• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.29-39
• /
• 2021

As damages to individuals, private sectors, and businesses increase due to newly occurring cyber attacks, the underlying network security problem has emerged as a major problem in computer systems. Therefore, NIDS using machine learning and deep learning is being studied to improve the limitations that occur in the existing Network Intrusion Detection System. In this study, a deep learning-based NIDS model study is conducted using the Convolution Neural Network (CNN) algorithm. For the image classification-based CNN algorithm learning, a discrete algorithm for continuity variables was added in the preprocessing stage used previously, and the predicted variables were expressed in a linear relationship and converted into easy-to-interpret data. Finally, the network packet processed through the above process is mapped to a square matrix structure and converted into a pixel image. For the performance evaluation of the proposed model, NSL-KDD, a representative network packet data, was used, and accuracy, precision, recall, and f1-score were used as performance indicators. As a result of the experiment, the proposed model showed the highest performance with an accuracy of 85%, and the harmonic mean (F1-Score) of the R2L class with a small number of training samples was 71%, showing very good performance compared to other models.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.3
• /
• pp.465-475
• /
• 2022

Due to the impact of COVID-19, people are paying attention to convenience and health, and the use of IoT devices to help them is increasing. In order to embed a lightweight security element in IoT devices that need to handle sensitive information even with limited resources, the development of a lightweight S-box is essential. Until 2021, it was common to develop a lightweight 4-bit S-box by a heuristic method, and to develop an extended structure or repeat the same operation for a larger size lightweight S-box. However, in January 2022, a paper that proposed a heuristic algorithm to find an 8-bit S-box with better differential uniformity and linearity than the S-box generated with an MISTY extended structure, although non-bijective, was published [1]. The heuristic algorithm proposed in this paper generates an S-box by adding AND operations one by one. Whenever an AND operation is added, they use a method that pre-removes the S-box for which the calculated differential uniformity does not reach the desired criterion. In this paper, we improve the performance of this heuristic algorithm. By increasing the amount of pre-removal using not only differential uniformity but also other differential property, and adding a process of calculating linearity for pre-removing, it is possible to satisfy not only differential security but also linear security.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.3
• /
• pp.96-106
• /
• 2009

Extending the scale of the distributed processing in a single Ethernet cluster is physically restricted by maximum ports per switch. This paper presents an implementation of MPI-based multicluster consisting of multiple Ethernet switches for extending the scale of distributed processing, and a asymptotical analysis for communication overhead through execution-time analysis model. To determine an optimum task partitioning, we analyzed the processing time for various partitioning schemes, and AAP(accumulator array partitioning) scheme was finally chosen to minimize the overall communication overhead. The scope of data partitioned in AAP was modified to fit for incremented nodes, and suitable load balancing algorithm was implemented. We tried to alleviate the communication overhead through exploiting the pipelined broadcast and flat-tree based result gathering, and overlapping of the communication and the computation time. We used the linear pipeline broadcast to reduce the communication overhead in intercluster which is interconnected by a single link. Experimental results shows nearly linear speedup by the proposed parallel distributed GHT implemented on MPI-based Ethernet multicluster with four 100Mbps Ethernet switches and up to 128 nodes of Pentium PC.

• Journal of Korea Water Resources Association
• /
• v.42 no.8
• /
• pp.643-657
• /
• 2009

The major reason to construct large dams is to store surplus water during rainy seasons and utilize it for water supply in dry seasons. Reservoir storage has to meet a pre-defined target to satisfy water demands and cope with a dry season when the availability of water resources are limited temporally as well as spatially. In this study, a Hedging rule that reduces total reservoir outflow as drought starts is applied to alleviate severe water shortages. Five stages for reducing outflow based on the current reservoir storage are proposed as the Hedging rule. The objective function is to minimize the total discrepancies between the target and actual reservoir storage, water supply and demand, and required minimum river discharge and actual river flow. Mixed Integer Linear Programming (MILP) is used to develop a multi-reservoir operation system with the Hedging rule. The developed system is applied for the Han River basin that includes four multi-purpose dams and one water supplying reservoir. One of the fours dams is primarily for power generation. Ten-day-based runoff from subbasins and water demand in 2003 and water supply plan to water users from the reservoirs are used from "Long Term Comprehensive Plan for Water Resources in Korea" and "Practical Handbook of Dam Operation in Korea", respectively. The model was optimized by GAMS/CPLEX which is LP/MIP solver using a branch-and-cut algorithm. As results, 99.99% of municipal demand, 99.91% of agricultural demand and 100.00% of minimum river discharge were satisfied and, at the same time, dam storage compared to the storage efficiency increased 10.04% which is a real operation data in 2003.

• Analytical Science and Technology
• /
• v.24 no.6
• /
• pp.533-543
• /
• 2011

The hERG (human ether-a-go-go related gene) ion channel is a main factor for cardiac repolarization, and the blockade of this channel could induce arrhythmia and sudden death. Therefore, potential hERG ion channel inhibitors are now a primary concern in the drug discovery process, and lots of efforts are focused on the minimizing the cardiotoxic side effect. In this study, $IC_{50}$ data of 202 organic compounds in HEK (human embryonic kidney) cell from literatures were used to develop predictive 2D-QSAR model. Multiple linear regression (MLR), Support Vector Machine (SVM), and artificial neural network (ANN) were utilized to predict inhibition concentration of hERG ion channel as machine learning methods. Population based-forward selection method with cross-validation procedure was combined with each learning method and used to select best subset descriptors for each learning algorithm. The best model was ANN model based on 14 descriptors ($R^2_{CV}$=0.617, RMSECV=0.762, MAECV=0.583) and the MLR model could describe the structural characteristics of inhibitors and interaction with hERG receptors. The validation of QSAR models was evaluated through the 5-fold cross-validation and Y-scrambling test.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.47 no.6
• /
• pp.12-18
• /
• 2010

Automatic Target Detection (ATD) systems that use forward-looking infrared (FLIR) consists of three stages. preprocessing, detection, and clutter rejection. All potential targets are extracted in preprocessing and detection stages. But, this results in a high false alarm rates. To reduce false alarm rates of ATD system, true targets are extracted in the clutter rejection stage. This paper focuses on clutter rejection stage. This paper presents a new clutter rejection technique using PCA features and stochastic features of clutters and targets. PCA features are obtained from Euclidian distances using which potential targets are projected to reduced eigenspace selected from target eigenvectors. CV is used for calculating stochastic features of edges in targets and clutters images. To distinguish between target and clutter, LDA (Linear Discriminant Analysis) is applied. The experimental results show that the proposed algorithm accurately classify clutters with a low false rate compared to PCA method or CV method

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.5C
• /
• pp.706-715
• /
• 2004

In this paper, to decide the watermark embedding positions and embed the watermark we use the subband tee structure which is presented in the wavelet domain and the edge information in the spatial domain. The significant frequency region is estimated by the subband searching from the higher frequency subband to the lower frequency subband. LH1 subband which has the higher frequency in tree structure of the wavelet domain is divided into 4${\times}$4 submatrices, and the threshold which is used in the watermark embedding is obtained by the blockmatrix which is consists by the average of 4${\times}$4 submatrices. Also the watermark embedding position, Keymap is generated by the blockmatrix for the energy distribution in the frequency domain and the edge information in the spatial domain. The watermark is embedded into the wavelet coefficients using the Keymap and the random sequence generated by LFSR(Linear feedback shift register). Finally after the inverse wavelet transform the watermark embedded image is obtained. the proposed watermarking algorithm showed PSNR over 2㏈ and had the higher results from 2％ to 8％ in the comparison with the previous research for the attack such as the JPEG compression and the general image processing just like blurring, sharpening and gaussian noise.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.11
• /
• pp.143-156
• /
• 2010

Faced with an increase of malicious threats from the Internet as well as local area networks, many companies are considering deploying a security system. To help a decision maker select a suitable security tool, this paper proposed a three-step integrated framework using linear fuzzy regression (LFR) and physical programming (PP). First, based on the experts' estimations on security criteria, analytic hierarchy process (AHP) and quality function deployment (QFD) are employed to specify an intermediate score for each criterion and the relationship among these criteria. Next, evaluation value of each criterion is computed by using LFR. Finally, a goal programming (GP) method is customized to obtain the most appropriate security tool for an organization, considering a tradeoff among the multi-objectives associated with quality, credibility and costs, utilizing the relative weights calculated by the physical programming weights (PPW) algorithm. A numerical example provided illustrates the advantages and contributions of this approach. Proposed approach is anticipated to help a decision maker select a suitable security tool by taking advantage of experts' experience, with noises eliminated, as well as the accuracy of mathematical optimization methods.

• Geophysics and Geophysical Exploration
• /
• v.7 no.3
• /
• pp.207-212
• /
• 2004

This article reviews recent developments in three-dimensional (3-D) magntotelluric (MT) imaging. The inversion of MT data is fundamentally ill-posed, and therefore the resultant solution is non-unique. A regularizing scheme must be involved to reduce the non-uniqueness while retaining certain a priori information in the solution. The standard approach to nonlinear inversion in geophysis has been the Gauss-Newton method, which solves a sequence of linearized inverse problems. When running to convergence, the algorithm minimizes an objective function over the space of models and in the sense produces an optimal solution of the inverse problem. The general usefulness of iterative, linearized inversion algorithms, however is greatly limited in 3-D MT applications by the requirement of computing the Jacobian(partial derivative, sensitivity) matrix of the forward problem. The difficulty may be relaxed using conjugate gradients(CG) methods. A linear CG technique is used to solve each step of Gauss-Newton iterations incompletely, while the method of nonlinear CG is applied directly to the minimization of the objective function. These CG techniques replace computation of jacobian matrix and solution of a large linear system with computations equivalent to only three forward problems per inversion iteration. Consequently, the algorithms are efficient in computational speed and memory requirement, making 3-D inversion feasible.

• Journal of Internet Computing and Services
• /
• v.8 no.3
• /
• pp.19-32
• /
• 2007

This paper present a new face recognition method based on LVQ neural net to construct a real time face recognition system. The previous researches which used PCA, LDA combined neural net usually need much time in training neural net. The supervised LVQ neural net needs much less time in training and can maximize the separability between the classes. In this paper, the proposed method transforms the input face image by PCA and LDA sequentially into low-dimension feature vectors and recognizes the face through LVQ neural net. In order to make the system robust to external light variation, light compensation is performed on the detected face by max-min normalization method as preprocessing. PCA and LDA transformations are applied to the normalized face image to produce low-level feature vectors of the image. In order to determine the initial centers of LVQ and speed up the convergency of the LVQ neural net, the K-Means clustering algorithm is adopted. Subsequently, the class representative vectors can be produced by LVQ2 training using initial center vectors. The face recognition is achieved by using the euclidean distance measure between the center vector of classes and the feature vector of input image. From the experiments, we can prove that the proposed method is more effective in the recognition ratio for the cases of still images from ORL database and sequential images rather than using conventional PCA of a hybrid method with PCA and LDA.