• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.631-634
• /
• 2004

In this paper, the Bayesian recurrent neural network (BRNN) is proposed to predict time series data. Among the various traditional prediction methodologies, a neural network method is considered to be more effective in case of non-linear and non-stationary time series data. A neural network predictor requests proper learning strategy to adjust the network weights, and one need to prepare for non-linear and non-stationary evolution of network weights. The Bayesian neural network in this paper estimates not the single set of weights but the probability distributions of weights. In other words, we sets the weight vector as a state vector of state space method, and estimates its probability distributions in accordance with the Bayesian inference. This approach makes it possible to obtain more exact estimation of the weights. Moreover, in the aspect of network architecture, it is known that the recurrent feedback structure is superior to the feedforward structure for the problem of time series prediction. Therefore, the recurrent network with Bayesian inference, what we call BRNN, is expected to show higher performance than the normal neural network. To verify the performance of the proposed method, the time series data are numerically generated and a neural network predictor is applied on it. As a result, BRNN is proved to show better prediction result than common feedforward Bayesian neural network.

• Convergence Security Journal
• /
• v.16 no.7
• /
• pp.3-11
• /
• 2016

With repid development of optical networking technology, Software-Defined Network (SDN) and Network Function Virtualization (NFV), high performance networking service, collaboration platform that enables collaborative research globally, drastically National Research Network (NRN) including Internet Service has changed. Therefore we compared and analyzed several world-class NRNs and took a view of future development strategy of the NRN. Also we suggest high speed security environment in super high bandwidth network with 40Gbps and 100Gbps optical transmission technology, network separation of NRN with Science DMZ to support high performance network transmission for science big data, building security environment for last-mile in campus network that supports programmability of IDS using BRO framework.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4246-4267
• /
• 2020

As the network goes deep into all aspects of people's lives, the number and the complexity of network traffic is increasing, and traffic classification becomes more and more important. How to classify them effectively is an important prerequisite for network management and planning, and ensuring network security. With the continuous development of deep learning, more and more traffic classification begins to use it as the main method, which achieves better results than traditional classification methods. In this paper, we provide a comprehensive review of network traffic classification based on deep learning. Firstly, we introduce the research background and progress of network traffic classification. Then, we summarize and compare traffic classification based on deep learning such as stack autoencoder, one-dimensional convolution neural network, two-dimensional convolution neural network, three-dimensional convolution neural network, long short-term memory network and Deep Belief Networks. In addition, we compare traffic classification based on deep learning with other methods such as based on port number, deep packets detection and machine learning. Finally, the future research directions of network traffic classification based on deep learning are prospected.

• ETRI Journal
• /
• v.25 no.6
• /
• pp.445-463
• /
• 2003

XML-based network management, which applies XML technologies to network management, has been proposed as an alternative to existing network management. The use of XML in network management offers many advantages. However, most existing network devices are already embedded with simple network management protocol (SNMP) agents and managed by SNMP managers. For integrated network management, we present the architectures of an XML-based manager, an XML-based agent, and an XML/SNMP gateway for legacy SNMP agents. We describe our experience of developing an XML-based network management system (XNMS), XML-based agent, and an XML/SNMP gateway. We also verify the effectiveness of our XML-based agent and XML/SNMP gateway through performance tests. Our experience with developing XNMS and XML-based agents can be used as a guideline for development of XML-based management systems that fully take advantage of the strengths of XML technologies.

• Proceedings of the IEEK Conference
• /
• 2005.11a
• /
• pp.707-710
• /
• 2005

In this paper, we propose a new communication network for on-chip communication. The network is based on a direct sequence code division multiple access (DS-CDMA) technique. The new communication network is suitable for a parallel processing system and also drastically reduces the I/O pin count. Our network architecture is mainly divided into a CDMA-based network interface (CNI), a communication channel, a synchronizer. The network includes a reverse communication channel for reducing latency. The network decouples computation task from communication task by the CNI. An extreme truncation is considered to simplify the communication link. For the scalability of the network, we use a PN-code reuse method and a hierarchical structure. The network elements have a modular architecture. The communication network is done using fully synthesizable Verilog HDL to enhance the portability between process technologies.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.12a
• /
• pp.10-13
• /
• 2001

Nonlinear Function Approximation of Moduled Neural Network Using Genetic Algorithm Neural Network consists of neuron and synapse. Synapse memorize last pattern and study new pattern. When Neural Network learn new pattern, it tend to forget previously learned pattern. This phenomenon is called to catastrophic inference or catastrophic forgetting. To overcome this phenomenon, Neural Network must be modularized. In this paper, we propose Moduled Neural Network. Modular Neural Network consists of two Neural Network. Each Network individually study different pattern and their outputs is finally summed by net function. Sometimes Neural Network don't find global minimum, but find local minimum. To find global minimum we use Genetic Algorithm.

• Journal of Information Technology Applications and Management
• /
• v.9 no.3
• /
• pp.97-111
• /
• 2002

This paper attempts to provide an overview of relationship between the characteristics of the network and competitive strategies. We review the theoretical background of the efficiency of network, which Is very important for the network-based industries. Network externality, positive feedback effects, bandwagon effects, economies of scale, economies of scope in network-base businesses are reviewed. Various network situations, including interconnection, and strategies are also discussed. In this purpose, simple but meaningful examples and cases are used to show the economic goals and means of network competition strategies. We try to link network strategies to the generic strategies and coopetition suggested by Porter and by Brandenburger and Nalebuff respectively. Since this study is an exploratory research, further studies on more complex network situation in the real work can be executed with taking advantage of this effort.

• Korean Management Science Review
• /
• v.32 no.3
• /
• pp.29-36
• /
• 2015

In this research, we analyze the sensitivity of the network density to the estimates for the Bass model parameters with both theoretical model and a simulation. Bass model describes the process that the non-adopters in the market potential adopt a new product or an innovation by the innovation effect and imitation effect. The imitation effect shows the word of mouth effect from the previous adopters to non-adopters. But it does not divide the underlying network structure from the strength of the influence over the network. With a network based Bass model, we found that the estimate for the imitation coefficient is highly sensitive to the network density and it is decreasing while the network density is decreasing. This finding implies that the interpersonal influence can be under-looked when the network density is low. It also implies that both of the network density and the interpersonal influence are important to facilitate the diffusion of an innovation.

• Journal of Communications and Networks
• /
• v.12 no.1
• /
• pp.11-18
• /
• 2010

Diversity-multiplexing tradeoff (DMT) is an efficient tool to measure the performance of multiple-input and multiple-output (MIMO) systems and cooperative systems. Recently, cooperative multicast system with wireless network coding stretched tremendous interesting due to that it can drastically enhance the throughput of the wireless networks. It is desirable to apply DMT to the performance analysis on the multicast system with wireless network coding. In this paper, DMT is performed at the three proposed wireless network coding protocols, i.e., non-regenerative network coding (NRNC), regenerative complex field network coding (RCNC) and regenerative Galois field network coding (RGNC). The DMT analysis shows that under the same system performance, i.e., the same diversity gain, all the three network coding protocols outperform the traditional transmission scheme without network coding in terms of multiplexing gain. Our DMT analysis also exhibits the trends of the three network coding protocols' performance when multiplexing gain is changing from the lower region to the higher region. Monte-Carlo simulations verify the prediction of DMT.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.05a
• /
• pp.195-198
• /
• 2000

We propose a new wavelet neural network structure, for which we apply new recurrent nodes to the network, in this paper for the dynamic system identification and control. We will construct the wavelet neural network by using wavelet frame function. The function does not have the best approximation property, but it may be possible to apply some modification to the structure of the network because the constriction of orthogonality is loosened a little. This wavelet neural network we propose can obtain previous state information by its structure of the network without any addition of input, though the conventional wavelet network needs additional previous state input for the improvement of the dynamic performance. In numerical experience, the performance of the new wavelet neural network we propose in the nonlinear system with uncertainity of parameter Is equal to that of the wavelet network which used the additional previous information input, superior to that of the conventional wavelet network.