Volume 4, Number 4, December 2003


Title:

Fault-based Distributed Recovery Block (FDRB) for Switching Systems
 

Author & Affiliation:

Byung Sun Lee (ETRI), Yong Rae Kwon (KAIST)
Abstract: (DRB) is a well-known fault-tolerant technique for real-time systems which treats hardware and software faults uniformly and provides a forward recovery scheme. However, DRB has the constraint that it uses only one version of a full application program running on a computing station as the smallest unit of software redundancy. Therefore, we propose a new fault tolerant technique specially designed to be more useful for large real-time, especially switching systems. Our technique is based on measuring the types and effects of faults combined with testing experience in switching system software development. This approach is Fault-based DRB (FDRB) with a self-checking and selective recovery mechanism added to it. We compared our approach with ordinary DRB using model analysis and simulations with injected faults. The reliability in FDRB is not less than in DRB due to the fault recovery of the recovery-handling programs and the fault recovery time is reduced without sacrificing software reliability.
   
Title:

Robust Image Watermarking using Rotationally Symmetric Masks

 
Author & Affiliation:

Hyon-Gon Choo & Whoi-Yul Kim (Hanyang University)

 
Abstract:

In this paper, a new watermarking method robust to geometric transformations such as rotation, scale and translation is proposed. In order to cope with the rotation of an image, we use a mask with a rotationally symmetric pattern onto the Fourier transformed image to utilize the translation invariance property of the Fourier transform. Watermark messages are embedded in forms of the positional variation and periodic energy distribution of a mask in the frequency domain. To extract the inserted information, radial projection of the energy in frequency domain is first calculated and its correlation with the reference profile determines the existence of the hidden information. Experimental results show that the proposed method is robust to various geometric transformations and filtering distortion.

 
Title:

Data Mining Methods for Alert Correlation Analysis

 
Author & Affiliation:

Moon Sun Shin & Keun Ho Ryu (Chungbuk National University)

 
Abstract:

With the growing deployment of host and network intrusion detection systems, managing the reports from these systems become critically important. We propose data mining methods for analysis of these alert reports. Because data mining tasks deal with the discovery of implicit data, we implement three mining methods to support the alert analyzer. The alert analyzer with mining engine can discover the interconnection and inter relationships. So the results of analyzing the alert data are used for the security policy server to construct the security policy rule efficiently in the framework of PBNM (Policy Based Network Management). It helps not only to manage the bad users and hosts but also to discover possible alert sequences. In this paper, we also provide the groundwork to apply the data mining techniques for alert correlation analysis.

 
Title: Handwritten Number Recognition using Contour Measure Band in the Presence of Boundary Box
 
Author & Affiliation: Young-Bin Kwon (Chung-Ang University)
 
Abstract: This paper deals with the problem of handwritten number recognition of ZIP code. In general, zip code recognition is accomplished in three steps. The first step is to find the location of the zip code area, the second step is to segment each number from the zip code area, and the third step is to recognize the segmented number. Among these three parts, this paper concentrates on the handwritten number recognition. We tried to recognize zip code using structural features. To extract the structural features of number, we propose a contour analysis method using measure band (MB), which compute the local features. The local features are merged into global features in accordance with neighbour features. Modules for each numeral calculate the score by comparing input features with its own features. Experimental results show sufficient correct recognition ratio on unconstrained real hand-written envelope numbers.
 
Title: A Classification Technique for Components based on Six Views
 
Author & Affiliation: Eun Sook Cho & Seo-Jung Lee (Dongduk Women’s University)
 
Abstract: As component-based reuse technology is being introduced, software components are more likely distributed on Internet. In order to promote Internet based on-line distribution, distribution model of components should be defined and a distribution system should also be constructed. Furthermore, to run the component distribution system effectively, a logical and practical schema for classifying components should be defined and standardized.In this paper, we propose a classification schema using BNF and representation standards of components using six different views. To verify the effectiveness of proposed classification and compare to other classification methods, we assess proposed classification by measuring hit ratio and correctness. By using the proposed methods that support six views on components, the search for right components and registering new components can be done more effectively.
 
Title: Role-based Model and Architecture for Workflow Systems
 
Author & Affiliation:

Kwang-Hoon Kim & Jae-Kang Won (Kyonggi University), Byong-Ok Jang (Korea Nazarene University)

 
Abstract: This paper formally defines a role-based model in order eventually to provide a theoretical basis for role-oriented distribution in designing and implementing a coordination system. We would, at first, apply the model to the typical coordination system - workflow. The formal role-based model is based on the extended information control net (ICN) that has been widely used for specifying workflow procedures in an organization. The role-based model is primarily devised so as for a workflow management system to be effectively adapted for the recent working behaviors in an office especially in oriental working cultures and environment. That is, the recent working behaviors and task assignment patterns have been changing from uni-cast based behavior to any-cast or/and multi-cast based behavior. To support the uni-cast behavior, which is the way of the traditional workflow systems, is that the workflow system pushes a task to an actor who is one of the actors associated with the role that is responsible to perform the task. The any-cast behavior means that the workflow system pushes a task to a role, then one of the actors associated with the role pulls the task to perform, on the other hand, the multi-cast behavior is for a group of actors associated with the role to cooperatively perform the task. The multi-cast behavior requires embedding cooperative (or real-time) groupware concepts into the workflow system – This is a hot issue on the engineering data processing area such as cooperative CAD/CAM systems. The role-based model should be the theoretical basis to embed the any-case behavior as well as the multi-cast behavior into workflow systems. This paper introduces, justifies, and formalizes notions of roles and their dependencies, which are associated with a workflow procedure. As a practical application of the role-based model, we try to design a role-based workflow architecture that should be appropriate for the basis of role-based workflow enactment.
 
Title:

An Adaptive Distributed Location Management Algorithm for Mobile Networks

 
Author & Affiliation:

Sun-Jin Oh (Semyung University), Sung-Kwang Chun (ETRI), Ihn-Han Bae (Catholic University of Daegu), Stephan Olariu (Old Dominion University)

 
Abstract:

An important issue in the design of future Personal Communication Service (PCS) networks is efficient management of location information. In this paper, we propose an adaptive distributed location management algorithm that stores the position of the mobile terminal (MT) in k out of the n location information databases (LIDs). The proposed algorithm chooses replication factor k adaptively according to both space locality of the LIDs in wireless mobile networks and location query popularity from re mote MTs to local MT. The performance of a proposed algorithm is to be evaluated by both an analytical model and a simulation. Based on the results of performance evaluation, we know that the performance of the proposed algorithm is better than that of Krishnamurthi's algorithm [5,6] regardless of call-mobility ratio.

 
Title:

Application Transition from IPv4 to IPv6 Network

 
Author & Affiliation:

YongGeun Hong, MyungKi Shin & HyoungJun Kim (Electronics and Telecommunications Research Institute)

 
Abstract:

In this paper, we will consider the application aspects of IPv6 transition from IPv4 to IPv6 network. As IPv6 is deployed, there are some problems that application developers and administrators face because of the difference between IPv4 mechanism and IPv6 mechanism. Having a dual stack node doesn’t mean having both IPv4 and IPv6 application. Call connections by DNS name resolution sometimes will fail. System administrators may have various versions of same application. This paper clarifies the problems occurring in transition periods between IPv4 applications and IPv6 applications. We will also describe guidelines to implement and choose the suitable application under the various transition environments.

 
Title: Synchronization and Optimization of Conformance Test Sequence for Protocol Testing
 
Author & Affiliation: Chul Kim (Yongin University)
 
Abstract: This paper presents a new technique for generating an optimum synchronizable test sequence which defines a set of phases that constructs a tester-related digraph from a protocol specification such that a minimum cost tour of the digraph with intrinsically synchronizable transfer sequences (ISTSs) can be used to generate an optimum synchronizable test sequence using synchronizable state identification sequences (SSISs). As a result, this hybrid approach with a heuristic and optimization technique provides a simple and elegant solution to the synchronization problem that arises in the distributed architecture of protocol testing.
 
Title: Clustering Analysis Technique for Component Identification and Integration
 
Author & Affiliation: Haeng-Kon Kim (Catholic University of Daegu)
 
Abstract: CBD is a software development approach based on reusable component and supports easy modification and evolution of software. For the success of this approach, a component must be developed with high cohesion and low coupling. In this paper, we propose the two types of clustering analysis technique based on affinity between use-cases and classes and propose component identification method applying to this technique. We also propose component reference model and CBD methodology framework and perform a case study to demonstrate how the affinity-based clustering technique is used in component identification method. Component identification method contains three tasks such as component extraction, component specification and component architecting. This method uses object-oriented concept for identifying component, which improves traceability from analysis to implementation and can automatically extract component. This method reflects the low coupling-high cohesion principle for good modularization about reusable component.
 
Title: A Framework for Adapting Software Process Maturity Models for Small Organizations
 
Author & Affiliation: Marta Lopez (Universidad Complutense de Madrid)
 
Abstract: This paper presents a framework for adapting maturity models for the purpose of performing software process assessments at small organizations. The results obtained at several small Spanish businesses were analyzed, and a series of factors capable of improving the process of adapting the maturity models applied in software process assessments were identified. The most important finding was that the features of the software process for assessment, defined as assessment criteria, need to be explicitly identified. Considering these criteria, relationships can be established with the part of the model that describes each feature and with the techniques used to gather data on that feature (questionnaire, interviews, etc.). These relationships were used to develop a framework by means of which a maturity model can be adapted to each organization and software process conducted.
 
Title: Issues Related to Mining Association Rules and Algorithms
 
Author & Affiliation: Hae Sool Yang (Hoseo University), Roger Lee (Central Michigan University)
 
Abstract: The development of the data storage technology makes it possible for mining the association rules from the large database through multiple passes over the database. But in some situations that is not an efficient way. Furthermore, the results may not be the interesting rules that the user expected. So provided a good standard to categorize the association rules, and along with the proper algorithms, is very important to optimize the process of discovering the association rules over the large database. From the framework for discovery of the association rules we find that a user interactive role in the frame has a significant effect on the success of the data mining process. In this article the variation of association rule and corresponding algorithm are discussed.