Hyunsang Youn

Deriving Queuing Network Model for UML for Software Performance Prediction

It is an important issue for software architects to estimate the performance of software in the early stage of development process due to the needs to verify QoS. Queueing network model is a very useful tool to analyze the performance of a system from abstract model. In this paper, we propose a transformation technique from UML into queueing network model. This approach avoids the need for a prototype implementation since we can determine the overall form of performance equation from the architectural design description. We prove the accuracy of derived queueing network model, which is summarized at 85 percent, through a ubiquitous commerce system which extends mobile commerce system developed by our prior work.

Automatic Detection of Design Pattern for Reverse Engineering

In maintenance, the lack of documentation leads to high costs of reverse engineering. Generally, design-pattern is a reusable solution to a commonly occurring problem in software design. If design-patterns could be captured and reused in reverse engineering, the reverse engineering would be very helpful those who develops and maintains software. So there have been many attempts to detect design-patterns during reverse engineering. However, the approaches suffer from serious drawbacks to its practical implementation; false positive, false negative rate, the number of detected patterns. In this paper, we propose a new taxonomy of GoF design patterns that can guide the reverse-engineering process. This approach not only combines static analysis with dynamic analysis but also adds what we call the implementation- specific analysis. We apply a number of existing and new applications, including PURE toolkit, JINI based home application system, project management tool, MP3 player, and we demonstrate that the reverse engineering process is more accurate.

Security based survivability risk analysis with extended HQPN

Analysis of software survivability in the early development phase is very important to validate and specify software architecture. Specifically, quantitative evaluation of survivability is very useful to determine the architecture and to estimate the risk. The risk factor can be quantified as a combination of the probability that a software system may be failed through security threat and the severity of the damages caused by the attack. In this paper, we devise a methodology for analysis of risk factor which originates from violations of security goal. We elaborate Extended Hierarchically combined Queueing Petri Nets (E-HQPN) to estimate the survival failure probability with regard to attack and combines it with the severity of the failure consequence obtained using the Functional Failure Analysis. We apply the methodology on the development of an e-business application using step-by-step approach.

Improved location management scheme based on autoconfigured logical topology in HMIPv6

Though some studies involving general micro-mobility exist, micro-mobility research concerning a mobile node (MN) moving between Mobile Anchor Points (MAP) is lacking. In Hierarchical Mobile IPv6 (HMIPv6), a MN sends a binding update (BU) message to a MAP when the MN performs a handoff. This scheme reduces the handoffs and BUs performance costs. However, the total signaling cost of HMIPv6 rapidly increases with the number of corresponding nodes (CN) for a MN moving among MAPs. In this paper, we propose a scheme based on logical topology with autonomous configuration allowing a MN to move around between MAPs without performing a BU. We perform the performance evaluation for packet delivery and location update costs through a novel analytic approach. We show that the proposed scheme is approximately two times better than the existing one in terms of the total signaling costs through performance analysis. Furthermore, our scheme also reduces the location management cost effectively.

Performability analysis method from reliability and availability

Availability and reliability are important attributes of dependability. Many dependability analysis methodologies measure and predict availability and reliability separately. However, those methodologies cannot reflect real system characteristics since availability and reliability are interdependent and occur simultaneously. We propose a performability analysis method which regards availability and reliability at the same time. The proposed performability analysis method contains discrete time Markov Chain modeling and its analysis.

Software Dependability Analysis Methodology

Dependability can be verified at the integration phase of the software development life cycle. However the dependability verification processes that inspect software dependability in the late period of development have an effect in the development cost (e.g. time, human resources). Therefore, it is a very important issue to verify any nonfunctional requirements in the early stages of the development process. In this paper, we propose a software dependability analysis methodology of distributed component based software by using HQPNs (Hierarchically Combined Queuing Petri Nets) modeling. We prove the validity of the proposed methodology by applying it to a video conference system development.

Service oriented architecture implementation model for enterprise collaboration

Since the computer era began, companies built various information systems in order to increase efficiency and fulfill the customer requirement. While the number of systems doubled, the number of interfaces linking the systems increased and the interface type became various. Accordingly, the management of interface became a big challenge. This paper suggests SOA as system interface and a step-by-step implementation model of SOA. In addition, a prototype of executing the model is shown for detail explanation. The prototype demonstrates how SOA could increase user convenience.

Next generation agent development supporting tool: case study

Agent based development is a new technology that has been recently used in many domains. However, the development of an agent-based system can be difficult and time-consuming for inexperienced developers. To address this problem, in this paper, we propose a pattern based agent development and support tool. This approach facilitates rapid agent development and addresses common design problem. The agent patterns are classified according to function. We apply this approach for agent development, based on a travel assistant scenario.

Evaluation of Real-Time Characteristics for Mission Critical Systems Using Hierarchically Profiled Petri Net

Development of mission-critical systems requires the verification and the validation (V&V) considering limited time, available resource and costs to analyze the possibility of mission accomplishment in system development life cycle. In this paper we propose a Petri Net based modeling formalism, called Hierarchically Profiled Petri Net (HPPN), which provides a unified solution for above problems through the profiled firing rule in which the state evolves according to the state transition rule. Additionally we propose an analysis technique, which represents the multi-aspect based state enumeration that considers the state of token and the state of marking. Expressive and analysis capabilities of the model are demonstrated with reference to an intercontinental ballistic missiles (ICBM) scenario.

Automatic generation techniques of a resource monitor based on deployment diagram

The overall goal of this paper is to improve development environments of self-healing software to facilitate self-managing capabilities such as self-healing. This paper proposes using deployment diagram for self-healing approach to determine problems arising in external environments.