• Title/Summary/Keyword: Training simulation

Search Result 1,362, Processing Time 0.025 seconds

Active pulse classification algorithm using convolutional neural networks (콘볼루션 신경회로망을 이용한 능동펄스 식별 알고리즘)

  • Kim, Geunhwan;Choi, Seung-Ryul;Yoon, Kyung-Sik;Lee, Kyun-Kyung;Lee, Donghwa
    • The Journal of the Acoustical Society of Korea
    • /
    • v.38 no.1
    • /
    • pp.106-113
    • /
    • 2019
  • In this paper, we propose an algorithm to classify the received active pulse when the active sonar system is operated as a non-cooperative mode. The proposed algorithm uses CNN (Convolutional Neural Networks) which shows good performance in various fields. As an input of CNN, time frequency analysis data which performs STFT (Short Time Fourier Transform) of the received signal is used. The CNN used in this paper consists of two convolution and pulling layers. We designed a database based neural network and a pulse feature based neural network according to the output layer design. To verify the performance of the algorithm, the data of 3110 CW (Continuous Wave) pulses and LFM (Linear Frequency Modulated) pulses received from the actual ocean were processed to construct training data and test data. As a result of simulation, the database based neural network showed 99.9 % accuracy and the feature based neural network showed about 96 % accuracy when allowing 2 pixel error.

The Possibility of Neural Network Approach to Solve Singular Perturbed Problems

  • Kim, Jee-Hyun;Cho, Young-Im
    • Journal of the Korea Society of Computer and Information
    • /
    • v.26 no.1
    • /
    • pp.69-76
    • /
    • 2021
  • Recentlly neural network approach for solving a singular perturbed integro-differential boundary value problem have been researched. Especially the model of the feed-forward neural network to be trained by the back propagation algorithm with various learning algorithms were theoretically substantiated, and neural network models such as deep learning, transfer learning, federated learning are very rapidly evolving. The purpose of this paper is to study the approaching method for developing a neural network model with high accuracy and speed for solving singular perturbed problem along with asymptotic methods. In this paper, we propose a method that the simulation for the difference between result value of singular perturbed problem and unperturbed problem by using neural network approach equation. Also, we showed the efficiency of the neural network approach. As a result, the contribution of this paper is to show the possibility of simple neural network approach for singular perturbed problem solution efficiently.

Evaluation of ATM usability test for improving financial life of Impaired elderly (인지저하 노인들의 금융생활 라이프 향상을 위한 ATM 사용성 평가)

  • Choi, Yoo-jung;Choi, Hun
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.24 no.1
    • /
    • pp.77-82
    • /
    • 2020
  • As Korea enters an aging age, social efforts to improve the IADL of the elderly are increasing. In this study, to improve the performance of financial management activities that the elderly is particularly burdened, we aim to learn the elderly through ATM simulation education contents so that they can use ATM smoothly. To this end, interviews were conducted with seniors to derive four major financial activities (deposits, withdrawals, deposit inquiries and bank account arrangements), and developed tablet PC-based ATM education contents identical to the existing bank ATM interfaces. The experiment was conducted on 20 elderly people in the Elderly Day Care Center, and their satisfaction, fatigue and performance were measured before and after education. The results of this study can provide ATM design guidelines for the elderly who have difficulty using ATM.

Resource Allocation for Heterogeneous Service in Green Mobile Edge Networks Using Deep Reinforcement Learning

  • Sun, Si-yuan;Zheng, Ying;Zhou, Jun-hua;Weng, Jiu-xing;Wei, Yi-fei;Wang, Xiao-jun
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.15 no.7
    • /
    • pp.2496-2512
    • /
    • 2021
  • The requirements for powerful computing capability, high capacity, low latency and low energy consumption of emerging services, pose severe challenges to the fifth-generation (5G) network. As a promising paradigm, mobile edge networks can provide services in proximity to users by deploying computing components and cache at the edge, which can effectively decrease service delay. However, the coexistence of heterogeneous services and the sharing of limited resources lead to the competition between various services for multiple resources. This paper considers two typical heterogeneous services: computing services and content delivery services, in order to properly configure resources, it is crucial to develop an effective offloading and caching strategies. Considering the high energy consumption of 5G base stations, this paper considers the hybrid energy supply model of traditional power grid and green energy. Therefore, it is necessary to design a reasonable association mechanism which can allocate more service load to base stations rich in green energy to improve the utilization of green energy. This paper formed the joint optimization problem of computing offloading, caching and resource allocation for heterogeneous services with the objective of minimizing the on-grid power consumption under the constraints of limited resources and QoS guarantee. Since the joint optimization problem is a mixed integer nonlinear programming problem that is impossible to solve, this paper uses deep reinforcement learning method to learn the optimal strategy through a lot of training. Extensive simulation experiments show that compared with other schemes, the proposed scheme can allocate resources to heterogeneous service according to the green energy distribution which can effectively reduce the traditional energy consumption.

SAD : Web Session Anomaly Detection based on Bayesian Estimation (베이지언 추정을 이용한 웹 서비스 공격 탐지)

  • 조상현;김한성;이병희;차성덕
    • Journal of the Korea Institute of Information Security & Cryptology
    • /
    • v.13 no.2
    • /
    • pp.115-125
    • /
    • 2003
  • As Web services are generally open for external uses and not filtered by Firewall, these result in attacker's target. Web attacks which exploit vulnerable web-applications and malicious users' requests cause economical and social problems. In this paper, we are modelling general web service usages based on user-web-session and detect anomal usages with Bayesian estimation method. Finally we propose SAD(Session Anomaly Detection) for detection unknown web attacks. To evaluate SAD, we made an experiment on attack simulation with web vulnerability scanner, whisker. The results show that the detection rate of SAD is over 90%, which is influenced by several features such as size of window or training set, detection filter method and web topology.

Types of Virtual Reality-based Safety Education Contents (가상현실 기반 안전교육 콘텐츠 유형 연구)

  • Chang, Sun-Hee;Chang, Hyo-Jin;Kim, Sung-Hoon
    • The Journal of the Korea Contents Association
    • /
    • v.21 no.1
    • /
    • pp.434-445
    • /
    • 2021
  • With the development of realistic content technology and related infrastructure, interest in the use of virtual reality content is growing. Accordingly, in the field of safety education, more and more cases of producing virtual reality-based safety education (VR safety education) contents such as experiencing disaster situations realistically by supplementing the shortcomings of existing lecture-style education. This study looked at the characteristics and effects of VR safety education compared to the existing safety education, and analyzed 104 VR safety education contents produced and disclosed to date into nine classifications based on content and form. Based on the degree of relevance between items and the two axes of 'interactionability' and 'vividness of the environment', VR safety education contents could be categorized into three types: tangible lecture type, simulation type, and game type. Through this study, we hope to contribute to the planning and production of quality VR safety education contents by considering the purpose of safety education and the characteristics of types with the expected effects.

A Study on the Circuit Design Methodology and Performance Evaluation for Hybrid Gate Driver (하이브리드 게이트 드라이버를 위한 회로 디자인 방법과 성능 평가에 관한 연구)

  • Cho, Geunho
    • Journal of IKEEE
    • /
    • v.25 no.2
    • /
    • pp.381-387
    • /
    • 2021
  • As Head-Mounted Displays(HMDs), which are mainly used to maximize realism in games and videos, have experienced increased demand and expanded scope of use in education and training, there is growing interest in methods to enhance the performance of conventional HMDs. In this study, a methodology to utilize Carbon NanoTubes(CNTs) to improve the performance of gate drivers that send control signals to each pixel circuit of the HMD is discussed. This paper proposes a new circuit design method that replaces the transistors constituting the buffer part of the conventional gate driver with transistors incorporating CNTs and compare the performance of the suggested gate drive with that of a gate driver comprising only conventional transistors via simulations. According to the simulation results, by including CNTs in the gate driver, the output voltage can be increased by approximately 0.3V compared to the conventional gate driver high voltage(1.1V) at a speed of 12.5 GHz and the gate width also can be reduced by up to 20 times.

Performance Analysis of Wireless Communication Systems Using Deep Learning Based Transmit Power Control in Nakagami Fading Channels (나카가미 페이딩 채널에서 딥러닝 기반 송신 전력 제어 기법을 이용하는 무선통신 시스템에 대한 성능 분석)

  • Kim, Donghyeon;Kim, Dongyon;Lee, In-Ho
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.24 no.6
    • /
    • pp.744-750
    • /
    • 2020
  • In this paper, we propose a deep learning based transmit power control (TPC) scheme to improve the spectral and energy efficiency of wireless communication systems. In the wireless communication system, the positions of multiple transceivers follow a uniform distribution, and the performances of spectral and energy efficiency for the proposed TPC scheme are analyzed assuming the Nakagami fading channels. The proposed TPC scheme uses batch normalization to improve spectral and energy efficiency in deep learning based training. Through simulation, we compare the results of the spectral and energy efficiency of the proposed TPC scheme and the conventional one for various area sizes that limit the position range of the transceivers and Nakagami fading factors. Comparing the performance results, we verify that the proposed scheme provides better performance than the conventional one.

Deep Learning based Frame Synchronization Using Convolutional Neural Network (합성곱 신경망을 이용한 딥러닝 기반의 프레임 동기 기법)

  • Lee, Eui-Soo;Jeong, Eui-Rim
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.24 no.4
    • /
    • pp.501-507
    • /
    • 2020
  • This paper proposes a new frame synchronization technique based on convolutional neural network (CNN). The conventional frame synchronizers usually find the matching instance through correlation between the received signal and the preamble. The proposed method converts the 1-dimensional correlator ouput into a 2-dimensional matrix. The 2-dimensional matrix is input to a convolutional neural network, and the convolutional neural network finds the frame arrival time. Specifically, in additive white gaussian noise (AWGN) environments, the received signals are generated with random arrival times and they are used for training data of the CNN. Through computer simulation, the false detection probabilities in various signal-to-noise ratios are investigated and compared between the proposed CNN-based technique and the conventional one. According to the results, the proposed technique shows 2dB better performance than the conventional method.

Mapless Navigation Based on DQN Considering Moving Obstacles, and Training Time Reduction Algorithm (이동 장애물을 고려한 DQN 기반의 Mapless Navigation 및 학습 시간 단축 알고리즘)

  • Yoon, Beomjin;Yoo, Seungryeol
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.25 no.3
    • /
    • pp.377-383
    • /
    • 2021
  • Recently, in accordance with the 4th industrial revolution, The use of autonomous mobile robots for flexible logistics transfer is increasing in factories, the warehouses and the service areas, etc. In large factories, many manual work is required to use Simultaneous Localization and Mapping(SLAM), so the need for the improved mobile robot autonomous driving is emerging. Accordingly, in this paper, an algorithm for mapless navigation that travels in an optimal path avoiding fixed or moving obstacles is proposed. For mapless navigation, the robot is trained to avoid fixed or moving obstacles through Deep Q Network (DQN) and accuracy 90% and 93% are obtained for two types of obstacle avoidance, respectively. In addition, DQN requires a lot of learning time to meet the required performance before use. To shorten this, the target size change algorithm is proposed and confirmed the reduced learning time and performance of obstacle avoidance through simulation.