Shinyoung Lim

Modeling of multiple agent based cryptographic key recovery protocol

When a receiver of a ciphertext message can not decrypt the message due to the fact that he has lost his private-key, the private-key of the receiver and session-key of the message need to be recovered. We demonstrate how we have modeled and analyzed a new type of multiple agent based key recovery protocol. It is characterized by key encapsulation approach, protocol generalization, secret choice of key recovery agents and fork/join of session-keys by random-keys. The proposed protocol is formally modeled by a new pictorial model, an extended cryptographic timed Petri net (ECTPN). Recoverability of a session-key as well as performance of the protocol is verified by using a reachability graph of the ECTPN.

Specification and analysis of n -way key recovery system by extended cryptographic timed Petri net

Abstract When a receiver of a ciphertext message cannot decrypt the message due to the fact that he has lost his private-key, the private-key of the receiver and session-key of the message need to be recovered. In this paper, we design a new n -way Key Recovery System ( n -KRS) that is characterized by key encapsulation approach, n -way key recovery, secret designation of key recovery agents, fork and join of session-key with random keys, recovery request by a new key pair, certificate attaching scheme, component-based software engineering, and conformance to international standards. n -KRS is specified and analyzed by Extended Cryptographic Timed Petri Net (ECTPN) model that is a new formal and graphical specification model for cryptographic protocols. Recoverability, performance characteristics of n -KRS are evaluated by means of reachability graph of ECTPN. A prototype of n -KRS is developed and evaluated by means of CC 2.0 based information security system evaluation scheme.

Effective Iris Recognition System by Optimized Feature Vectors and Classifier

This paper presents an effective system for recognizing the identity of a living person on the basis of iris patterns that is one of the physiological and biological features with high reliability. To represent the iris pattern efficiently, a new method for optimizing the dimension of feature vectors using wavelet transform is proposed. In order to increase the recognition accuracy of competitive learning algorithm, an efficient initialization of the weight vectors and a new method to determine the winner are also proposed. With all of these novel mechanisms, the experimental results showed that the proposed system could be used for personal identification in an efficient and effective manner.

Modeling of refreshing sleep aid service based on ubiquitous computing

This paper describes that we propose context-aware computing based model in ubiquitous computings subset for refreshing of humans sleep and examine technological possibility. Human can do fatigue recovery and growth through sleep and we propose methodology with model form in more effective and most suitable context for refreshing sleep in this paper.

Service compatibility and interaction issues in context-aware intelligent ubiquitous home

In this article, we discuss interaction issues about conflicts and interferences between various context information and services in the intelligent ubiquitous home environment. Context information conflicts occur when classifying services based on various sensing information. Service interactions are generated when services are applied in different controls simultaneously for the same device. In order to solve the problem, the paper utilizes the context-aware based service compatibility and interaction solving mechanism in intelligent ubiquitous home

Secure Online Internet Reservation of E-Government Service (ORGS) Using Java Card Applications Toolkit (J-CAT)

e-Government service will be beneficial to citizen for its efficiency and security. From the citizen’s point of view, although the e-Government service brings citizens online service, citizens need to visit at least one of the governmental offices for receiving specific service such as job interview, ID card issuing, or attending social events. These kinds of service often accompany with payments for the service charge. It means that the citizen has to do three things: request for reservation, payment for the service, and visit at least one of the governmental offices. In this paper, the Online Internet Reservation of e-Government Service(ORGS), one of the smart card based models for online Internet reservation accompanying with payment, is proposed for its efficiency and security of these kinds of service. Java card security functions are part of the building components of the Java Card Applications Toolkit(J-CAT). The J-CAT is designed to provide developers with security, easiness and efficiency in developing Java card applications. In this paper, the proposed ORGS implemented by the J-CAT provides users, developers, and managers with security, easiness, and efficiency.

Efficient iris recognition system by optimization of feature vectors and classifier

This paper presents an efficient system for recognizing the identity of a living person on the basis of iris patterns that are one of the physiological and biological features with high reliability. After various preprocessing are conducted for the iris data acquired by a CCD camera and an image grabber, feature vectors are extracted using Wavelet transform. A competitive learning neural network is used to classify the patterns. To represent the iris pattern efficiently, a new method for optimizing the dimension of feature vectors is proposed without any influence to the system performance. In order to increase the recognition accuracy of compjetitive learning algorithm, an efficient initialization of the weight vectors and a new method to determine the winner are also proposed. With all of these novel mechanisms, the experimental results showed that the proposed system could be used for personal identification in an efficient and effective manner.