Ruizhong Wei

Efficient K-anonymization for privacy preservation

Privacy preservation during cooperation has become an interesting issue in the last few years. This problem attracted much research work. k-anonymization is an efficient approach to protect data privacy. However, k-anonymization problem was proven NP-hard though the idea of k-anonymizafion is not complex. In this paper, we propose two simple but very efficient algorithms, which work for numeric and categorical data respectively, can minimize information loss as low as possible. We show that these algorithms can produce better performance comparing to other known algorithms.

An Efficient Trust Model for Multi-Agent Systems

Agent-based e-business opens up a computational direction, in which software agents behave on behalf of their owners. However, enabling agents to make decisions and exploiting information to other unknown agents have introduced some security challenges among them is trust relationship while agents are dealing with unknown environments and their residences. Despite existing of some sophisticated proposed approaches addressing the above problem, most of them are suffering from complexity and therefore drawback from development. We propose an efficient trust model for distributed systems mainly focusing on multi-agent systems. We provide a feasible mechanism using some well-known cryptographic techniques such that not only it addresses the above issue but also guarantees the security and resistibility of the model against some attacks

A multi-queue algorithm for DDoS attacks

We proposed an algorithm for gateway and router to prevent DDoS attacks. The algorithm combines two simple congestion control methods. Simulation results show that our algorithm efficiently increases the throughput of normal flows under DDoS attacks comparing to common Drop Tail algorithm.

Indoor localization of ubiquitous heterogeneous devices

We investigated the particular case of heterogeneous devices used in location based applications. The results suggest applying relative signal strength readings to overcome variations in the wireless characteristics of different devices. An experiment was used to determine an optimum data collection interval. Finally, a full end-to-end reference implementation is constructed. The results show that heterogeneous commodity devices can achieve good localization without a calibration period or expert site surveying.

Indoor location based on WiFi

In this paper we consider indoor location based on WiFi signals under the condition that the location provider does not know the physical location of the access points. This scenario is different from most previous researches on indoor location. We proposed several algorithms for this purpose including set calibration data, calculate position and track moving rout. Results from some experiments show that our scheme is useful for many kind buildings such as Malls and office buildings.

Zero-knowledge trust negotiation

Electronic business or on-line cooperation transactions happen regularly over the internet. Such a transaction usually involves a service provider who provides a certain service (i.e., perform an on-line purchase) and a service requester who requests the service. In order to decide whether a service requester can access a service, a distributed access control system can be used. Traditional identity-based access control systems usually require pre-register, which is too rigid to adapt to the rapid developing on-line cooperation. Trust-based access control provides open authentication and access control. The flexibility that it introduces could boost the on-line cooperation significantly. However, it is vulnerable to attacks that lead to leakage of sensitive information. Furthermore, certain credentials (such as, credit card number) are too sensitive to release for some people even through proper release policies. This paper introduces the Zero-knowledge protocol for credential verification, and presents a trust-based access control framework that incorporates this protocol. This system keeps the highly sensitive credentials secret; while at the same time proceed with the trust negotiation.

Applying secret sharing schemes to service reputation

A successful result of choosing, invoking, integrating and composing services in a collaborative environment depends directly on locating reliable services with respect to their performances, histories, feedbacks from their customers, and generally their reputations. Nevertheless, lacking a secure mechanism to create countable reputation information may lead service providers to exaggerate their performances. We propose to apply threshold schemes under the context of service reputation towards providing an unconditionally secure reputation credits and systems for services and their providers. We argue that applying such a secure mechanism for reputation of services is not forgeable. As a result, any service requester was able to rely on valuable reputation information and choose the most reliable services with confidence.

Vector signature for face recognition

In this paper, we proposed a vector signature scheme for face recognition. Using the signature, both the database size and communication bandwidth can be reduced. And the privacy of the face image is also improved. Some experimental implementation shows the potential of the new proposal.

The complexity of a data privacy protection algorithm

To share data with other parties in computer collaboration is an important issue of data security. k-anonymity and ℓ-diversity are used to protect privacy and secure data sharing. It was proved that optimal k-anonymity are NP hard for k ≥ 3 and that there is a polynomial algorithm for optimal 2-anonymity. This paper proves that an optimal 2-diversity is NP hard. Since ℓ-diversity must be ℓ-anonymity, it follows that the ℓ-diversity is NP hard for ℓ ≥ 2. The result shows that finding polynomial algorithm for efficient (but not optimal) ℓ-diversity is important.

Implement privacy for an OMS

Currently, the use of transfer information is becoming critical for organizations. Collaborative activities such as coordination, decision making and negotiation depend on it. Organizational memory systems (OMS) have been proposed to accumulate, organize, preserve and share diverse information coming from various sources. However, the need to provide privacy capability to all the users of an OMS requires a good strategy for acquiring and structuring the information. This paper presents a new privacy mechanism strategy for the information stored in the OMS.