• Journal of Digital Convergence
• /
• v.12 no.11
• /
• pp.379-386
• /
• 2014

In this paper, the processing speed of Grabcut algorithm in order to efficiently improve the GPU (Graphics Processing Unit) for processing the data from the method. Grabcut algorithm has excellent performance object detection algorithm. Grabcut existing algorithms to split the foreground area and the background area, and then background and foreground K-cluster is assigned a cluster. And assigned to gradually improve the results, until the process is repeated. But Drawback of Grabcut algorithm is the time consumption caused by the repetition of clustering. Thus GPGPU (General-Purpose computing on Graphics Processing Unit) using the repeated operations in parallel by processing Grabcut algorithm to effectively improve the processing speed of the method. We proposed method of execution time of the algorithm reduced the average of about 95.58%.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.16 no.4
• /
• pp.407-417
• /
• 2003

Enen though several GA-based optimization algorithms have been successfully applied to complex optimization problems in various engineering fields, GA-based optimization methods are computationally too expensive for practical use in the field of structural optimization, particularly for large- scale problems. Furthermore, a successful implementation of GA-based optimization algorithm requires a cumbersome and trial-and-error routine related to setting of parameters dependent on a optimization problem. Therefore, to overcome these disadvantages, a high-performance GA is developed in the form of distributed hybrid genetic algorithm for structural optimization on a cluster of personal computers. The distributed hybrid genetic algorithm proposed in this paper consist of a simple GA running on a master computer and multiple μ-GAs running on slave computers. The algorithm is implemented on a PC cluster and applied to the minimum weight design of steel structures. The results show that the computational time required for structural optimization process can be drastically reduced and the dependency on the parameters can be avoided.

• Journal of Digital Contents Society
• /
• v.18 no.5
• /
• pp.811-820
• /
• 2017

For content-based audio retrieval which is one of main functions in audio service, the techniques for extracting fingerprints from the audio source, storing and indexing them in a database are widely used. However, if the fingerprints of new audio sources are continually inserted into the database, there is a problem that space efficiency as well as audio retrieval performance are gradually deteriorated. Therefore, there is a need for techniques to support efficient expansion of audio database without periodic reorganization of the database that would increase the system operation cost. In this paper, we design a content-based audio retrieval system that solves this problem by using MapReduce and NoSQL database in a cluster computing environment based on the Shazam's fingerprinting algorithm, and evaluate its performance through a detailed set of experiments using real world audio data.

• Journal of KIISE:Information Networking
• /
• v.30 no.5
• /
• pp.649-655
• /
• 2003

Network-based computing such as cluster computing requires a reliable high-speed transport protocol. TCP is a representative reliable transport protocol on the Internet, which implements many mechanisms, such as flow control, congestion control, and retransmission, for reliable packet delivery. This paper, however, finds out that Myrinet does not incur any packet losses caused by network congestion. In addition, we ascertain that Myrinet supports reliable and ordered packet delivery. Consequently, most of reliable routines implemented in TCP produce unnecessarily additional overheads on Myrinet. In this paper, we show that we can attain the reliability only by flow control on Myrinet and propose a new reliable protocol based on UDP named RUM (Reliable UDP on Myrinet) that performs a flow control. As a result, RUM achieves a higher throughput by 45% than TCP and shows a similar one-way latency to UDP.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.5
• /
• pp.942-947
• /
• 2008

A cluster server usually consists of a front end node and multiple backend nodes. Though increasing the number of bookend nodes can result in the more QoS(Quality of Service) streams for clients, the possibility of failures in backend nodes is proportionally increased. The failure causes not only the stop of all streaming service but also the loss of the current playing positions. In this paper, when a backend node becomes a failed state, the recovery mechanisms are studied to support the unceasing streaming service. The basic techniques are hewn as providing very high speed data transfer rates suitable for the video streaming. However, without considering the architecture of cluster-based VOD server, the application of these basic techniques causes the performance bottleneck of the internal network for recovery and also results in the inefficiency CPU usage of backend nodes. To resolve these problems, we propose a new failure recovery mechanism based on the pipeline computing concept.

• KNOM Review
• /
• v.23 no.1
• /
• pp.43-50
• /
• 2020

As General Purpose Graphics Processing Unit (GPGPU) recently plays an essential role in high-performance computing, several cloud service providers offer GPU service. Most cluster orchestration platforms in a cloud environment using containers allocate the integer number of GPU to jobs and do not allow a node shared with other jobs. In this case, resource utilization of a GPU node might be low if a job does not intensively require either many cores or large size of memory in GPU. GPU virtualization brings opportunities to realize kernel concurrency and share resources. However, performance may vary depending on characteristics of applications running concurrently and interference among them due to resource contention on a node. This paper proposes GPU container co-execution framework with multiple server creation and execution based on Kubernetes, container orchestration platform for measuring interference which may be occurred by sharing GPU resources. Performance changes according to scheduling policies were investigated by executing several jobs on GPU. The result shows that optimal scheduling is not possible only considering GPU memory and computing resource usage. Interference caused by co-execution among applications is measured using the framework.

• Journal of Internet Computing and Services
• /
• v.18 no.2
• /
• pp.61-73
• /
• 2017

In this paper, we propose a Documents Cluster Labeling method using information content of words in clusters to understand what the clusters imply. To do so, we calculate the weight and frequency of the words. These two measures are used to determine the weight among the words in the cluster. As a nest step, we identify the candidate labels using the WordNet. At this time, the candidate labels are matched to least common hypernym of the words in the cluster. Finally, the representative labels are determined with respect to information content of the words and the weight of the words. To prove the superiority of our method, we perform the heuristic experiment using two kinds of measures, named the suitability of the candidate label ($Suitability_{cl}$) and the appropriacy of representative label ($Appropriacy_{rl}$). In applying the method proposed in this research, in case of suitability of the candidate label, it decreases slightly compared with existing methods, but the computational cost is about 20% of the conventional methods. And we confirmed that appropriacy of the representative label is better results than the existing methods. As a result, it is expected to help data analysts to interpret the document cluster easier.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.11
• /
• pp.105-111
• /
• 2009

The study aims to propose the intelligent clustering technique that calculates the distance by improving the problems of multi-hop clustering technique for inter-vehicular secure communications. After calculating the distance between vehicles with no connection for rapid transit and clustering it, the connection between nodes is created through a set distance vale. Header is selected by the distance value between nodes that become the identical members, and the information within a group is transmitted to the member nodes. After selecting the header, when the header is separated due to its mobility, the urgent situation may occur. At this time, the information transfer is prepared to select the new cluster header and transmit it through using the intelligent cluster provided from node by the execution of programs included in packet. The study proposes the cluster technique of the intelligent distance estimation for the mobile Ad-hoc network that calculates the cluster with the Store-Compute-Forward method that adds computing ability to the existing Store-and-Forward routing scheme. The cluster technique of intelligent distance estimation for the mobile Ad-hoc network suggested in the study is the active and intelligent multi-hop cluster routing protocol to make secure communications.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.8
• /
• pp.203-211
• /
• 2013

Due to the rapid growth of mobile devices and applications, mobile cloud computing is becoming an important platform in the development of cloud services. However, the mobile cloud computing is facing many challenges in terms of the computing resources and communications. One of them is the performance issue between mobile devices and cloud server. In the paper, we implemented a hadoop-based android mobile application, called mobile TPC-C, and used it for evaluating the performance aspect between mobile devices and cloud server. The mobile TPC-C was implemented based on the existing TPC-C, to make it possible to execute on top of android mobile devices. The performance measurement using mobile TPC-C was executed on various transactions while changing the number of mobile clients. By comparing it to the evaluation on the personal PC, we tried to point out the important aspects affecting the performance improvement between mobile clients and cloud server.

• Journal of KIISE:Computer Systems and Theory
• /
• v.36 no.5
• /
• pp.351-359
• /
• 2009

Recently with increase of data in the Internet, platform technologies that can process huge data effectively such as Google platform and Hadoop are regarded as worthy of notice. In this kind of platform, there exist network I/O overheads to send task outputs due to the MapReduce operation which is a programming model to support parallel computation in the large cluster system. In this paper, we suggest applying of TIPC (Transparent Inter-Process Communication) protocol for reducing network I/O overheads and increasing network performance in the distributed computing environments. TIPC has a lightweight protocol stack and it spends relatively less CPU time than TCP because of its simple connection establishment and logical addressing. In this paper, we analyze main features of the Hadoop-based distributed computing system, and we build an experimental model which can be used for experiments to compare the performance of various protocols. In the experimental result, TIPC has a higher bandwidth and lower CPU overheads than other protocols.