• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06c
• /
• pp.407-410
• /
• 1994

In this paper, we describe the new unsuperivised learning algorithm, SASOM. It can solve the defects of the conventional SOM that the state of network can't converge to the minimum point. The proposed algorithm uses the object function which can evaluate the state of network in learning and adjusts the learning rate flexibly according to the evaluation of the object function. We implement the simulated annealing which is applied to the conventional network using the object function and the learning rate. Finally, the proposed algorithm can make the state of network converged to the global minimum. Using the two-dimensional input vectors with uniform distribution, we graphically compared the ordering ability of SOM with that of SASOM. We carried out the recognitioin on the new algorithm for all Korean phonemes and some continuous speech.

• Journal of Information Processing Systems
• /
• v.2 no.3 s.4
• /
• pp.183-188
• /
• 2006

A RMON agent system, which locates on a subnet, collects the network traffic information for management by retrieving and analyzing all of the packets on the subnet. The RMON agent system can miss some packets due to the high packet analyzing overhead when the number of packets on the subnet is huge. In this paper, we have developed a light-weight RMON agent system that can handle a large amount of packets without packet loss. Our RMON agent system has also been designed such that its functionality can be added dynamically when needed. To demonstrate the dynamic reconfiguration capability of our RMON agent system, a simple port scanning attack detection module is added to the RMON agent system. We have also evaluated the performance of our RMON agent system on a large network that has a huge traffic. The test result has shown our RMON agent system can analyze the network packets without packet loss.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.34C no.7
• /
• pp.70-81
• /
• 1997

In this paper, we propose HMCNN(hybrid multiple component neural networks) that enhance performance of MCNN by adapting new pattern partitioning algorithm which can cluster many input patterns efficiently. Added neural network performs similar learning procedure that of kohonen network. But it dynamically determine it's number of output neurons using algorithms that decide self-organized number of clusters and patterns in a cluster. The proposed network can effectively be applied to problems of large data as well as huge networks size. As a sresutl, proposed pattern partitioning network can enhance performance results and solve weakness of MCNN like generalization capability. In addition, we can get more fast speed by performing parallel learning than that of other supervised learning networks.

• Journal of information and communication convergence engineering
• /
• v.16 no.3
• /
• pp.160-165
• /
• 2018

This paper designed and developed the image processing system of integrating feature extraction and matching by using convolutional neural network (CNN), rather than relying on the simple method of processing feature extraction and matching separately in the image processing of conventional image recognition system. To implement it, the proposed system enables CNN to operate and analyze the performance of conventional image processing system. This system extracts the features of an image using CNN and then learns them by the neural network. The proposed system showed 84% accuracy of recognition. The proposed system is a model of recognizing learned images by deep learning. Therefore, it can run in batch and work easily under any platform (including embedded platform) that can read all kinds of files anytime. Also, it does not require the implementing of feature extraction algorithm and matching algorithm therefore it can save time and it is efficient. As a result, it can be widely used as an image recognition program.

• Journal of the military operations research society of Korea
• /
• v.18 no.2
• /
• pp.66-95
• /
• 1992

Shortest paths are one of the simplest and most widely used concepts in deterministic networks. A decison of troops movement route can be analyzed in the network with a shortest path algorithm. But in reality, the value of arcs can not be determined in the network by crisp numbers due to imprecision or fuzziness in parameters. To account for this reason, a fuzzy network should be considered. A fuzzy shortest path can be modeled by general fuzzy mathematical programming and solved by fuzzy dynamic programming. It can be formulated by the fuzzy network with lingustic variables and solved by the Klein algorithm. This paper focuses on a revised fuzzy shortest path algorithm and an application is discussed.

• Journal of Communications and Networks
• /
• v.14 no.2
• /
• pp.169-178
• /
• 2012

In this paper, we study the advantages of using range extension cooperative transmission (CT) in multi-hop energy harvesting wireless sensor networks (EH-WSNs) from the network layer perspective. EH-WSNs rely on harvested energy, and therefore, if a required service is energy-intensive, the network may not be able to support the service successfully. We show that CT networks that utilize both range extension CT and non-CT routing can successfully support services that cannot be supported by non-CT networks. For a two-hop toy network, we show that range extension CT can provide better services than non-CT. Then, we provide a method of determining the supportable services that can be achieved by using optimal non-CT and CT routing protocols for EH-WSNs. Using our method and network simulations, we justify our claim that CT networks can provide better services than nonCT networks in EH-WSNs.

• Journal of the Korea Society for Simulation
• /
• v.7 no.2
• /
• pp.1-16
• /
• 1998

This paper presents an automated methodology for campus network design and performance analysis using the rule-based SES and DEVS modeling & simulation techniques. Proposed methodology for structural design and performance analysis can be utilized not only in the early stage of network design for selecting configurable candidate from all possible design alternatives, but also in simulation verification for generating performance data. Our approach supercedes conventional methodologies in that, first, it can support the configuration automation by utilizing the knowledge of design expert ; second, it can provide the simulation-based performance evaluation ; third, it is established on the basis of the well-formalized framework so that it can support a hierarchical and modular system design. Several simulation tests performed on a campus network example will illustrate our technique.

• Journal of Communications and Networks
• /
• v.10 no.4
• /
• pp.384-395
• /
• 2008

Random network coding can be viewed as a single block code applied to all source packets. To manage the concomitant high coding complexity, source packets can be partitioned into generations; block coding is then performed on each set. To reach a better performance-complexity tradeoff, we propose a novel concatenated network code which mixes generations while retaining the desirable properties of generation-based coding. Focusing on the code's erasure performance, we show that the probability of successfully decoding a generation on erasure channels can increase substantially for any erasure rate. Using both analysis (for small networks) and simulations (for larger networks), we show how the code's parameters can be tuned to extract best performance. As a result, the probability of failing to decode a generation is reduced by nearly one order of magnitude.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2242-2244
• /
• 2003

This paper considers a networked control system (NCS) that consists of an inverted cart pendulum, a digital controller, and a controller area network (CAN) in which the actuator and sensors of the pendulum are connected to form a closed-loop system. The worst-case message response time (WCMRT) in the CAN is analyzed and the analysis results are applied to the target control system. For the case where the control system cannot satisfy the WCMRT condition and therefore time delays are inevitable, the Luck and Ray method is used to compensate the network-induced time delays. Simulations are carried out to show the feasibility of the proposed scheme.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.11 no.2
• /
• pp.205-212
• /
• 2011

The high performance intelligent robots have become to use a lot of sensors, actuators and controllers which are connected via bus-type networks while the point to point connection was widely used in-robot networks of existing robots. The bus topology makes the robot system share data internally and it can also reduce the weight of the robot because of smaller wiring harness. The CAN is one of serial bus communication protocols typically used for in-vehicle networks and now it can also be used for in-robot networks because of its low cost and reliability. A performance evaluation system for this type of networked system is developed to analyze the real-time operation and the reliability of a serial-bus network, and the usefulness of the developed evaluation system is shown by some experiments for a network of many CAN nodes.