• Journal of Internet Computing and Services
• /
• v.4 no.2
• /
• pp.39-45
• /
• 2003

Throughput response characteristics depending on the network bandwidth allocation is needed to be modeled to devise adaptive control mechanism to support QoS of the local network. In this study, we propose a dynamic system model that reveals the response characteristics of network. The adaptive traffic feedback control is applied to this model. And we simulate this system for optimization of adaptive control mechanism.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2001.10a
• /
• pp.446-449
• /
• 2001

A LLU system is developed for efficient use of existing local loop. By this system, new entrant ran use the local loop indifferently comparing with incumbent telecommunications operator. To implement the LLU, bottom-up typed LRIC model by network redesigning was accepted for costing system in Korea. In this paper, local loop redesigning methodology is presented to build a bottom-up typed LRIC model.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.13 no.5
• /
• pp.917-923
• /
• 2010

The local area weather information is very important element to estimate where the air-pollutant will flow. But the existing NBC hazard prediction model does not consider the local area weather information. So, in this paper, we present SN-HPM that uses the local area wether information to perform more accurate and reliable estimate, and embody it to program.

• The Journal of Korea Robotics Society
• /
• v.17 no.4
• /
• pp.425-430
• /
• 2022

As robots are considered one of the mainstream digital transformations, robots with machine vision becomes a main area of study providing the ability to check what robots watch and make decisions based on it. However, it is difficult to find a small object in the image mainly due to the flaw of the most of visual recognition networks. Because visual recognition networks are mostly convolution neural network which usually consider local features. So, we make a model considering not only local feature, but also global feature. In this paper, we propose a detection method of a small marker on the object using deep learning and an algorithm that considers global features by combining Transformer's self-attention technique with a convolutional neural network. We suggest a self-attention model with new definition of Query, Key and Value for model to learn global feature and simplified equation by getting rid of position vector and classification token which cause the model to be heavy and slow. Finally, we show that our model achieves higher mAP than state of the art model YOLOr.

• Journal of Distribution Science
• /
• v.14 no.11
• /
• pp.75-82
• /
• 2016

Purpose - In order for actively pursuing medium and long term policies of Gangwon region to be effectively and efficiently driven, efficacious and practical development strategies are needed. In terms of regional revitalization in most regions that are dependent on the primary industry like Gangwon-do Province, the maintaining of local community becomes difficult and there are limitations on the support from the central government and local governments. Therefore, local communities need to implement measures not only to be financially independent but also maintain and activate themselves. And community business can be adopted to be a proper strategy to cope with this change. This study drew importance of a community business model appropriate for Gangwon-do region to figure out success factors. Research design, data, and methodology - This study aimed to come up with importance of community business model for Gangwon-do region by using AHP Method. AHP Method, which was developed by Professor Saaty in 1970', is a methodology to simplify complex problems for a rational decision making. A survey targeting related public officials and expert group was carried out and a total of 30 questionnaires were collected for the analysis. Results - Analysis model used in this study was to prioritize community business models of Gangwon-do region. The second hierarchy was divided according to local restoration type, local resource utilization type, environment improvement type, and life support type. The third hierarchy consisted of 5 items such as network, the middle structure, program, government support, and human resources to measure each importance. As a result, in the second hierarchy, local resource utilization type had the highest importance. In the third hierarchy, the middle structure had the highest importance, followed by government support, program, network, and human resources. Collectively, the results suggested that important critical factors of community business model of Gangwon-do region was the importance of local resource utilization model and the middle structure. Conclusions - Not only should projects that are already operating in the region but next community business projects that are planning in the Gangwon-do region should be practically operated in view of the importance and the models derived from this study.

• Journal of Internet of Things and Convergence
• /
• v.10 no.2
• /
• pp.77-84
• /
• 2024

Over the past few years, IoT edges have begun to emerge based on new low-latency communication protocols such as 5G. However, IoT edges, despite their enormous advantages, pose new complementary threats, requiring new security solutions to address them. In this paper, we propose a cloud environment-based IoT edge architecture model that complements IoT systems. The proposed model acts on machine learning to prevent security threats in advance with network traffic data extracted from IoT edge devices. In addition, the proposed model ensures load and security in the access network (edge) by allocating some of the security data at the local node. The proposed model further reduces the load on the access network (edge) and secures the vulnerable part by allocating some functions of data processing and management to the local node among IoT edge environments. The proposed model virtualizes various IoT functions as a name service, and deploys hardware functions and sufficient computational resources to local nodes as needed.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2002.11a
• /
• pp.126-131
• /
• 2002

In constructing a superhigh communications network, investment in broadband of local loop is the most necessary. In this paper, it is performed to extract an optimum local loop by means of a comparative study after investigation and analysis of each construction structure of local loop. This paper is presented the way of construction of an economically optimum superhigh network by measurement in various circumstances, comparison and analysis of cost per node and star and ring in topology for constructing an optimum superhigh local loop.

• Smart Structures and Systems
• /
• v.23 no.4
• /
• pp.393-403
• /
• 2019

Applying more features gives us better accuracy in modeling; however, increasing the inputs causes the curse of dimensions. In this paper, a new structure has been proposed for fault detecting and identifying (FDI) of high-dimensional systems. This structure consist of two structure. The first part includes Auto-Encoders (AE) as Deep Neural Networks (DNNs) to produce feature engineering process and summarize the features. The second part consists of the Local Model Networks (LMNs) with LOcally LInear MOdel Tree (LOLIMOT) algorithm to model outputs (multiple models). The fault detection is based on these multiple models. Hence the residuals generated by comparing the system output and multiple models have been used to alarm the faults. To show the effectiveness of the proposed structure, it is tested on single-shaft industrial gas turbine prototype model. Finally, a brief comparison between the simulated results and several related works is presented and the well performance of the proposed structure has been illustrated.

• Wind and Structures
• /
• v.15 no.1
• /
• pp.43-64
• /
• 2012

Many potential small wind turbine locations are near obstacles such as buildings and shelterbelts, which can have a significant, detrimental effect on the local wind climate. A neural network-based model has been developed which predicts mean wind speed and turbulence intensity at points in an obstacle's region of influence, relative to unsheltered conditions. The neural network was trained using measurements collected in the wakes of 18 scale building models exposed to a simulated rural atmospheric boundary layer in a wind tunnel. The model obstacles covered a range of heights, widths, depths, and roof pitches typical of rural buildings. A field experiment was conducted using three unique full scale obstacles to validate model predictions and wind tunnel measurements. The accuracy of the neural network model varies with the quantity predicted and position in the obstacle wake. In general, predictions of mean velocity deficit in the far wake region are most accurate. The overall estimated mean uncertainties associated with model predictions of normalized mean wind speed and turbulence intensity are 4.9% and 12.8%, respectively.

• Journal of IKEEE
• /
• v.27 no.3
• /
• pp.313-318
• /
• 2023

This paper proposes a neural network applying fine-tuning as a way to improve the performance of Video Classification based on Vision Transformer. Recently, the need for real-time video image analysis based on deep learning has emerged. Due to the characteristics of the existing CNN model used in Image Classification, it is difficult to analyze the association of consecutive frames. We want to find and solve the optimal model by comparing and analyzing the Vision Transformer and Non-local neural network models with the Attention mechanism. In addition, we propose an optimal fine-tuning neural network model by applying various methods of fine-tuning as a transfer learning method. The experiment trained the model with the UCF101 dataset and then verified the performance of the model by applying a transfer learning method to the UTA-RLDD dataset.