• Annual Conference of KIPS
• /
• 2021.05a
• /
• pp.82-85
• /
• 2021

최근 초연결화를 근간으로 한 스마트 홈 구성을 위해 스마트 홈 내부에 센서를 탑재한 디바이스가 증가하고 있으며, 이를 효과적으로 사용하기 위해 빅데이터 처리 시스템이 활발하게 도입되고 있다. 그러나 기존 빅데이터 처리 시스템은 분산노드에 할당되기 전 모든 요청이 클러스터 드라이버로 향하기 때문에 동시에 많은 요청이 발생하는 경우 분할 작업을 관리하는 클러스터 드라이버에 병목현상이 발생함에 따라 네트워크를 공유하는 클러스터 전체의 성능감소로 이어진다. 특히 작은 데이터 처리를 지속적으로 요청하는 스마트 홈 디바이스에서 지연율이 더 크게 나타난다. 이에 본 논문에서는 동시간에 빈번한 요청이 발생하는 스마트 홈 환경에서 효과적인 데이터 처리를 위한 기계학습 기반 캐싱 시스템을 설계하였다.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.13 no.3
• /
• pp.60-71
• /
• 2017

Smart devices such as smart phones, smart TVs, and smart pads have become essential devices in recent years. As the popularity and demand grows, the performance of smart devices is also getting better and users are dealing with a lot of things such as education and business using smart devices instead of desktop. However, smart devices that still have poor performance compared to desktop, even with improved performance, have difficulty running high performance applications due to limited resources. In this paper, we propose a load balancing algorithm applying the characteristics of smart devices to overcome the resource limitations of devices. in order to verify the algorithm, we implemented the algorithm after adding the distributed processing system service in Android platform. After constructing the cluster on the smart device, various experiments were conducted. Through the analysis of the test results, it is confirmed that the proposed algorithm efficiently improves the overall distributed processing performance by effectively aggregating different amounts of computing resources in heterogeneous smart devices.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.6
• /
• pp.1441-1448
• /
• 2013

The propose of this study is to apply establish the foundation for development of elder driver friendly smart cluster. At first, vehicle dashboard trends were analyzed. Secondly, we presents the structure of elder driver friendly smart cluster and explains android based meter cluster. also, we shows the implementation details and experimental result of elder driver friendly smart cluster system. And we presents a summary and conclusions.

• It's Smart Media
• /
• v.1 no.4
• /
• pp.40-48
• /
• 2012

• It's Smart Media
• /
• v.4 no.1
• /
• pp.1-3
• /
• 2015

• Journal of the Economic Geographical Society of Korea
• /
• v.19 no.4
• /
• pp.799-811
• /
• 2016

After the participatory government began, various cluster policies in explicit and tacit forms had been promoted. However, an opinion of coming up with new policy alternative different from the existing one is recently brought up for strengthening the competitiveness of industrial agglomerations. This research attempts to discuss the ways in which both a smart specialization strategy and a triple-helix innovation system approach, as an alternative approach to regional industrial policy, are theoretically associated with the existing cluster policy. Through this discussion, it highlights that post-cluster policy should be not just based on regional specificity, but also facilitated by establishing the consensus space of innovation on the bassis of voluntary cooperation among industry, academy and government. It also stresses that it is necessary to focus on nurturing a new industry by systematic and intensive investment and the diversification of industrial cluster for reinforcing competitiveness of local universities and revitalizing practical cooperation between industry and university.

• Journal of the Economic Geographical Society of Korea
• /
• v.19 no.4
• /
• pp.764-781
• /
• 2016

During the last two decades, industrial cluster policies for promoting regional economic growth and industrial development have been flourishing all over the world. Even though cluster policies have partly contributed to regional industrial growth and innovation capabilities, they have long been blamed for regional industrial lock-ins and declining regional industrial resilience because of applying homeogenous cluster policies and regional specialized strategic industrial promotion policies for various localities, which are based on so-called 'selection and concentration' principle. This paper suggests postcluster policy focused on placed-based smart specialization and regional business platform strategies.

• Smart Media Journal
• /
• v.8 no.4
• /
• pp.46-52
• /
• 2019

The size of data generated and collected at industrial sites or in public institutions is growing rapidly. The existing data processing server often handles the increasing data by increasing the performance by scaling up. However, in the big data era, when the speed of data generation is exploding, there is a limit to data processing with a conventional server. To overcome such limitations, a distributed cluster computing system has been introduced that distributes data in a scale-out manner. However, because distributed cluster computing systems distribute data, inefficient use of network bandwidth can degrade the performance of the cluster as a whole. In this paper, we propose a scheme that compresses data when transmitting data in a Hadoop cluster considering network bandwidth. The proposed scheme considers the network bandwidth and the characteristics of the compression algorithm and selects the optimal compression transmission scheme before transmission. Experimental results show that the proposed scheme reduces data transfer time and size.

• Smart Media Journal
• /
• v.6 no.3
• /
• pp.68-74
• /
• 2017

Big data processing processes various types of data such as files, images, and video to solve problems and provide insightful useful information. Currently, various platforms are used for big data processing, but many organizations and enterprises are using Hadoop for big data processing due to the simplicity, productivity, scalability, and fault tolerance of Hadoop. In addition, Hadoop can build clusters on various hardware platforms and handle big data by dividing into a name node (master) and a data node (slave). In this paper, we use a fully distributed mode used by actual institutions and companies as an operation mode. We have constructed a Hadoop cluster using a low-power and low-cost single board for smooth experiment. The performance analysis of Name node is compared through the same data processing using single board and laptop as name nodes. Analysis of influence by number of data nodes increases the number of data nodes by two times from the number of existing clusters. The effect of the above experiment was analyzed.

• Smart Media Journal
• /
• v.8 no.1
• /
• pp.43-50
• /
• 2019

As numerous devices are connected through a wireless network, there exist many studies conducted to efficiently connect the devices. While earlier studies often use clustering for efficient device management, there is a load-intensive cluster node which may lead the entire network to be unstable. In order to solve this problem, we propose a scheduling model for centralized unequal chain clustering for efficient management of sensor nodes. For the cluster configuration, this study is based on the cluster head range and the distance to the base station(BS). The main vector projection technique is used to construct clustering with concentricity where the positions of the base stations are not the same. We utilize a multiple radio access interface, multiple-input multiple-output (MIMO), for data transmission. Experiments show that cluster head energy consumption is reduced and network lifetime is improved.