• Convergence Security Journal
• /
• v.21 no.2
• /
• pp.57-66
• /
• 2021

Deep neural networks (DNNs) provide excellent performance for image, speech, and pattern recognition. However, DNNs sometimes misrecognize certain adversarial examples. An adversarial example is a sample that adds optimized noise to the original data, which makes the DNN erroneously misclassified, although there is nothing wrong with the human eye. Therefore studies on defense against adversarial example attacks are required. In this paper, we have experimentally analyzed the success rate of detection for adversarial examples by adjusting various parameters. The performance of the ensemble defense method was analyzed using fast gradient sign method, DeepFool method, Carlini & Wanger method, which are adversarial example attack methods. Moreover, we used MNIST as experimental data and Tensorflow as a machine learning library. As an experimental method, we carried out performance analysis based on three adversarial example attack methods, threshold, number of models, and random noise. As a result, when there were 7 models and a threshold of 1, the detection rate for adversarial example is 98.3%, and the accuracy of 99.2% of the original sample is maintained.

• Journal of Broadcast Engineering
• /
• v.24 no.2
• /
• pp.306-314
• /
• 2019

Since the resolutions of videos increase rapidly, there are continuing needs for effective video compression methods despite an increase in the transmission bandwidth. In order to satisfy such a demand, a reconstructive video coding (RVC) method by using a super resolution has been proposed. Since RVC reduces the resolution of the input video, when frames are compressed to the same size, the number of bits per pixel increases, thereby reducing coding artifacts caused by video coding. However, RVC method using super resolution is not effective in all target bitrates. Comparing the size of the loss generated while downsizing the resolution and the size of the loss caused by the video compression, only when the size of loss generated in the video compression is larger, RVC method can perform the improved compression performance compared to direct video coding. In particular, since HEVC has considerably higher compression performance than the previous standard video codec, it can be experimentally confirmed that the compression distortions become larger than the distortions of downsizing the resolution only in the very low-bitrate conditions. In this paper, we applied RVC based HEVC in various video types and measured the target bitrates that RVC method can be effectively applied.

• Korean Journal of Remote Sensing
• /
• v.38 no.5_1
• /
• pp.511-521
• /
• 2022

Cloud detection and removal from satellite images is an essential process for topographic observation and analysis. Threshold-based cloud detection techniques show stable performance because they detect using the physical characteristics of clouds, but they have the disadvantage of requiring all channels' images and long computational time. Cloud detection techniques using deep learning, which have been studied recently, show short computational time and excellent performance even using only four or less channel (RGB, NIR) images. In this paper, we confirm the performance dependence of the deep learning network according to the heterogeneous learning dataset with different resolutions. The DeepLabV3+ network was improved so that channel features of cloud detection were extracted and learned with two published heterogeneous datasets and mixed data respectively. As a result of the experiment, clouds' Jaccard index was low in a network that learned with different kind of images from test images. However, clouds' Jaccard index was high in a network learned with mixed data that added some of the same kind of test data. Clouds are not structured in a shape, so reflecting channel features in learning is more effective in cloud detection than spatial features. It is necessary to learn channel features of each satellite sensors for cloud detection. Therefore, cloud detection of heterogeneous sensors with different resolutions is very dependent on the learning dataset.

• Journal of Positioning, Navigation, and Timing
• /
• v.13 no.3
• /
• pp.257-268
• /
• 2024

The Korea Augmentation Satellite System (KASS) satellite was successfully launched and service is being started. By receiving messages transmitted from the KASS satellite, users can employ the messages to improve positioning accuracy or to verify the integrity of Global Positioning System (GPS). In this paper, we propose a test method for a beam-formed GPS receiver developed to improve the survivability of KASS augmented messages that can enhance positioning accuracy even in an environment with jamming or interfering signals. Through the test method proposed in this paper, quantitative verification is performed for a beam-formed GPS receiver aimed at maintaining the augmented navigation solution to which KASS augmented messages are applied by tracking the KASS signal as much as possible under conditions where the jamming signal strength is gradually increasing. In addition, the proposed test method includes three conditions; first, a static lab test method for repeated verification of functions under the same conditions; second, a static outdoor test method for performance verification in an operating environment of a platform equipped with the beam-formed GPS receiver; and finally, a dynamic lab test method for performance verification of a moving platform equipped with the beam-formed GPS receiver toward a jammer. In this paper, we propose a method for simulating the jamming signal incident direction through the phase delay of an RF cable designed to prevent unintentional jamming signal emission in both lab and outdoor tests, and a method of applying test software for injecting a jamming signal to compare the survivability performance consistently according to the presence or absence of beamforming signal processing. Through the proposed test method, it was verified that the augmented navigation solution could be output for a longer time period when the beamforming signal processing was applied to the KASS satellite signal in the beam-formed GPS receiver.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.6
• /
• pp.669-675
• /
• 2014

In this paper, we described the investigated theoretical time synchronization performances and experiment results obtained by commercially provided PTP (Precise Time Protocol) modules when the time of a slave clock is synchronized to the master clock. In the case of the theoretical performance analysis, we investigated 3 types of clock levels such as Crystal Oscillator (XO), TCXO (Temperature Compensated XO) and OCXO (Oven Controlled XO). From the analysis, it was observed that the synchronization performance is greatly influenced by the synchronization period and the required performance under 1 us can be achieved by using XO level clocks when the synchronization period is less than 2 seconds and the uncertainty of the propagation delay is under 100 ns. For the experiments using commercial PTP modules, the synchronization performance was investigated for direct, through 1 hub and through 2 hubs connections between the master clock and the slave clock. From the experiment results, we observed that time synchronization under 90 ns with 1,000 seconds observation interval can be achieved in the case of direct connection.

• Journal of Mechanical Science and Technology
• /
• v.16 no.10
• /
• pp.1187-1200
• /
• 2002

This paper is concerned with the design or an LMI (Linear Matrix Inequality) -based H$\infty$ controller for a line of sight (LOS) stabilization system and with its robustness performance. The linearization of the system is necessary to analyze various nonlinear characteristics, but the linearization entails modeling uncertainties which reduce its performance. In addition, the stability of the LOS can be adversely affected by angular velocity disturbances while the vehicle is moving. As the vehicle accelerates, all the factors that are Ignored and simplified for the linearization tend to Inhibit the performance of the system. The robustness in the face of these uncertainties needs to be assured. This paper employs H$\infty$ control theory to address these problems and the LMI method to provide a suitable controller with minimal constraints for the system. Even though the system matrix does not have a full rank, the proposed method makes it possible to design a H$\infty$ controller and to deal with R and S matrices for reducing the system order. It can be also shown that the proposed robust controller has a better disturbance attenuation and tracking performance. The LMI method is also used to enhance the applicability of the proposed reduced-order H$\infty$ controller for the system given. The LMI-based H$\infty$ controller has superior disturbance attenuation and reference input tracking performance, compared with that of the conventional controller under real disturbances.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.3
• /
• pp.246-252
• /
• 2018

Recent studies have been received much attention on underwater laser communication, which is capable of high data rate. However, in underwater laser communication, distortions caused by absorption and scattering induced performance degradation. A typical way to improve performance is to apply channel coding technique. In the beginning of studies, simple methods such as RS and BCH coding techniques were applied. However, due to distance expansion and performance improvement, channel coding methods with low error probability such as LDPC coded method were applied. In this paper, we analyzed the performance according to the size of the code word N, the distance between the transceivers and the size of the M of the M-ary PPM modulation scheme. Simulation results show that parameter M of M-ary PPM is most effect on performance.

• Journal of Korean Society for Quality Management
• /
• v.48 no.2
• /
• pp.297-307
• /
• 2020

Purpose: The purpose of this study is to analyze the factors that affect ammunition performance by applying data mining techniques to the Ammunition Stockpile Reliability Program (ASRP) data of the 155mm propelling charge. Methods: The ASRP data from 1999 to 2017 have been utilized. Logistic regression and decision tree analysis were used to investigate the factors that affect performance of ammunition. The performance evaluation of each model was conducted through comparison with an artificial neural networks(ANN) model. Results: The results of this study are as follows; logistic regression and the decision tree analysis showed that major defect rate of visual inspection is the most significant factor. Also, muzzle velocity by base charge and muzzle velocity by increment charge are also among the significant factors affecting the performance of 155mm propelling charge. To validate the logistic regression and decision tree models, their classification accuracies have been compared with the results of an ANN model. The results indicate that the logistic regression and decision tree models show sufficient performance which conforms the validity of the models. Conclusion: The main contribution of this paper is that, to our best knowledge, it is the first attempt at identifying the significant factors of ASPR data by using data mining techniques. The approaches suggested in the paper could also be extended to other types ammunition data.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.18 no.4
• /
• pp.1-8
• /
• 2014

Overexpanded flow of an axisymmetric thruster nozzle is numerically simulated to investigate effects of nozzle pressure ratio (NPR) on the shock structure and thrust performance. The Reynolds-averaged Navier-Stokes equations with k-${\omega}$ SST turbulence model are solved utilizing FLUENT solver. As the NPR is raised, thrust performance monotonically increases with the shock structure and flow-separation point being pushed toward the nozzle exit. It is also discussed that the flow structure at nozzle-exit plane which is immediately affected by a position of nozzle-interior shocks and expansion waves, has strong influence upon the thrust performance of thruster nozzle.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.11 no.5
• /
• pp.689-696
• /
• 2000

In this paper, the cascaded co-channel interference canceller using the dead zone decision is proposed to reduced the co-channel interference in the DS/CDMA communication with the multipath fading and its performance is analyzed. The cancellation scheme creates the replica of the co-channel interference using the data decision of correlator output and removes it at the next stage. The canceller uses the dead zone decision which is detecting strong signal to obtain a good performance in the mobile communication. The results of performance analysis show that the proposed cancellation scheme has the improved performance and the number of user is increased 4 times compared with the conventional receiver, and increased 17% compared with the hard decision. And also the perfect power control is required along with increasing the number of user.