• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.1
• /
• pp.114-123
• /
• 2018

This paper concerns a design and learning method of softmax function neural networks based on K-means clustering. The partial discharge data Information is preliminarily processed through simulation using an Epoxy Mica Coupling sensor and an internal Phase Resolved Partial Discharge Analysis algorithm. The obtained information is processed according to the characteristics of the pattern using a Motor Insulation Monitoring System program. At this time, the processed data are total 4 types that void discharge, corona discharge, surface discharge and slot discharge. The partial discharge data with high dimensional input variables are secondarily processed by principal component analysis method and reduced with keeping the characteristics of pattern as low dimensional input variables. And therefore, the pattern classifier processing speed exhibits improved effects. In addition, in the process of extracting the partial discharge data through the MIMS program, the magnitude of amplitude is divided into the maximum value and the average value, and two pattern characteristics are set and compared and analyzed. In the first half of the proposed partial discharge pattern classifier, the input and hidden layers are classified by using the K-means clustering method and the output of the hidden layer is obtained. In the latter part, the cross entropy error function is used for parameter learning between the hidden layer and the output layer. The final output layer is output as a normalized probability value between 0 and 1 using the softmax function. The advantage of using the softmax function is that it allows access and application of multiple class problems and stochastic interpretation. First of all, there is an advantage that one output value affects the remaining output value and its accompanying learning is accelerated. Also, to solve the overfitting problem, L2-normalization is applied. To prove the superiority of the proposed pattern classifier, we compare and analyze the classification rate with conventional radial basis function neural networks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.10
• /
• pp.4054-4071
• /
• 2015

Named-Data Networking (NDN) is one of the promising approaches for the Future Internet to cope with the explosion and current usage pattern of Internet traffic. Content provider mobility in the NDN allows users to receive real-time traffic when the content providers are on the move. However, the current solutions for managing these mobile content providers suffer several issues such as long handover latency, high cost, and non-optimal routing path. In this paper, we survey main approaches for provider mobility in NDN and propose an optimal scheme to support the mobile content providers in the large-scale NDN domain. Our scheme predicts the movement of the provider and uses state information in the NDN forwarding plane to set up an optimal new routing path for mobile providers. By numerical analysis, our approach provides NDN users with better service access delay and lower total handover cost compared with the current solutions.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.7
• /
• pp.805-812
• /
• 2005

Smart home is one of the ubiquitous environment platforms with various complex sensor-and-control network. In this paper, a now learning methodology for learning user's behavior preference pattern is proposed in the sense of reductive user's cognitive load to access complex interfaces and providing personalized services. We propose a fuzzy inductive learning methodology based on life-long learning paradigm for knowledge discovery, which tries to construct efficient fuzzy partition for each input space and to extract fuzzy association rules from the numerical data pattern.

• Knowledge Management Research
• /
• v.9 no.3
• /
• pp.59-76
• /
• 2008

Due to rapid increase in number of Internet users, the volume of domestic Internet shopping is expanding dramatically. Although the Internet shopping mall business is steadily growing, it has structural weakness, and in general, the business does not generate profit. In this context, this research identifies characteristics of the Internet shopping non-adopters, and suggests strategies to strengthen competitiveness of domestic Internet shopping malls and to create a new demand. Analysis of MCR data in 2007 showed that out of 4,298 respondents, 2,206 people adopted Internet shopping (51.3%). and 2,092 did not (48.7%). The survey measured 28 items regarding a consumption pattern and a lifestyle of the Internet users, and the analysis result showed that the pattern can be categorized in seven groups. Based on the analysis, the research suggests that the domestic Internet shopping malls adopt strategies to increase consumers' access frequency to tile Internet service, provide high-end goods, diversify transaction methods, make online shopping more convenient, cater to diversified consumer demands based on demographic data, provide price comparison more Internet shopping mails, and provide sufficient and useful information to consumers.

• Journal of KIISE:Software and Applications
• /
• v.32 no.5
• /
• pp.416-427
• /
• 2005

Web navigation mining is a method of discovering Web navigation patterns by analyzing the Web access log data. However, it is admitted that the log data contains noisy information that leads to the incorrect recognition of user navigation path on the Web's hyperlink structure. As a result, previous Web navigation mining systems that exploited solely the log data have not shown good performance in discovering correct Web navigation patterns efficiently, mainly due to the complex pre-processing procedure. To resolve this problem, this paper proposes a technique of amalgamating the Web's hyperlink structure information with the Web access log data to discover navigation patterns correctly and efficiently. Our implemented Web navigation mining system called SPMiner produces a WebTree from the hyperlink structure of a Web site that is used trl eliminate the possible noises in the Web log data caused by the user's abnormal navigational activities. SPMiner remarkably reduces the pre-processing overhead by using the structure of the Web, and as a result, it could analyze the user's search patterns efficiently.

• Journal of Broadcast Engineering
• /
• v.10 no.2
• /
• pp.182-189
• /
• 2005

In this paper, an asymmetric WDM-EPON transmission scheme is proposed to be used in a high speed access network system, which is required to implement the convergence of broadcast and communication. WDM is used for downstream transmission from OLT to access nodes, satisfying wide bandwidth requirement for broadcasting and various multimedia services. And an EPON scheme, which is cheaper than WDM, is applied to upstream transmission where less bandwidth is required. A transmission test in physical layer was performed successfully and the results are provided. If ONUs are to be used in a home gateway, its protocol should be appropriate to its traffic pattern. Voice is sensitive to a time delay while data is not. A new dynamic bandwidth assignment protocol for PON system, which can cope with various types of data in access network is proposed and its performance is analysed. A maximum cycle time is specified to achieve the QoS of signals sensitive to time delay. And a minimum window is specified to prevent the downstream control signals from uprising. It is shown by simulation that the proposed EPON protocol can provide a better performance than previous ones.

• Journal of the Institute of Convergence Signal Processing
• /
• v.3 no.3
• /
• pp.27-35
• /
• 2002

A knowledge-based image retrieval technique is image searching methods using some features from the queried image. The materials in this study are automobile head lamps. The input data is composed of characters and images which have various pattern. The numbers, special symbols, and general letters are under the category of the character. The image informations are made up of the distribution of pixel data, statistical analysis, and state of pattern which are useful for the knowledge data. In this paper, we implemented a retrieval system for the scientific crime detection at traffic accident using the proposed multi knowledge-based image retrieval technique. The values for the multi knowledge-based image features were extracted from color and gray scale each. With this 22 features, we improved the retrieval efficiency about the color information and pattern information. Visual basic, crystal report and MS access DB were used for this application. We anticipate the efficient scientific detection for the traffic accident and the tracking of suspicious vehicle.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.3
• /
• pp.85-93
• /
• 2012

In this paper, I designed the satellite system, using the MF-TDMA (Multi Frequency-Time Division Multiple Access) DAMA (Demand-Assigned Multiple Access) method, that allocates the satellite resources more effectively to prevent a large quantity data discard when Earth Terminals (ETs) request their satellite network resource to Network Controller (NC) by using their history information, such as traffic pattern or traffic class which have been receiving, and the network controller allocates the satellite network resource and send to earth terminals by using his history information that earth terminals have been requested already. The U.S. Military Warfighter Information Network-Tactical (WIN-T) community has selected the L-3 Linkabit MF-TDMA DAMA Network Centric Waveform (NCW) as the networking standard for full-mesh IP over SHF satellite communications. In the MF-TDMA DAMA satellite network, network controller allocates the satellite network resources and enables maximum 255 earth terminals to communicate each other for periodic satellite network resource requests of earth terminals.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.6B
• /
• pp.569-578
• /
• 2002

This paper proposes a new multicarrier code division multiple access (CDMA) system. The proposed multicarrier CDMA system provides the advantages that the transmission bandwidth is more efficiently utilized by using divided spreading sequence, time and frequency diversity is achieved in frequency selective nultipath (acting channel, and inter-carrier interference (ICI) can be minimized by using specific data and code pattern. In this system, transmitted data bits are serial-to-parallel converted to some parallel branches. On each branch each bit is direct-sequence spread-spectrum modulated by divided spreading sequences and transmitted using orthogonal carriers. The receiver providers a Rake for each carrier, and the outputs of Rakes are combined to get time and frequency diversity. This multicarrier CDMA system allows additional flexibility in the choice of system parameters. Upon varying system parameters, bit error rate (BER) performance is examined for the proposed multicarrier CDMA system. Simulation results show that the proposed multicarrier CDMA scheme can achieve better performance than the other types of conventional multicarrier CDMA systems.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.24 no.2
• /
• pp.315-323
• /
• 1987

Generally, general-purpose image processing system is so expensive that not so many users easily can access the system. In this paper attemps have been made to design and describe a general and economical image processing system for real-time aplications such as image data compression, pattern recognition and target tracking. The system comprises an operator console, image data acquisition/display sistem and IBM PC/XT. The system also utilizes a high speed Fairchild 16-bit microprocessor with ALU speed of 375 nsec for system control, algrithm execution and user computation. The system also can digitize /display a 256x 256x 8 bit image in real time and store two frames of images. All image pixels are directly accessible by the microprocessor for fast and efficient computation. Some experimental and illustrative results such as target tracking are presented to show the efficient performance of the system.