• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2007년도 Proceedings of ISRS 2007
• /
• pp.101-104
• /
• 2007

Efficient Query processing and optimization are critical for reducing network traffic and decreasing latency of query when accessing and manipulating sensor data of large-scale sensor networks. Currently it has been studied in sensor database projects. These works have mainly focused on in-network query processing for sensor networks and assumes homogeneous sensor networks, where each sensor network has same hardware and software configuration. In this paper, we present a framework for efficient query processing over heterogeneous sensor networks. Our proposed framework introduces query processing paradigm considering two heterogeneous characteristics of sensor networks: (1) data dissemination approach such as push, pull, and hybrid; (2) query processing capability of sensor networks if they may support in-network aggregation, spatial, periodic and conditional operators. Additionally, we propose multi-query optimization strategies supporting cross-translation between data acquisition query and data stream query to minimize total cost of multiple queries. It has been implemented in WSN middleware, COSMOS, developed by ETRI.

• ETRI Journal
• /
• 제39권2호
• /
• pp.202-212
• /
• 2017

With the advent of the Internet-of-Things paradigm, the amount of data production has grown exponentially and the user demand for responsive consumption of data has increased significantly. Herein, we present DART, a fast and lightweight stream processing framework for the IoT environment. Because the DART framework targets a geospatially distributed environment of heterogeneous devices, the framework provides (1) an end-user tool for device registration and application authoring, (2) automatic worker node monitoring and task allocations, and (3) runtime management of user applications with fault tolerance. To maximize performance, the DART framework adopts an actor model in which applications are segmented into microtasks and assigned to an actor following a single responsibility. To prove the feasibility of the proposed framework, we implemented the DART system. We also conducted experiments to show that the system can significantly reduce computing burdens and alleviate network load by utilizing the idle resources of intermediate edge devices.

• 로봇학회논문지
• /
• 제3권3호
• /
• pp.255-261
• /
• 2008

This paper describes a network framework that support network based humanoid. The framework utilizes middleware such as CORBA (ACE/TAO) that provides PnP capability for network based humanoid. The network framework transfers data gathered from a network based humanoid to a processing group that is distributed on a network. The data types are video stream, audio stream and control data. Also, the network framework transfers service data produced by the processing group to the network based humanoid. By using this network framework, the network based humanoid can provide high quality of intelligent services to user.

• 품질경영학회지
• /
• 제48권4호
• /
• pp.553-566
• /
• 2020

Purpose: The purpose of this study was to design a framework for generating one-class classification algorithm based on Hyper-Rectangle(H-RTGL) in a distributed environment connected by network. Methods: At first, we devised one-class classifier based on H-RTGL which can be performed by distributed computing nodes considering model and data parallelism. Then, we also designed facilitating components for execution of distributed processing. In the end, we validate both effectiveness and efficiency of the classifier obtained from the proposed framework by a numerical experiment using data set obtained from UCI machine learning repository. Results: We designed distributed processing framework capable of one-class classification based on H-RTGL in distributed environment consisting of physically separated computing nodes. It includes components for implementation of model and data parallelism, which enables distributed generation of classifier. From a numerical experiment, we could observe that there was no significant change of classification performance assessed by statistical test and elapsed time was reduced due to application of distributed processing in dataset with considerable size. Conclusion: Based on such result, we can conclude that application of distributed processing for generating classifier can preserve classification performance and it can improve the efficiency of classification algorithms. In addition, we suggested an idea for future research directions of this paper as well as limitation of our work.

• Journal of Information Processing Systems
• /
• 제2권2호
• /
• pp.124-129
• /
• 2006

With the explosive increase in the generation and utilization of spatiotemporal data sets, many research efforts have been focused on the efficient handling of the large volume of spatiotemporal sets. With the remarkable growth of ubiquitous computing technology, mining from the huge volume of spatiotemporal data sets is regarded as a core technology which can provide real world applications with intelligence. In this paper, we propose a 3-tier knowledge discovery framework for spatiotemporal data mining. This framework provides a foundation model not only to define the problem of spatiotemporal knowledge discovery but also to represent new knowledge and its relationships. Using the proposed knowledge discovery framework, we can easily formalize spatiotemporal data mining problems. The representation model is very useful in modeling the basic elements and the relationships between the objects in spatiotemporal data sets, information and knowledge.

• Journal of Electrical Engineering and Technology
• /
• 제6권4호
• /
• pp.573-579
• /
• 2011

To meet the scalability and performance requirements of data analyses, which often involve voluminous data, efficient parallel or concurrent algorithms and frameworks are essential. We present a high-performance Korean morphological analyzer which employs the MapReduce framework on the graphics processing unit (GPU). MapReduce is a programming framework introduced by Google to aid the development of web search applications on a large number of central processing units (CPUs). GPUs are designed as a special-purpose co-processor. Their programming interfaces are typically formulated for graphics applications. Compared to CPUs, GPUs have greater computation power and memory bandwidth; however, GPUs are more difficult to program because of the design of their architectures. The performance of the Korean morphological analyzer using the MapReduce framework on the GPU is evaluated in comparison with the CPU-based model. The proposed Korean Morphological analyzer shows promising scalable performance on distributed computing with the GPU.

• Journal of Information Processing Systems
• /
• 제16권6호
• /
• pp.1238-1249
• /
• 2020

Throughout the world, aging populations and doctor shortages have helped drive the increasing demand for smart healthcare systems. Recently, these systems have benefited from the evolution of the Internet of Things (IoT), big data, and machine learning. However, these advances result in the generation of large amounts of data, making healthcare data analysis a major issue. These data have a number of complex properties such as high-dimensionality, irregularity, and sparsity, which makes efficient processing difficult to implement. These challenges are met by big data analytics. In this paper, we propose an innovative analytic framework for big healthcare data that are collected either from IoT wearable devices or from archived patient medical images. The proposed method would efficiently address the data heterogeneity problem using middleware between heterogeneous data sources and MapReduce Hadoop clusters. Furthermore, the proposed framework enables the use of both fog computing and cloud platforms to handle the problems faced through online and offline data processing, data storage, and data classification. Additionally, it guarantees robust and secure knowledge of patient medical data.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제5권10호
• /
• pp.477-482
• /
• 2016

기업의 신속한 의사결정 및 전략적 정책 결정을 위해 빅데이터에 대한 고도 분석이 필수적으로 요구됨에 따라 대량의 데이터를 복수의 노드에 분산하여 처리하는 하둡 또는 스파크와 같은 분산 처리 플랫폼이 주목을 받고 있다. 최근 공개된 Spark SQL은 Spark 환경에서 SQL 기반의 분산 처리 기법을 지원하고 있으나, 기계학습이나 그래프 처리와 같은 반복적 처리가 요구되는 고도 분석 분야에서는 효율적 처리가 불가능한 문제가 있다. 따라서 본 논문은 이러한 문제점을 바탕으로 Spark 환경에서 고도 분석 지원을 위한 SQL 기반의 빅데이터 최적처리 엔진설계와 처리 프레임워크를 제안한다. 복수의 조건과 다수의 조인, 집계, 소팅 연산이 필요한 복합 SQL 질의를 분산/병행적으로 처리할 수 있는 최적화 엔진과 관계형 연산을 지원하는 기계학습 최적화하기 위한 프레임워크를 설계한다.

• 한국정보통신학회논문지
• /
• 제23권10호
• /
• pp.1179-1187
• /
• 2019

오늘날의 기술 환경에서 대다수의 빅 데이터 기반 애플리케이션 및 솔루션은 스트리밍 데이터의 실시간 처리를 기반으로 한다. 빅 데이터 스트림의 실시간 처리 및 분석은 빅 데이터 기반 애플리케이션 및 솔루션 개발에서 중요한 역할을 한다. 특히 해사 분야 데이터 처리 환경에서도 데이터의 폭발적 증대에 따른 대용량 실시간 데이터를 빠르게 처리 및 분석할 수 있는 기술 개발의 필요성이 가속화되고 있다. 따라서 본 논문에서는 다양한 빅 데이터 처리를 위한 오픈소스 기술 중에 적합한 오픈소스로 NiFi, Kafka, Druid의 특징을 분석하여 한국형 e-Navigation 서비스에서 해사 분야 서비스 분석에 필요한 외부 연계 필요 정보들을 상시 최신 정보로 제공할 수 있도록 실시간 데이터 처리를 위한 개방형 데이터 프레임워크 기술 적용의 기초를 마련하고자 한다.

• 정보과학회 논문지
• /
• 제44권1호
• /
• pp.84-94
• /
• 2017

최근 다양하고 방대한 양의 데이터를 처리하기 위해 빅데이터의 특성인 5V(Volume, Variety, Velocity, Veracity, Value) 중에서도 속도(Velocity)의 중요성이 강조되면서 대량의 데이터를 빠르고 정확하게 처리하는 기술인 실시간 스트림 처리(Real-time Stream processing)를 위해 많은 연구가 진행되고 있다. 본 논문에서는 실시간 빅데이터 처리를 위해 대표적인 실시간 객체 모델인 TMO(Time-triggered Message-triggered Object) 개념을 도입한 Squall 프레임워크를 제시하고, 단일 노드에서 동작하는 Squall 프레임워크와 그 동작들에 대해 기술한다. TMO는 작업을 수행할 때, 특정 조건에 대해 실시간으로 처리하는 비주기적인 처리방법과 일정 시간 간격동안 주기적인 처리를 지원하는 객체 모델이다. 따라서 Squall 프레임워크는 실시간 빅데이터의 실시간 이벤트 스트림 및 마이크로-배치 처리를 동시에 지원하고, 기존 아파치 스톰과 스파크 스트리밍 대비 상대적으로 우수한 성능을 제공한다. 하지만 Squall은 대부분의 프레임워크에서 제공되는 다중 노드에서의 실시간 분산처리를 위한 추가적인 개발이 필요하다. 결론적으로, TMO 모델의 장점은 실시간 빅데이터 처리시 기존 아파치의 스톰이나 스파크 스트리밍의 단점들을 극복할 수 있다. 이러한 TMO 모델은 실시간 빅데이터 처리에 있어 유용한 모델로서의 가능성을 가지고 있다.