• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.5 no.4
• /
• pp.626-648
• /
• 2011

Energy consumption of wireless access networks is in permanent increase, which necessitates development of more energy-efficient network management approaches. Such management schemes must result with adaptation of network energy consumption in accordance with daily variations in user activity. In this paper, we consider possible energy savings of wireless local area networks (WLANs) through development of a few integer linear programming (ILP) models. Effectiveness of ILP models providing energy-efficient management of network resources have been tested on several WLAN instances of different sizes. To cope with the problem of high computational time characteristic for some ILP models, we further develop several heuristic algorithms that are based on greedy methods and local search. Although heuristics obtains somewhat higher results of energy consumption in comparison with the ones of corresponding ILP models, heuristic algorithms ensures minimization of network energy consumption in an amount of time that is acceptable for practical implementations. This confirms that network management algorithms will play a significant role in practical realization of future energy-efficient network management systems.

• Journal of Information Processing Systems
• /
• v.18 no.1
• /
• pp.12-25
• /
• 2022

In recent years, deep convolutional neural networks have made significant progress in the research of single image super-resolution. However, it is difficult to be applied in practical computing terminals or embedded devices due to a large number of parameters and computational effort. To balance these problems, we propose CSRNet, a lightweight neural network based on channel split residual learning structure, to reconstruct highresolution images from low-resolution images. Lightweight refers to designing a neural network with fewer parameters and a simplified structure for lower memory consumption and faster inference speed. At the same time, it is ensured that the performance of recovering high-resolution images is not degraded. In CSRNet, we reduce the parameters and computation by channel split residual learning. Simultaneously, we propose a double-upsampling network structure to improve the performance of the lightweight super-resolution network and make it easy to train. Finally, we propose a new evaluation metric for the lightweight approaches named 100_FPS. Experiments show that our proposed CSRNet not only speeds up the inference of the neural network and reduces memory consumption, but also performs well on single image super-resolution.

• Journal of Korean Society of Transportation
• /
• v.11 no.1
• /
• pp.87-103
• /
• 1993

The subject this paper is the signal control strategy under oversaturated conditions. The nature of traffic control for oversaturation is essentially different from the standard control modes. While under non-saturated situation traffic control is needed for the sake of safety and efficiency, the throughput is essential under oversaturated conditions. Therefore berth objective and strategies differ. For an oversaturated stream the cycle time and the signal offset are thought to be of rather secondary importance. For this case the green split may well be the most important control variable to serve the excessive demand. Up to now, however, most efforts have concentrated on the strategy with the concept which lies just on the extension of Webster's. "Green-split Coordination Strategy for Over-Saturated Networks", presents newly contrived three types of strategies named Forward-coordination, Backward-coordination and Network-coordination respectively and describes the algorithms with the evaluations. The forward coordination strategy treats the forward wave of flow between two signals. The aim is to prevent the outbreak of queue due to the accumulation of temporary excess of demand in near-saturation or saturation flow. The backward coordination strategy treats the backward rave of flow between two signals. The goal is to prevent the waste of green time caused by the exit block at the upstream signal. for this purpose a feedback regulation is provided of the upstream green-split so that the inflow-outflow balance is kept zero. The resultant surplus of green time is alloted to other signal stages. Also here the examination is made of the appropriate value of the feedback control parameter. The network coordination strategy is operated to maximize the network throughput in a specific direction applying a bang-bang control at the bottleneck intersection. This is a type of intervenient control for policy reasons. For this strategy the green-split coordinations, particuarly the backward coordination, are essential as the tactical elements. In order to evaluate the preposed strategies those are compared with the latest existing strategy called saturation-degree-ratio control by the simulation experiments in an assumed 4$\times$4 grid network. The results are satisfactory showing a 10-15% reduction in delays and a 15% increase in network capacity.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.5
• /
• pp.178-188
• /
• 2020

This paper proposes a lane recognition CNN network using split-attention network as a backbone to extract feature. Split-attention is a method of assigning weight to each channel of a feature map in the CNN feature extraction process; it can reliably extract the features of an image during the rapidly changing driving environment of a vehicle. The proposed deep neural networks in this paper were trained and evaluated using the Tusimple data set. The change in performance according to the number of layers of the backbone network was compared and analyzed. A result comparable to the latest research was obtained with an accuracy of up to 96.26, and FN showed the best result. Therefore, even in the driving environment of an actual vehicle, stable lane recognition is possible without misrecognition using the model proposed in this study.

• The KIPS Transactions:PartA
• /
• v.8A no.4
• /
• pp.359-366
• /
• 2001

This paper overviews the design of SANique$^{TM}$ -a shred file system for Linux cluster based on SAN environment. SANique$^{TM}$ has the capability of transferring user data from network-attached SAN disks to client applcations directly without the control of centralized file server system. The paper also presents the characteristics of each SANique$^{TM}$ subsystem: CFM(Cluster File Manager), CVM(Cluster Volume Manager), CLM(Cluster Lock Manager), CBM(Cluster Buffer Manager) and CRM(Cluster Recovery Manager). Under the SANique$^{TM}$ design layout, then, the syndrome of '||'&'||'quot;split-brain'||'&'||'quot; in shared file system environments is described and defined. The work first generalizes and illustrates possible situations in each of which a shared file system environment may split into two or more pieces of separate brain. Finally, the work describes the SANique$^{TM}$ approach to the given "split-brain"problem using SAN disk named "split-brain" and develops the overall recovery procedure of shared file systems.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.54 no.5
• /
• pp.798-807
• /
• 2021

The directivity characteristics of acoustic transducers for conventional single-beam echo sounders considerably limit the detection of fish-size information in acoustic field surveys. To overcome this limitation, using the split-aperture technique to estimate the direction of arrival of single-echo signals from individual fish distributed within the sound beam represents the most reliable method for fish-size classification. For this purpose, we design and develop a split-beam data acquisition and processing system to obtain fish-size information in conjunction with a 50-kHz single-beam echo sounder. This split-beam data acquisition and processing system consists of a notebook PC, a field-programmable gate array board, an external single-transmitter module with a matching network, and four-channel receiver modules operating at a frequency of 50-kHz. The functionality of the developed split-beam data processor is tested and evaluated. Acoustic measurements in an experimental water tank showed that the developed data acquisition and processing system can be used as a fish-sizing echo sounder to estimate the size distribution of individual fish, although an external single-transmitter module with a matching network is required.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.111-113
• /
• 2000

We discuss the problem of resource reservation, such as bandwidth, for delay sensitive application in wireless networking environment. As multimedia application is becoming a critical role of the current mobile network, the resource reservation become very important to support the real-time service. In this paper, we propose and evaluate a new resource reservation protocol, called Split Reservation Protocol, in wireless network. The simulation results show that our proposed protocol outperforms an existing protocol called MPSVP in terms of network overhead.

• Information Systems Review
• /
• v.2 no.1
• /
• pp.151-160
• /
• 2000

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.4A
• /
• pp.388-398
• /
• 2011

In this paper, we introduce an EMSP(Efficient Mobility Support Protocol) for mobile sensor network with mobility-aware. We propose virtual cluster and node split scheme considering movements of mobile nodes. The existing M-LEACH protocol suffers from communication cost spent on JOIN request information during invitation phase. To address this issue, the large boundary of the cluster in LUR-tree can reduce superfluous update cost. In addition to the expansion of the cluster, the proposed approach exploits node split algorithms used in R-tree in order to uniformly form a cluster. The simulated results show that energy-consumption has less up to about 40% than LEACH-C and 8% than M-LEACH protocol. Finally, we show that the proposed scheme outperforms those of other in terms of lifetime of sensor fields and scalability in wireless sensor network.

• ETRI Journal
• /
• v.34 no.3
• /
• pp.352-360
• /
• 2012

This paper proposes a low-cost, low-power, and high-capacity optical-electrical-optical-type reach extender that can provide 3R frame regeneration and remote management to increase the reach and split ratio with no change to a legacy time division multiple access passive optical network. To provide remote management, the extender gathers information regarding optical transceivers and link status per port and then transmits to a service provider using a simple network management protocol agent. The extender can also apply to an Ethernet passive optical network (E-PON) or a gigabit-capable PON (G-PON) by remote control. In a G-PON, in particular, it can provide burst mode signal retiming and burst-to-continuous mode conversion at the upstream path through a G-PON transmission convergence frame adaptor. Our proposed reach extender is based on the quad-port architecture for cost-effective design and can accommodate both the physical reach of 60 km and the 512 split ratios in a G-PON and the physical reach of 80 km and the 256 split ratios in an E-PON.