel encryption scheme in our third work. Apart from preserving user conditional privacy, this work also enforces the receiver to give a publicly verifiable proof so as to convince others that it is honest during the process of identifying the actual message sender. In our forth work, we show our special interest in the access control encryption, or ACE for short, and find this primitive can inherently achieve user conditional privacy preservation to some extent. we present a newly constructed ACE scheme in this work, and our scheme has advantages over existing ACE schemes in two aspects. Firstly, our ACE scheme is more reliable than existing ones since we utilize a distributed sanitizing algorithm and thus avoid the so called single point failure happened in ACE systems with only one sanitizer. Then, since the ciphertext and key size of our scheme is more compact than that of the existing ACE schemes, our scheme enjoys better scalability.

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.