• Proceedings of the IEEK Conference
• /
• 2003.07d
• /
• pp.1251-1254
• /
• 2003

In this paper, Experience Sensitive Cumulative Neural Network (ESCNN) is introduced, which can cumulate the same or similar experiences. As the same or similar training patterns are cumulated in the network, the system recognize more important information in the training patterns. The functions of forgetting less important informations and attending more important informations resided in the training patterns are surveyed and implemented by simulations. The system behaves well under the noisy circumstances due to its forgetting and/or attending properties, even in 50 percents noisy environments. This paper also describes the creation of the generalized patterns for the input training patterns.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.3
• /
• pp.659-667
• /
• 2013

If the regression model or the propensity model is correct, the unbiasedness of the estimator using doubly robust imputation can be guaranteed. Using a neural network instead of a logistic regression model for the propensity model, the estimators using doubly robust imputation are approximately unbiased even though both assumed models fail. We also propose a doubly robust estimator of ratio form using population information of an auxiliary variable. We prove some properties of proposed theory by restricted simulations.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.11
• /
• pp.1571-1574
• /
• 2014

In this paper, we propose a unidirectional transmission of critical information obtained by keyboard hacking or kernel and keyboard driver hacking even though the computer is not connected to the external network. We show the hacking can be attempted in the proposed method to show the way preventing such attempts in advance. Firewalls and other various methods are used to prevent the hacking from the external network but the hacking is also attempted in various ways to detour the firewall. One of the most effective way preventing from the hacking attack is physically disconnect the internal intranet systems from the external internet and most of the government systems, military systems and big corporate systems are using this way as on one of the protection method. In this paper, we show the feasibility of transmission of security codes, etc via the short message to the external network on the assumption that a hacking program such as Trojan Horse is installed on the computer systems separated from the external network. Previous studies showed that the letters on the monitor can be hijacked by electromagnetic analysis on the computer to obtain the information even though the system is not connected ti the network. Other studies showed that the security code hint can obtained by analyzing the power consumption distribution of CPU. In this paper, the power consumption distribution of externally accessible power line is analyzed to obtain the information and the information can be transmitted to the external network. Software controlling the CPU and GPU usage is designed to control the power supply of computer. The sensors such as the Rogowski coils can be used on the external power line to collect the data of power consumption change rates. To transmit the user password by short message, due to the capacitive components and the obstacle from other power supply, A very slow protocol are used.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.12A
• /
• pp.1181-1187
• /
• 2010

This paper proposes a mobile AI (Artificial Intelligence) conducting decision-making in the game through education for intelligent character on the basis of Neural Network. Neural Network is learned through the input/output value of the algorithm which defines the game rule and the problem solving method. The learned character is able to perceive the circumstances and make proper action. In this paper, the mobile AI using Neural Network has been step-by-step designed, and a simple game has been materialized for its functional experiment. In this game, the goal, the character, and obstacles exist on regular 2D space, and the character, evading obstacles, has to move where the goal is. The mobile AI can achieve its goals in changing environment by learning the solution to several problems through the algorithm defined in each experiment. The defined algorithm and Neural Network are designed to make the input/output system the same. As the experimental results, the suggested mobile AI showed that it could perceive the circumstances to conduct action and to complete its mission. If mobile AI learns the defined algorithm even in the game of complex structure, its Neural Network will be able to show proper results even in the changing environment.

• Journal of Communications and Networks
• /
• v.13 no.5
• /
• pp.518-524
• /
• 2011

A high assurance IP encryption (HAIPE) compliant protocol accelerator is proposed for military networks consisting of red (or classified) networks and black (or unclassified) networks. The boundary between red and black sides is assumed to be protected via a HAIPE device. However, the IP layer encryption introduces challenges for bandwidth on demand satellite communication. The problems experienced by transmission control protocol (TCP) over satellites are well understood: While standard modems (on the black side) employ TCP performance enhancing proxy (PEP) which has been shown to work well, the HAIPE encryption of TCP headers renders the onboard modem's PEP ineffective. This is attributed to the fact that under the bandwidth-on-demand environment, PEP must use traditional TCP mechanisms such as slow start to probe for the available bandwidth of the link (which eliminates the usefulness of the PEP). Most implementations recommend disabling the PEP when a HAIPE device is used. In this paper, we propose a novel solution, namely broadband HAIPE-embeddable satellite communications terminal (BHeST), which utilizes dynamic network performance enhancement algorithms for high latency bandwidth-on-demand satellite links protected by HAIPE. By moving the PEP into the red network and exploiting the explicit congestion notification bypass mechanism allowed by the latest HAIPE standard, we have been able to regain PEP's desired network enhancement that was lost due to HAIPE encryption (even though the idea of deploying PEP at the modem side is not new). Our BHeST solution employs direct video broadcast-return channel service (DVB-RCS), an open standard as a means of providing bandwidth-on-demand satellite links. Another issue we address is the estimation of current satellite bandwidth allocated to a remote terminal which is not available in DVBRCS. Simulation results show that the improvement of our solution over FIX PEP is significant and could reach up to 100%. The improvement over the original TCP is even more (up to 500% for certain configurations).

• Journal of KIBIM
• /
• v.5 no.1
• /
• pp.54-62
• /
• 2015

The Cities that recently developed have been applied to private network for establishing information communication network system. The local governments planning or pursuing U-City construction should also choose the private network in consideration of operation and maintenance. In viewpoint of agency operating u-City, it is necessary to integrate traditional and new network. However, there has been lack of guides to choose U-City network considering the economic analysis between private and leased network. This study analyzed the characteristics of private and leased network, and the cost-benefit by estimating the network cost and communication demand focused on U-services that are recently applied. This study purpose a guide for efficient U-City information network selected by estimating ROI(Return On Investment) and BEP(Break Even Point) for establishing private and leased network.

• ETRI Journal
• /
• v.37 no.2
• /
• pp.380-386
• /
• 2015

Guaranteeing quality of service over a multihop wireless network is difficult because end-to-end (ETE) delay is accumulated at each hop in a multihop flow. Recently, research has been conducted on network coding (NC) schemes as an alternative mechanism to significantly increase the utilization of valuable resources in multihop wireless networks. This paper proposes a new section-based joint NC and scheduling scheme that can reduce ETE delay and enhance resource efficiency in a multihop wireless network. Next, this paper derives the average ETE delay of the proposed scheme and simulates a TDMA network where the proposed scheme is deployed. Finally, this paper compares the performance of the proposed scheme with that of the conventional sequential scheduling scheme. From the performance analysis and simulation results, the proposed scheme gives more delay-and energy-efficient slot assignments even if the NC operation is applied, resulting in a use of fewer network resources and a reduction in ETE delay.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.587-591
• /
• 2003

This study develops a neural network for solving optimization problems. Hopfield network has been used for such problems, but it frequently gives abnormal solutions or non-optimal solutions. Moreover, it takes much time for solving a solution. To overcome such disadvantages, this study adopts a neural network whose output nodes change with a small value at every evolution, and the proposed neural network is applied to solve ALB (Assembly Line Balancing) problems . Given a precedence diagram and a required number of workstations, an ALB problem is solved while achieving even distribution of workload among workstations. Here, the workload variance is used as the index of workload deviation, and is reflected to an energy function. The simulation results show that the proposed neural network yields good results for solving ALB problems with high success rate and fast execution time.

• Journal of Satellite, Information and Communications
• /
• v.7 no.3
• /
• pp.47-53
• /
• 2012

In this paper, we present an algorithm called redundant logical paths (RLP) for efficient HSR traffic reduction. It creates redundant logical paths between each HSR node and all the other nodes. Eventually, a logical full-meshed network or paths will be established among all HSR node types, except the Quadbox type, which is used only for interconnection. The logical full-meshed network will be used instead of using the standard HSR protocol that depends on the concepts of the duplication and forwarding of the received frame until it reaches the destination node. The RLP algorithm results in significantly less frame traffic because there is no random forwarding as in the standard HSR protocol. For the sample network in this paper simulation results showed a 61.5-80% reduction in network frame traffic compared to the standard HSR. Our algorithm will avoid latency issues in the network and even network congestion, thus improving network efficiency.

• Journal of information and communication convergence engineering
• /
• v.7 no.3
• /
• pp.241-246
• /
• 2009

Advances in mobile and wireless networking technologies have enabled mobile ad hoc networks applicable to a wide range of areas. Many cases of even such ad hoc networks demand to be accessible to the global network. Owing to be diversified in features depending on applications, however, some those networks may consist of devices or nodes which do not facilitate all the same transport protocols. This results in unreachable situations of establishing ad hoc network: such as non-existence of all required access points, faults or contention in a path of particular protocol communication. Interworking between different transport protocols can alleviate such problems. This paper proposes an interworking scheme to improve data transport performance and network accessibility, especially in such a mobile ad hoc network that is applicable to data communications among ships or user's convenient u-health services. Modeling and simulation analysis are used to verify the improvement.