Young-Sik Chung

A concurrent program debugging environment using real-time replay

A common debugging strategy involves re-execution of a program (on a given input) over and over, each time gaining more information about bugs. Such techniques can fail in concurrent programs. Because of variations in message latencies and process scheduling, different runs on the given input may produce different results. This non-repeatability rules out the reproduction of errors which is the cornerstone of conventional debugging techniques. So, guaranteeing reproducibility is major issue in the concurrent program debugging. This paper discusses the design and implementation of a concurrent program debugging environment which replays the error-occurred execution and debugs the errors using a recorded event history file and input data.

VoD streaming synchronization mechanism with multiple users for service continuity

In this paper, we propose multiple users-based streaming service synchronization mechanism for streaming delivery and synchronization in wire-wireless service environment and analyze the available network performance. Current structure of streaming service for multiple users is not designed to be suitable for the purpose of service continuity. So, multiple users cannot be provided with seamless media experience in wire-wireless service environment. Extending the concept from one user to multiple users, we identify the possibility of service continuity. We introduce the concept of streaming service synchronization, and implement VoD streaming service using proposed synchronized mechanism, and then analyze the available network performance. Finally we discuss the feasibility and practicality in aspect of service deployment.

High-level CHILL debugging system in cross-development environments

CHILL is a concurrent programming language, especially for implementing telecommunications software. ETRI has used the language for the development of switching software. Generally, switching software is characterized by real-time execution, parallel and distributed processing, large scale source code, and high complexity. The software is also developed on cross development environments. In this paper, we present a CHILL cross debugging system, in which the system can test and debug the programs on remote system from local system. The system provides the traditional debugging features such as executing and tracing a program, listing source code, setting breakpoints, examining and setting program locations, and single-stepping. It also provides parallel and distributed real-time debugging, a powerful command language, signal sending capability, and graphical user interface for switching software.

Integrated frame switch network guaranteeing customized end-to-end qualities

As a unified transport network, we propose a network transport technology, uEthernet (universal Ethernet), that can provide a straight forward way to reserve network resources in time-aware manner over time-unaware requests, and can support all types of service with a customized quality requirement via a single network. We have introduced a definition of the customized end-to-end quality, and the concept that assigns the network resources for loss-free average rate reservation while keeping synchronization of representation between consecutive arrived application data. The proposed network enables the transport of end user data traffic that has wide-range of bandwidth requirements, such as several kbps for a sensor network application, 220 Mbps, even 1.5 Gbps for immersive media services. Nevertheless, by adopting simple and reliable control and signaling mechanism, the architecture is scalable and efficient to deliver QoS for synchronous traffic as well as asynchronous traffic in a combined manner

Design and implementation of software simulation platform for ATM switching systems

Since the target system, where the software for an ATM switching system will be loaded, has limited resources and various functional constraints, it is difficult to develop and test ATM software on the target system. Therefore, its development and testing are accomplished usually by a general UNIX system. However, testing environment developed until now has emphasized concurrent processing simulation of CHILL language, which means that there have been a great deal of restrictions in testing software of switching systems. In order to resolve this problem, we develop software simulation platform system (SSPS), which is a testing environment to observe and analyze actions of software of ATM switching systems by simulating target environment including the operating system of switching systems at a general UNIX system. SSPS is characterized in that it supports distributed concurrent environment, provides a communication function with an operation and management system, add facilitates extension to testing environment for other systems. The test coverage over all ATM S/W blocks has increased by SSPS.

Hybrid Cache Architecture Using Big Data Analysis for Content Delivery Network

Hybrid cache architecture is the purpose of reducing network traffic in real time. For this, this paper suggest of multi-layer cache, collaborative cache and adaptive transformation using big-data analysis. Multi-layer cache compose of parent cache and edge cache. Parent cache has large scaled DB. Edge cache has small memory and high performance network ability. Each content stripes into edge cache. And each cache has a monitor agent for collecting cache status. This monitor agent sends log status to big data platform. Big data analysis engine analyses the information and determines reconfiguration of current cache layout using analysis result.

Debugging Environment for a Large Scale Telecommunication System

Testing and debugging of real-time, embedded, concurrent software system is costly and time consuming. In order to reduce the development costs of the large-scale telecommunication system, TDX-10, we developed a CHILL Debugging Environment(CDE). CDE is a system that helps us to develop software for large-scale telecommunication systems with high productivity and reliability. It gives us effective debugging facilities such as logic tests, signal tracing for integration test, and assists us in cross debugging the software that runs on a target system. With CDE, we can not only test and debug the software on a host computer without downloading them onto a target system but also cross debug the software that runs on a target system. To simulate the software on the host, CDE provides facilities for concurrent processing that are available in the target system. This paper describes the design and implementation of CHILL debugging environment for a large-scale telecommunication system.

Real-time adaptive streaming mechanism with device capability for resource efficiency

In this paper, we propose a device capability-based adaptive streaming mechanism for resource efficiency in wire-wireless service environment and analyze resource utilization efficiency. If the same multiple devices share streaming content, this technique is to allow a transformation thread to run as a minimum. Current structure of real-time streaming service for multiple devices is not designed to be suitable for the heterogeneous device based service. So, multiple devices cannot be provided with efficient media experience in wire-wireless service environment. By extending the concept from one device to multiple devices, we identify the possibility of device capability based real-time streaming service. In this paper, we seek to provide some insights on this viewpoint through empirical experiments driven by wireless streaming service. Our results suggest that real-time adaptive streaming for resource efficiency is an effective solution for delivering real-time streaming experience.

Design and implementation of u-service manager in open service framework

User centric service in mobile environment has become a new issue in the area of service convergence with the advent of versatile mobile devices. We propose an open service framework with support for service mobility that executes contents and applications in a dynamic environment. We put emphasis on u-service manager of the open service framework in this paper. For the u-service manager, mechanisms for device profiling, user preference learning, and profile-based service recommendation are designed to enable device-capability-aware service recommendation functionality. A prototype system integrates functionalities with proprietary protocol to support multiple device platforms. The prototype shows the capability of the u-service manager that controls service sessions in real-time under service mobility. Also it shows that the u-service manager provides the dynamic configuration environment appropriate to the user and the targeted terminal device.

A Design and Implementation of u-Service Framework

Service mobility has become a new issue in the area of service convergence with the advent of versatile mobile devices. Hence, we propose an open service framework with support for service mobility that executes contents and applications in a dynamic environment. For the framework, the concept and role of a device and its application for a service are redefined into the new concepts of device, execution engine, and service. Mechanisms for device profiling, user preference learning, and profile-based service recommendation are designed to enable device-capability-aware service recommendation functionality. A prototype system integrates functionalities with proprietary protocol and a content transcoder to support multiple device platforms. The prototype shows the capability of the framework in service mobility support and its advancement into an open international application platform.