• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.1
• /
• pp.35-57
• /
• 2014

Multi-criteria decision making (MCDM) algorithms play an important role in ensuring quality of service in an integrated HetNets (Heterogeneous Networks). The primary objective of this paper is to develop a multi-criteria vertical handoff decision algorithm (VHDA) for best access network selection in an integrated Wireless Local Area Network (WLAN)/Universal Mobile Telecommunications System (UMTS)/Worldwide Interoperability for Microwave Access (WiMAX) system. The proposed design consists of two parts, the first part is the evaluation of an Analytic Hierarchy Process (AHP) to decide the relative weights of handoff decision criteria and the second part computes the final score of the weights to rank network alternatives using Simple Additive Weighting (SAW). SAW ranks the network alternatives in a faster and simpler manner than AHP. The AHP-SAW mathematical model has been designed, evaluated and simulated for streaming video type of traffic. For other traffic type, such as conversational, background and interactive, only simulation results have been discussed and presented in brief. Simulation results reveal that the hierarchical modelling and computing provides optimum solution for access network selection in an integrated environment as obtained results prove to be an acceptable solution to what could be expected in real life scenarios.

• Electronics and Telecommunications Trends
• /
• v.35 no.5
• /
• pp.30-42
• /
• 2020

Unlike in previous generations of mobile technology, machine learning (ML)-based self-healing research trend are currently attracting attention to provide high-quality, effective, and low-cost 5G services that need to operate in the HetNets scenario where various wireless transmission technologies are added. Self-healing plays a vital role in detecting and mitigating the faults, and confirming that there is still room for improvement. We analyzed the research trend in self-healing framework and ML-based fault detection, fault diagnosis, and fault compensation. We propose that to ensure that self-healing is a proactive instead of being reactive, we have to design an ML-based self-healing framework and select a suitable ML algorithm for fault detection, diagnosis, and outage compensation.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.43 no.4
• /
• pp.511-523
• /
• 2023

Research on the integration of unmanned aerial vehicles and deep learning for reinforced concrete damage detection is actively underway. Convolutional neural networks have a high impact on the performance of image classification, detection, and segmentation as backbones. The MobileNet, a pre-trained convolutional neural network, is efficient as a backbone for an unmanned aerial vehicle-based damage detection model because it can achieve sufficient accuracy with low computational complexity. Analyzing vanilla convolutional neural networks and MobileNet under various conditions, MobileNet was evaluated to have a verification accuracy 6.0~9.0% higher than vanilla convolutional neural networks with 15.9~22.9% lower computational complexity. MobileNetV2, MobileNetV3Large and MobileNetV3Small showed almost identical maximum verification accuracy, and the optimal conditions for MobileNet's reinforced concrete damage image feature extraction were analyzed to be the optimizer RMSprop, no dropout, and average pooling. The maximum validation accuracy of 75.49% for 7 types of damage detection based on MobilenetV2 derived in this study can be improved by image accumulation and continuous learning.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.1
• /
• pp.65-73
• /
• 2006

Network based VPN(Virtual Private Network) using MPLS(Multiprotocol Label Switching) technology, called PE(Provider Edge router) based MPLS VPN, is regarded as a good solution for intranets or ext3nets because of the low cost and the flexibility of the service provision. In this paper, we describe a mechanism that allows the VPN users to move from one site to another site of the VPN network based on the BGP-E MPLS technology. This mechanism is designed for PE(Provider Edge) routers of the backbone network. PE routers connected to the VPN sites establish a new MPLS path to the mobile node after they detect movement of the mobile VPN node. The new location may belong to the same VPN or to different VPN. We desisted VPN management and control functions of the PE routers in order to interface with the Mobile IP protocol and support the QoS mechanism. The pilot implementation and performance measurement were carried out on a simulation using COVERS tool.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.10a
• /
• pp.355-358
• /
• 2007

In wireless mobile communication systems, priority of voice service through high speed data and multimedia transmission requires increased service diversification. Research is being carried out in this environment, on the call admission control techniques to guarantee the diversified service's QoS. SRN (Stochastic Reward Net) is an extended version of Petri nets, well know modeling and analysis tool. In this paper, we develop SRN call admission control model considering the 4 classes of services in the 4th generation IEE 802.16e mobile communication Technology.

• Journal of IKEEE
• /
• v.23 no.2
• /
• pp.448-453
• /
• 2019

Under the conflicting objectives of maximum user satisfaction and fast launching, there exist great needs for automated mobile app testing. In automated app testing, detection of usage-state changes is one of the most important issues for minimizing human intervention and testing of various usage scenarios. Because conventional approaches utilizing pre-collected training examples can not handle the rapid evolution of apps, we propose a novel method detecting changes in usage-state through graph-entropy. In the proposed method, widgets in a screen shot are recognized through DNNs and 'onverted graphs. We compared the performance of the proposed method with a SIFT (Scale-Invariant Feature Transform) based method on 20 real-world apps. In most cases, our method achieved superior results, but we found some situations where further improvements are required.

• Journal of the Korean Society of Safety
• /
• v.30 no.5
• /
• pp.44-51
• /
• 2015

As the domestic communication industry has been made rapid progresses, the construction of base stations are also expanded yearly from more than 86,000 ones in the year 2005 to over 300,000 ones in 2013. While the new construction and rebuilt works have been rapidly increased, there has been not tried about the analysis as well as the control of hazards for the construction works. This study is conducted to do not only the focus interview for 27 expert groups who are engaged in the construction works and safety, but also the questionnaires for AHP analysis which is suggested by the Korea Occupational Safety and Health Agency for 92 persons who are engaged in the construction works in order to analyze the hazards of construction works for base stations. The high ranked risk works which are acquired from the two methods are fall under installation of antennae, struck by the hand tools and construction materials under installation of antennae, fall during going up and down the steel tower and telegraph poles and the fall by electric shock during the underground cable works. The control for the above hazards are the installation of fall protection equipments, the installation of safety nets for falling objects and the installation of protection equipment of electric shock.

• Journal of IKEEE
• /
• v.22 no.3
• /
• pp.693-699
• /
• 2018

CNN based object recognitions have reported splendid results. However, the recognition of mobile apps raises an interesting challenge that recognition performance of similar widgets is not consistent. In order to improve the performance, we propose a noble method utilizing relations between input widgets. The recognition process flows from the Faster R-CNN based recognition to enhancement using a relation recognizer. The relations are represented as vector translation between objects in a relation space. Experiments on 323 apps show that our method significantly enhances the Faster R-CNN only approach.

• 한국사회정책
• /
• v.25 no.4
• /
• pp.199-231
• /
• 2018

The development of ICT has led to changes in the pattern and the meaning of work and requires restructuring of the existing social security system, which was established chiefly for the wage workers in the industrial economy. However, while economic activities within the digital platform markets are different from industrial labour, there is still a lack of discussion on what social problems platform workers can face and how to cope with them. As a comprehensive sketch of economic activities in platform economy, this study identifies the types of platform workers and analyses three characteristics of their economic activities - flexibility, virtuality, and connectivity. It then examines what social risks can be derived form the three characteristics. This research lastly suggests alternative social safety nets and policies to alleviates the social risks and problems that platform workers may face in the digital society.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.3
• /
• pp.277-288
• /
• 2013

3GPP LTE-Advanced (Third Generation Partnership Project Long Term Evolution-Advanced) as a next generation mobile communication standard introduced small base stations such as femto cells or pico cells, and D2D (Device-to-Device) communications between mobiles in the proximity in order to satisfy the needs of rapidly growing wireless data traffic. A diverse range of topics has been studied to solve various interference situations which may occur within a single cell. In particular, an introduction of a small base station along with D2D communication raises important issues of how to increase the channel capacity and frequency efficiency in HetNets (Heterogeneous Networks). To this end, we propose in this paper methods to manage the interference between the macro cell and other small cells in the HetNet to improve the frequency efficiency. The proposed CCN (Cluster Coordinator Node)-assisted ICI (Inter-Cell Interference) avoidance methods exploit the CCN to control the interference in HetNet comprising of an MeNB (Macro enhanced Node-B) and a large number of small cells. A CCN which is located at the center of a number of small cells serves to avoid the interference between macro cell and small cells. We propose methods of resource allocation to avoid ICI for user equipments within the CCN coverage, and evaluate their performance through system-level computer simulations.