• 한국컴퓨터산업학회논문지
• /
• 제9권2호
• /
• pp.63-68
• /
• 2008

병렬 스케줄링의 목적은 다중프로세서 병렬시스템 환경에서 병렬성을 가진 응용 프로그램에 대하여 최소 동기화 오버헤드(Synchronization overhead) 및 병목현상(Bottleneck) 그리고 부하균등(Load balance)을 달성하도록 스케줄링을 수행하는데 있다. 본 논문에서는 기존의 대표적인 병렬수행 스케줄링 알고리즘들을 분석하고 각 방법들로부터 문제점들을 도출한다. 이는 향후 효율적인 알고리즘을 설계하는데 도움이 필 것이다.

• Journal of Information Processing Systems
• /
• 제17권5호
• /
• pp.892-904
• /
• 2021

An image semantic segmentation model is proposed based on improved ENet network in order to achieve the low accuracy of image semantic segmentation in complex environment. Firstly, this paper performs pruning and convolution optimization operations on the ENet network. That is, the network structure is reasonably adjusted for better results in image segmentation by reducing the convolution operation in the decoder and proposing the bottleneck convolution structure. Squeeze-and-excitation (SE) module is then integrated into the optimized ENet network. Small-scale targets see improvement in segmentation accuracy via automatic learning of the importance of each feature channel. Finally, the experiment was verified on the public dataset. This method outperforms the existing comparison methods in mean pixel accuracy (MPA) and mean intersection over union (MIOU) values. And in a short running time, the accuracy of the segmentation and the efficiency of the operation are guaranteed.

• 한국멀티미디어학회논문지
• /
• 제25권1호
• /
• pp.72-81
• /
• 2022

Visual Object Tracking is known as the most fundamental problem in the field of computer vision. Object tracking localize the region of target object with bounding box in the video. In this paper, a custom CNN is created to extract object feature that has strong and various information. This network was constructed as a Siamese network for use as a feature extractor. The input images are passed convolution block composed of a bottleneck layers, and features are emphasized. The feature map of the target object and the search area, extracted from the Siamese network, was input as a local proposal network. Estimate the object area using the feature map. The performance of the tracking algorithm was evaluated using the OTB2013 dataset. Success Plot and Precision Plot were used as evaluation matrix. As a result of the experiment, 0.611 in Success Plot and 0.831 in Precision Plot were achieved.

• Current Optics and Photonics
• /
• 제5권6호
• /
• pp.617-626
• /
• 2021

A parallel Monte Carlo photon migration algorithm for graphics processing units that implements an improved load-balancing strategy is presented. Conventional parallel Monte Carlo photon migration algorithms suffer from a computational bottleneck due to their reliance on a simple load-balancing strategy that does not take into account the different length of the mean free paths of the photons. In this paper, path-partition load balancing is proposed to eliminate this computational bottleneck based on a mathematical formula that parallelizes the photon path tracing process, which has previously been considered non-parallelizable. The performance of the proposed algorithm is tested using three-dimensional photon migration simulations of a human skin model.

• International Journal of Computer Science & Network Security
• /
• 제22권3호
• /
• pp.292-302
• /
• 2022

The radar tomographic imaging is based on the Radar Cross-Section "RCS" of the materials of a shape under examination and investigation. The RCS varies as the conductivity and permittivity of a target, where the target has a different material profile than other background objects in a scene. In this research paper, we use Hierarchical Performance Modeling "HPM" and a framework developed earlier to determine/spot bottleneck(s) for pattern recognition of materials using a combination of the Single Layer Perceptron (SLP) technique and tomographic images in radar systems. HPM provides mathematical equations which create Objective Functions "OFs" to find an average performance metric such as throughput or response time. Herein, response time is used as the performance metric and during the estimation of it, bottlenecks are found with the help of OFs. The obtained results indicate that processing images consumes around 90% of the execution time.

• 대한전자공학회논문지TC
• /
• 제47권7호
• /
• pp.51-58
• /
• 2010

무선 메쉬 네트워크 (Wireless Mesh Network)는 넓은 지역에 저렴한 비용으로 무선 인터넷 서비스를 제공할 수 있는 기술로 많은 관심을 받고 있다. 하지만 무선 메쉬 네트워크는 게이트웨이 주변 노드로 트래픽이 집중되는 병목현상이 발생하는 문제가 있다. 이러한 병목현상을 해결하기 위해서는 멀티 라디오를 사용하는 무선 메쉬 네트워크 구성이 필요하지만, 무분별한 안테나 설치는 전체 시스템 구축비용의 증가를 초래하게 된다. 이에 본 논문에서는 네트워크 트래픽 용량에 따라 최적화된 메쉬 클러스터 규모를 결정하고, 각 메쉬 라우터의 무선 채널 사용량에 따라 안테나 수의 차등적 배치를 통해 병목 현상 해소 및 시스템 구축비용의 최소화가 가능한 시스템 비용 최적화 모델을 제안한다. 제안 최적화 모델의 성능 분석을 위해 CPLEX를 사용해 사용자 트래픽과 게이트웨이 용량에 따른 네트워크 규모 및 최적의 안테나 수를 결정하는 최적해를 도출하였다. 그 결과 제안된 최적화 모델을 통해 멀티 라디오 무선 메쉬 네트워크 구성 시 병목 현상 해결 및 시스템 구축비용의 현저한 감소를 확인할 수 있었다.

• 실천공학교육논문지
• /
• 제6권1호
• /
• pp.51-55
• /
• 2014

산업체 수요에 기반한 공학혁신형 산업의료원 교과목과 운영 사례에 대해서 소개한다. 산업의료원 교과목은 산업체에서 제공한 애로기술 문제를 교수와 학생이 산업의료원 교과목 수업을 통해서 해결한다. 산업체는 애로기술을 제시하고 담당교수는 제시된 애로기술을 해결하는 강의 계획을 수립해서 학생들과 함께 산업체 애로기술을 해결하는 과정을 수행하고 학생들은 그것을 통해서 현장적응력을 키웠다. 학생들은 교과목 운영 목적 인식 정도, 수행기간의 적절성, 커뮤니케이션의 원활성, 전공지식의 이해 및 활용에 기여, 문제해결능력, 협업 학습 능력, 진로선택의 기회, 교과목 추천 등에 높은 평가를 하였다. 산업의료원은 한 학기 과정보다는 두 학기로 운영한 경우가 학생들의 만족도가 높게 평가된다. 그 이유는 산업의료원에서 제시하는 애로기술 문제가 한 학기에 해결책을 찾기 어렵기 때문이다. 산업의료원 교과목을 통해서 산업체에 취업연계를 가질 수 있는 기회가 될 수 있다.

• 한국정보과학회논문지:시스템및이론
• /
• 제26권9호
• /
• pp.1042-1054
• /
• 1999

공유메모리 병렬프로그램의 오류수정에서 경합의 탐지는 중요하다. 왜냐하면 경합은 잘못된 수행 결과를 초래할 뿐만 아니라, 의도하지 않은 프로그램의 비결정적인 수행을 유발하여 오류수정을 어렵게 하기 때문이다. 특히 최초경합의 탐지는 더욱 중요하다. 그 이유는 최초경합을 제거함으로써 나머지 경합들을 방지할 수도 있기 때문이다. 기존의 수행중 경합 탐지기법들은 접근별 보고방식을 기반으로 하는데, 이 기법들은 임의 공유변수에 대한 병행 쓰레드들의 모든 접근사건들을 검사하기 위해서 접근역사라는 유일한 공유정보를 이용하므로 탐지과정에 심각한 병목현상을 유발시킨다. 그러나, 최초경합 탐지를 위한 경우 이러한 병목현상은 크게 개선될 수 있다. 본 논문에서는, 각 접근사건 검사를 위해 각 쓰레드에 공유되지 않는 독립적인 접근역사를 별개로 두고, 경합을 보고하는 시점인 쓰레드 합류시점에서만 공유되는 접근역사를 이용하도록 함으로써 병목현상을 개선하여 최초경합을 탐지할 수 있는 새로운 수행중 탐지기법을 제안한다. 그러므로 본 기법은 최초경합을 보다 효율적으로 탐지할 수 있기 때문에 수행중 경합 탐지를 더욱 효율적이고 실용적으로 할 수 있다. Abstract Detecting races is important for debugging shared-memory parallel programs, because the races lead to unintended nondeterministic executions of the programs as well as erroneous result and then make debugging programs difficult. Especially, detecting the first races is more important. The reason is that the removal of the first races can make other races disappear. Most existing on-the-fly techniques to detect the races are based on per- access reporting method incurring the serious central bottleneck, because the techniques use unique shared information called access history for checking all accesses of concurrent threads to a shared variable. Such bottleneck, however, can be improved considerably in case of detecting first races. This paper presents a new on-the-fly technique which detects the first races with reduced bottleneck through checking each accesses with private access histories and finally reporting races with shared access histories. Therefore, this technique makes on-the-fly race detection more efficient and practical.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권4호
• /
• pp.1406-1423
• /
• 2014

Multi-mode device which combines multiple access technologies into a device will offer more cost-effective solution than a sole access implementation. Its concurrent multipath transfer (CMT) technology can transmit media flows over multiple end-to-end paths simultaneously, which is essential to select at least two paths from all available paths. At real networks, different paths are likely to overlap each other and even share bottleneck, which can weaken the path diversity gained through CMT. Spurred by this observation, it is necessary to select multiple independent paths as much as possible to avoid underlying shared bottleneck between topologically joint paths. Recent research in this context has shown that different paths with shared bottleneck can weaken the path diversity gained through CMT. In our earlier work, a grouping-based multipath selection (GMS) mechanism is introduced and developed. However, how to estimating the selection is still to be resolved. In this paper, we firstly introduce a Selection Correctness Index (SCI) to evaluate the correctness of selection results in actual CMT experiment. Therefore, this metric is helpful to discuss and validate the accuracy of the output paths. From extensive experiments with a realized prototype, the proposed scheme provides better evaluation tool and criterion in various network conditions.

• Journal of Communications and Networks
• /
• 제12권1호
• /
• pp.52-66
• /
• 2010

Since large objects consume substantial resources, web proxy caching incurs a fundamental trade-off between performance (i.e., hit-ratio and latency) and overhead (i.e., resource usage), in terms of caching and relaying large objects to users. This paper investigates how and to what extent the current dedicated-server based web proxy caching scheme is affected by large file transfers in a high bandwidth campus network environment. We use a series of trace-based performance analyses and profiling of various resource components in our experimental squid proxy cache server. Large file transfers often overwhelm our cache server. This causes a bottleneck in a web network, by saturating the network bandwidth of the cache server. Due to the requests for large objects, response times required for delivery of concurrently requested small objects increase, by a factor as high as a few million, in the worst cases. We argue that this cache bandwidth bottleneck problem is due to the fundamental limitations of the current centralized web proxy caching model that scales poorly when there are a limited amount of dedicated resources. This is a serious threat to the viability of the current web proxy caching model, particularly in a high bandwidth access network, since it leads to sporadic disconnections of the downstream access network from the global web network. We propose a peer-to-peer cooperative web caching scheme to address the cache bandwidth bottleneck problem. We show that it performs the task of caching and delivery of large objects in an efficient and cost-effective manner, without generating significant overheads for participating peers.