Vuze is a popular file-sharing client. When looking for content, Vuze selects from its list of neighbours, a set of 20 nodes to be contacted; the selection is performed such that the neighbours closest to the content in terms of Vuze ID are contacted first. To improve efficiency of its searches, Vuze implements a network coordinate system: from the set of 20 to-be-contacted nodes, queries are sent to the closest nodes in terms of network distance, which is calculated by the difference in network coordinates. However, network coordinate systems are inherently insecure and a malicious peer can lie about its coordinate to appear closest to every peer in the network. This allows the malicious peer to bias next-hop choices for victim peers such that queries will be sent to the attacker, thus hijacking every search query. In our experiments, almost 20% of the search queries are hijacked; the cost of performing this attack is minimal – less than $112/month.

Light encryption device (LED) is a 64 bit lightweight block cipher proposed by Guo et al. at CHES 2011, and its key size is primarily defined as 64 and 128 bits. This study studies differential fault analysis (DFA) of LED using the technique of Super-Sbox analysis. Under various fault models, the fault pattern propagation rule of the Super-Sbox can be obtained, based on which the efficiency of fault attack on LED can be greatly improved. For LED-64, under the nibble-based fault model, a random nibble fault at the 30th round can reduce the size of key search space to 2⁷–2²⁰ (average 2^14.02). Even if a random nibble fault is injected into the 29th round, the size of the key search space can also be reduced to about 2^17.43–2^17.72 (average 2^17.65) using early-abort technique. Although under the byte-based fault model, a random byte fault at the 30th round can reduce the size of the key space to 2⁷–2¹⁶ (average 2^11.92). If the adversary has the capability of injecting two random nibble faults at some specified rounds, then the above fault attack on LED-64 can be similarly extended to LED-128, and the size of the exhaustive search space for the 128 bit key can be reduced to 2¹⁵–2^27.94 (average 2^21.96). These results demonstrate that Super-Sbox is a powerful technique that can be used to obtain significant improvements in the key filtration, and thus improve the efficiency of DFA on some special ciphers.

The MARS-like structure is a generalised Feistel structure. Unified impossible differential (UID) method is an effective method to discover impossible differential characteristics for block cipher structures. In this study, for a specific kind of MARS-like structure, the authors use UID to show that when n, the number of subblocks, is even, there always exist 3n − 1 rounds impossible differentials. Moreover, the authors prove that when n is odd, the MARS-like structure has impossible differentials for any number of rounds, which is a clear but interesting result.

Unattended wireless sensor networks (UWSNs) are operated in hostile environments without constant supervision by a trusted sink; so it faces the risk of compromising by adversaries (ADVs). In this study, proposed is a cooperative hybrid self-healing randomised distributed (CHSHRD) scheme, a new mechanism to enhance the confidentiality of the data collected by UWSNs. The proposed scheme employs both proactive and reactive peers to ensure both backward secrecy and data reliability. It helps the unattended sensors to self-heal and restore their backward secrecy by asking for help from the best qualified neighbours to generate a new secret key to regain their secrecy. The sick sensors also use the best qualified neighbours to distribute the data parts among them to protect the data from eavesdropping and this will enhance the data reliability. In this study, they also present a powerful, realistic and agile ADV model and show how CHSHRD scheme can result in sensor regaining secrecy and achieving high data reliability, despite the adversary efforts to the contrary. The evaluation of the proposed scheme relies on both theoretical probabilistic results and simulation results that compare the proposed scheme to other protection schemes. The results show that this hybrid scheme provides better protection than other schemes that use either proactive or reactive peers.

(t, n) threshold secret sharing is a cryptographic mechanism to divide and disseminate information among n participants in a way that at least t(t ≤ n) of them should be present for the original data to be retrieved. This has practical applications in the protection of secure information against loss, destruction and theft. In this study, the authors propose a new multi-secret sharing scheme which is based on Hermite interpolation polynomials. Using the properties of discrete logarithm over elliptic curves and bilinear maps, they have created a verifiable scheme in which there is no need for a secure channel and every participant chooses their own share. This feature does not let the dealer cheat. The proposed method is dynamic to the changes in the number and value of the secrets as well as the threshold. In addition, it has the multi-use property which reduces the cost of secret distribution in multiple rounds of operation. The public values used in the proposed scheme are less than those of schemes providing similar features and the computations are also less complex. At the end of this study, they have compared the author's scheme with the similar ones against a comprehensive set of key features used in secret sharing.

The aim of this study is to provide a generic implementation strategy for honeytokens deployed within a database management system that leverages the aspect-oriented paradigm to contain the insider threat. This approach is tested by developing a proof-of-concept prototype in an aspect-oriented language, namely AspectJ. This study also reflects on design and implementation challenges involved in the deployment of honeytokens to contain the insider threat. Consequently, aspect-orientation is proposed as a means to resolve some of these challenges.

A distributed key generation (DKG) protocol is a central component in distributed cryptosystems. It allows a group of participants to jointly generate a pair of keys (private key and public key) without assuming any trusted participant. The public key is output in clear but only authorised subgroups of participants are able to reconstruct or use the private key. Existing literature on DKG protocols assumes equal authority for participants. In this study, the authors consider the problem of DKG in groups with hierarchical structure where the authorised subsets can be defined by a hierarchical threshold access structure. They first propose a verifiable hierarchical threshold secret sharing protocol. On the basis of this construction, they then propose a new DKG protocol with hierarchical threshold access structure for discrete-logarithm-based cryptosystems. It is shown that the proposed protocols satisfy all the security requirements.