• Journal of KIISE:Computer Systems and Theory
• /
• v.26 no.12
• /
• pp.1493-1503
• /
• 1999

ATM Clustering System과 같이 SAN(System Area Network) 환경에서 동작하는 시스템은 낮은 지연시간과 넓은 대역폭의 네트워크가 필수적이나 기존의 에러 복구 프로토콜들은 이러한 요구를 충족시키기에는 큰 오버헤드를 가지고 있다. 제안된 새로운 에러 복구 프로토콜은 ATM Clustering System 환경에서 최적의 성능을 나타내는 light-weight 프로토콜로 에러가 없는 상황과 에러 복구가 진행중인 상황에 따라 acknowledgement 주기를 적응적으로 변화시키는 adaptive acknowledgement scheme를 제안하여 적용하였다. 제안된 프로토콜은 상용 툴인 SDT를 이용한 논리 검증 받았고, DEVSim++ 환경에서의 성능 분석을 통해 프로토콜이 최상의 성능을 보이기 위한 파라메터 값을 찾았고, 이 값을 적용하였을 때의 성능을 기존의 프로토콜과 비교하여 제안된 프로토콜이 더 우수함을 확인하였다.Abstract While a system working with SAN, such as ATM Clustering System, requires a network with low latency and wide bandwidth, the previous error recovery protocols have a serious network overhead to satisfy this requirement. The suggested error recovery protocol is a light-weight protocol which can shows its best performance at ATM Clustering System and uses a newly suggested adaptive acknowledgement scheme. In the adaptive acknowledgement scheme, the period of acknowledgement is dynamically changed depending on the state of the network. We proved the logical correctness of our protocol with SDT and did performance analysis with DEVSim++. From the analysis, we found the optimal parameter values for best performance and showed that our protocol works better than the previous error recovery protocols.

• Journal of Internet Computing and Services
• /
• v.6 no.3
• /
• pp.85-95
• /
• 2005

Whereas the traditional multicast schemes based on Mobile IP can support a limited number of very large multicast groups, the Xcast protocol can support a very large number of small multicast groups, In the Xcast, the source node encodes the list of destinations in the Xcast header. Therefore, the maximum packet size in the network limits the number of destinations that a Xcast packet may have. In this paper, an enhanced Xcast protocol is proposed to solve the multicast group size limitation of the existing Xcast protocol in providing multicast service for mobile nodes. Moreover, the SIP (a very flexible control plane protocol) is integrated with the proposed multicast scheme to provide mobility awareness on the application layer. The simulation results verify that the proposed multicast scheme not only increases the packet delivery ratio and the data packet forwarding efficiency but also achieves low latency of packets in the network.

• Journal of Advanced Navigation Technology
• /
• v.18 no.5
• /
• pp.437-444
• /
• 2014

The integrated modular avionics (IMA) system has quite a number of line repalceable moduels (LRMs) in a cabinet. The LRM performs functions like line replaceable units (LRUs) in federated architecture. The video processing module (VPM) acts as a video bus bridge and gateway of ARINC 818 avionics digital video bus (ADVB). The VPM is a LRM in IMA core system. The ARINC 818 video interface and protocol standard was developed for high-bandwidth, low-latency and uncompressed digital video transmission. FPGAs of the VPM include video processing function such as ARINC 818 to DVI, DVI to ARINC 818 convertor, video decoder and overlay. In this paper we explain how to implement VPM's Hardware. Also we show the verification results about VPM functions and IP core performance.

• KIISE Transactions on Computing Practices
• /
• v.23 no.7
• /
• pp.417-423
• /
• 2017

Recently, many embedded devices that have the computing capability required for deep learning have become available; hence, many new applications using these devices are emerging. However, these embedded devices have an architecture different from that of PCs and high-performance servers. In this paper, we propose a method that improves the performance of deep-learning framework by considering the architecture of an embedded device that shares memory between the CPU and the GPU. The proposed method is implemented in Caffe, an open-source deep-learning framework, and is evaluated on an NVIDIA Jetson TK1 embedded device. In the experiment, we investigate the image recognition performance of several state-of-the-art deep-learning networks, including AlexNet, VGGNet, and GoogLeNet. Our results show that the proposed method can achieve significant performance gain. For instance, in AlexNet, we could reduce image recognition latency by about 33% and energy consumption by about 50%.

• Proceedings of the Korean Society for Cognitive Science Conference
• /
• 2005.05a
• /
• pp.151-153
• /
• 2005

We investigated the brain substrate of repetition priming on the implicit memory taskusing low-resolution electromagnetic tomography (LORETA) with high-density 128 channel EEG and individual MRI as a realistic head model. Thirteen right-handed, healthy subjects performed a word/nonword discrimination task, in which the words and nonwords were presented visually,and some of the words appeared twice with a lag of one or five items. All of the subjects exhibited repetition priming with respect to the behavioral data, in which a faster reaction time was observed to the repeated word (old word) than to the first presentation of the word (new word). The old words elicited more positive-going potentials than the new words, beginning at 200 ms and lasting until 500 ms post-stimulus. We conducted source reconstruction using LORETA at a latency of 400 ms with the peak mean global field potentials and used statistical parametric mapping for the statistical analysis. We found that the source elicited by the old words exhibited a statistically significant current density reduction in the left inferior frontal gyrus. This is the first study to investigate the generators of repetition priming using voxel-by-voxel statistical mapping of the current density with individual MRI and high-density EEG.

• The Journal of the Korea Contents Association
• /
• v.12 no.4
• /
• pp.31-39
• /
• 2012

DVC(Distributed Video Coding) techniques have been attracting a lot of research works since these enable us to implement the light-weight video encoder and to provide good coding efficiency by introducing the feedback channel. However, the feedback channel causes the decoder to increase the decoding complexity and requires very high decoding latency because of numerous iterative decoding processes. So, in order to reduce the decoding delay and then to implement in a real-time environment, this paper proposes several parity bit estimation methods which are based on the temporal correlation, spatial correlation and spatio-temporal correlations between LDPCA frames on each bit plane in the consecutive video frames in pixel-domain Wyner-Ziv video coding scheme and then the performances of these methods are compared in fast DVC scheme. Through computer simulations, it is shown that the adaptive spatio-temporal correlation-based estimation method and the temporal correlation-based estimation method outperform others for the video frames with the highly active contents and the low active contents, respectively. By using these results, the proposed estimation schemes will be able to be effectively used in a variety of different applications.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.12
• /
• pp.913-917
• /
• 2009

Flash memories have asymmetric I/O costs for read and write in terms of latency and energy consumption. However, the ratio of these costs is dependent on the type of storage. Moreover, it is becoming more common to use two flash memories on a system as an internal memory and an external memory card. For this reason, buffer cache replacement algorithms should consider I/O costs of device as well as possibility of reference. This paper presents WWCLOCK(Write-Weighted CLOCK) algorithm which directly uses I/O costs of devices along with recency and frequency of cache blocks to selecting a victim to evict from the buffer cache. WWCLOCK can be used for wide range of storage devices with different I/O cost and for systems that are using two or more memory devices at the same time. In addition to this, it has low time and space complexity comparable to CLOCK algorithm. Trace-driven simulations show that the proposed algorithm reduces the total I/O time compared with LRU by 36.2% on average.

• Journal of Korean Neurosurgical Society
• /
• v.55 no.6
• /
• pp.370-374
• /
• 2014

To present a case of cauda equina syndrome (CES) caused by chronic inflammatory demyelinating polyneuropathy (CIDP) which seemed clinically similar to Charcot-Marie-Tooth disease type1 (CMT1). CIDP is an immune-mediated polyneuropathy, either progressive or relapsing-remitting. It is a non-hereditary disorder characterized by symmetrical motor and sensory deficits. Rarely, spinal nerve roots can be involved, leading to CES by hypertrophic cauda equina. A 34-year-old man presented with low back pain, radicular pain, bilateral lower-extremity weakness, urinary incontinence, and constipation. He had had musculoskeletal deformities, such as hammertoes and pes cavus, since age 10. Lumbar spine magnetic resonance imaging showed diffuse thickening of the cauda equina. Electrophysiological testing showed increased distal latency, conduction blocks, temporal dispersion, and severe nerve conduction velocity slowing (3 m/s). We were not able to find genetic mutations at the PMP 22, MPZ, PRX, and EGR2 genes. The pathologic findings of the sural nerve biopsy revealed thinly myelinated nerve fibers with Schwann cells proliferation. We performed a decompressive laminectomy, intravenous IgG (IV-IgG) and oral steroid. At 1 week after surgery, most of his symptoms showed marked improvements except foot deformities. There was no relapse or aggravation of disease for 3 years. We diagnosed the case as an early-onset CIDP with cauda equine syndrome, whose initial clinical findings were similar to those of CMT1, and successfully managed with decompressive laminectomy, IV-IgG and oral steroid.

• Journal of Satellite, Information and Communications
• /
• v.11 no.3
• /
• pp.37-42
• /
• 2016

In this paper, in today's highly complex video and broadcast operations, broadcasters are constantly challenged to reliably deliver low-latency, high-quality video to multiscreen audiences on-air and online. The Adaptive Bit Rate (ABR) protocols enable internet video to a wide range of multiscreen devices. However, video quality is often marginal and would prove unacceptable for valued linear broadcast content delivered to the Big Screen today. The Media information processing technology advances in ABR enables service providers to take control and offer quality managed linear video services to ALL screens in the home, including the Big Screen, with a single unified IP Video infrastructure. The New Multiscreen-Assisted ABR (MSA-ABR) delivery management system proposed using Cloud based multicast-assisted ABR for a broadcast facility that performs routing of contribution content and online publishing services within a virtual, centralized cloud infrastructure.

• Journal of Advanced Navigation Technology
• /
• v.15 no.1
• /
• pp.54-60
• /
• 2011

The development of wireless sensor networks (WSN) can be motivated by several types of applications. However, these applications demand an energy-efficient WSN that can prolong the network lifetime and can provide high throughput, low latency and delay. Designing wireless sensor networks with the capability of prolonging network lifetime catch the attention of many researchers in wireless system and network field. Contrasts with Mobile Ad Hoc Network system, Wireless Sensor Networks designs focused more on survivability of each node in the network instead of maximizing data throughput or minimizing end-to-end delay. In this paper, we will study part of data link layer in Open Systems Interconnection (OSI) model, called medium access control (MAC) layer. Since the MAC development of energy aware MAC Protocol for wireless sensor layer controls the physical radio part, it has a large impact on the overall energy consumption and the lifetime of a node. This paper proposes a analytical approach that tries to reduce idle energy consumption, and shows the increasement of network end-to-end arrival rate due to efficiency in energy consumption from time slot management.