• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.1
• /
• pp.20-34
• /
• 2021

Today, IoT devices are flooding, and traffic is increasing rapidly. The Internet of Things creates a variety of added value through connections between devices, while many devices are easily targeted by attackers due to security vulnerabilities. In the IoT environment, security diagnosis has problems such as having to provide different solutions for different types of devices in network situations where various types of devices are interlocked, personal leakage of security solutions themselves, and high cost, etc. To avoid such problems, a TCP-based active scan was presented. However, the TCP-based active scan has limitations that it is difficult to be applied to real-time systems due to long detection times. To complement this, this study uses UDP-based approaches. Specifically, a lightweight active scan algorithm that effectively identifies devices using UPnP protocols (SSDP, MDNS, and MBNS) that are most commonly used by manufacturers is proposed. The experimental results of this study have shown that devices can be distinguished by more than twice the true positive and recall at an average time of 1524 times faster than Nmap, which has a firm position in the field.

• Journal of Information Processing Systems
• /
• v.18 no.1
• /
• pp.1-11
• /
• 2022

Online video telephony systems have been increasingly used in several industrial areas because of coronavirus disease 2019 (COVID-19) spread. The existing session initiation protocol (SIP)-based video call system is being usefully utilized, however, there is a limitation that it is very inconvenient for users to transmit additional information during conversation to the other party in real time. To overcome this problem, an enhanced scheme is presented based on augmented real-time reality (AR). In this scheme, augmented information is automatically searched from the Internet and displayed on the user's device during video telephony. The proposed approach was qualitatively evaluated by comparing it with other conferencing systems. Furthermore, to evaluate the feasibility of the approach, we implemented a simple network application that can generate SIP call requests and answer with AR object pre-fetching. Using this application, the call setup time was measured and compared between the original SIP and pre-fetching schemes. The advantage of this approach is that it can increase the convenience of a user's mobile phone by providing a way to automatically deliver the required text or images to the receiving side.

• Journal of Information Processing Systems
• /
• v.18 no.2
• /
• pp.245-257
• /
• 2022

In the Internet of Things (IoT) era, the types of devices used by one user are becoming more diverse and the number of devices is also increasing. However, a forensic investigator is restricted to exploit or collect all the user's devices; there are legal issues (e.g., privacy, jurisdiction) and technical issues (e.g., computing resources, the increase in storage capacity). Therefore, in the digital forensics field, it has been a challenge to acquire information that remains on the devices that could not be collected, by analyzing the seized devices. In this study, we focus on the fact that multiple devices share data through account synchronization of the online platform. We propose a novel way of identifying the user's interest through analyzing the remnants of targeted advertising which is provided based on the visited websites or search terms of logged-in users. We introduce a detailed methodology to pick out the targeted advertising from cache data and infer the user's interest using deep learning. In this process, an improved learning model considering the unique characteristics of advertisement is implemented. The experimental result demonstrates that the proposed method can effectively identify the user interest even though only one device is examined.

• Nuclear Engineering and Technology
• /
• v.53 no.5
• /
• pp.1483-1490
• /
• 2021

Many studies are undertaken into nuclear power plants (NPPs) in preparation for accidents exceeding design standards. In this paper, we analyze the applicability of various wireless communication technologies as accident countermeasures in different NPP environments. In particular, a commercial wireless communication module (WCM) is investigated by measuring leakage current and packet error rate (PER), which vary depending on the intensity of incident radiation on the module, by testing at a Co-60 gamma-ray irradiation facility. The experimental results show that the WCMs continued to operate after total doses of 940 and 1097 Gy, with PERs of 3.6% and 0.8%, when exposed to irradiation dose rates of 185 and 486 Gy/h, respectively. In short, the lower irradiation dose rate decreased the performance of WCMs more than the higher dose rate. In experiments comparing the two communication protocols of request/response and one-way, the WCMs survived up to 997 and 1177 Gy, with PERs of 2% and 0%, respectively. Since the request/response protocol uses both the transmitter and the receiver, while the one-way protocol uses only the transmitter, then the electronic system on the side of the receiver is more vulnerable to radiation effects. From our experiments, the tested module is expected to be used for design-based accidents (DBAs) of "Category A" type, and has confirmed the possibility of using wireless communication systems in NPPs.

• The Journal of Korea Robotics Society
• /
• v.17 no.3
• /
• pp.264-272
• /
• 2022

Human-robot cooperative tasks are increasingly required in our daily life with the development of robotics and artificial intelligence technology. Interactive reinforcement learning strategies suggest that robots learn task by receiving feedback from an experienced human trainer during a training process. However, most of the previous studies on Interactive reinforcement learning have required an extra feedback input device such as a mouse or keyboard in addition to robot itself, and the scenario where a robot can interactively learn a task with human have been also limited to virtual environment. To solve these limitations, this paper studies training strategies of robot that learn table balancing tasks interactively using deep reinforcement learning with human's facial expression feedback. In the proposed system, the robot learns a cooperative table balancing task using Deep Q-Network (DQN), which is a deep reinforcement learning technique, with human facial emotion expression feedback. As a result of the experiment, the proposed system achieved a high optimal policy convergence rate of up to 83.3% in training and successful assumption rate of up to 91.6% in testing, showing improved performance compared to the model without human facial expression feedback.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.2
• /
• pp.559-575
• /
• 2023

3D human pose estimation is widely applied in various fields, including action recognition, sports analysis, and human-computer interaction. 3D human pose estimation has achieved significant progress with the introduction of convolutional neural network (CNN). Recently, several researches have proposed the use of multiview approaches to avoid occlusions in single-view approaches. However, as the number of cameras increases, a 3D pose estimation system relying on a CNN may lack in computational resources. In addition, when a single host system uses multiple cameras, the data transition speed becomes inadequate owing to bandwidth limitations. To address this problem, we propose a distributed real-time 3D pose estimation framework based on asynchronous multiple cameras. The proposed framework comprises a central server and multiple edge devices. Each multiple-edge device estimates a 2D human pose from its view and sendsit to the central server. Subsequently, the central server synchronizes the received 2D human pose data based on the timestamps. Finally, the central server reconstructs a 3D human pose using geometrical triangulation. We demonstrate that the proposed framework increases the percentage of detected joints and successfully estimates 3D human poses in real-time.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.11a
• /
• pp.21-22
• /
• 2009

유비쿼터스 사회(Ubiquitous Society)로의 진입에 발맞추어 USN(Ubiquitous Sensor Network) 기반의 인간 중심적 편의 시스템에 관한 연구가 활발하게 진행되고 있다. 그 중 대형 시장을 갖는 차량 편의시설에 관한 연구는 지능형 차량 시스템(Intelligent Car System)을 중심으로 활발히 이루어지고 있다. 지능형 차량 시스템의 주요 연구는 자동 항법 장치, 사고 예방 장치, 자가 진단 시스템 등 탑승자의 편의성과 안전성을 중심으로 진행되었다. 그러나 탑승자의 사고 발생 시 응급 상황 처리를 위한 지원 시스템은 미미하다. 탑승자 부상으로 사고 신고를 하지 못할 경우, 사고지점 확인, 탑승자의 위급(현재) 상황, 부상 정보와 같은 정보를 얻을 수 없어 응급 상황 대처에 신속하지 못할 수 있다. 따라서 본 논문은 다양한 센서를 이용하여 차량의 정보를 수집하고, 사고 판단 시 차량 위치 정보, 탑승자 상황 정보를 응급 기관에 전달할 수 있는 차량용 스마트 안전장치를 설계 및 구현하였다. 테스트를 위해 Intel PXA255 MCU와 AM-3AXIS(3축 가속 센서), MDSM-1000A(지자기 센서), RX-M800S CDMA, GPS520, Alpha cam, Flex Sensor로 시스템을 제작하였으며 모의 도로 모형에서 테스트 하였다.

• Telecommunications review
• /
• v.24 no.4
• /
• pp.468-480
• /
• 2014

The current-generation avionics systems are based on a federated architecture, where an electronic device runs a single software module or application that collaborates with other devices through a network. This architecture makes the internal system architecture very complicate, and gives rise to issues of Size, Weight, and Power (SWaP). In this paper, we show that the partitioning defined by ARINC 653 can efficiently deal with the SWaP issues on small unmanned aerial vehicles, where the SWaP issues are extremely severe. We especially install the integrated mission system on real hexacopter and quadcopter and perform successful flight tests. The presented software technology for integrated mission system and software consolidation methodology can provide a valuable reference for other SWaP sensitive real-time systems.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2011.11a
• /
• pp.718-721
• /
• 2011

웹 서비스를 대상으로 하는 DDoS(Distributed Denial-of-Service) 공격 또는 유해 트래픽 유입을 탐지 또는 차단하기 위한 목적으로 HTTP(Hypertext Transfer Protocol) 트래픽을 실시간으로 분석하는 기능은 거의 모든 네트워크 트래픽 보안 솔루션들이 탑재하고 있는 필수적인 요소이다. 하지만, HTTP 트래픽의 실시간 데이터 측정 양이 시간이 지날수록 기하급수적으로 증가함에 따라, HTTP 트래픽을 실시간 패킷 단위로 분석한다는 것에 대한 성능 부담감은 날로 커지고 있는 실정이다. 이제는 응용 어플리케이션 차원에서는 성능에 대한 부담감을 해소할 수 없기 때문에 고비용의 소프트웨어 가속기나 하드웨어에 의존적인 전용 장비를 탑재하여 해결하려는 시도가 대부분이다. 본 논문에서는 현재 대부분의 PC 에 탑재되어 있는 그래픽 카드의 GPU(Graphics Processing Units)를 범용적으로 활용하고자 하는 GPGPU(General-Purpose computation on Graphics Processing Units)의 연구에 힘입어, NVIDIA사의 CUDA(Compute Unified Device Architecture)를 사용하여 네트워크 트래픽에서 HTTP 패킷 추출성능을 응용 어플리케이션 차원에서 향상시켜 보고자 하였다. HTTP 패킷 추출 연산만을 기준으로 GPU 의 연산속도는 CPU 에 비해 10 배 이상의 높은 성능을 얻을 수 있었다.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.36 no.2
• /
• pp.175-185
• /
• 2023

A hybrid supercapacitor is a promising energy storage device in view of its excellent capacitive performance. Commercial three-dimensional foam nickel (Ni) can be used as an ideal framework due to an interconnected network structure. However, its application as an electrode material for supercapacitors is limited due to its low specific capacity. Herein, we report a successful growth of MnO₂ on the surface of graphene by a one-step hydrothermal method; thus, forming a three-dimensional MnO₂-graphene-Ni hybrid foam. Our results show that the mixed structure of MnO₂ with nanoflowers and nanorods grown on the graphene/Ni foam as a hybrid electrode delivers the maximum specific capacitance of 193 F·g^-1 at a current density 0.1 A·g^-1. More importantly, the hybrid electrode retains 104% of its initial capacitance after 1,000 charge-discharge cycles at 1 A·g^-1; thus, showing the potential application as a stable supercapacitor electrode.