• Electronics and Telecommunications Trends
• /
• v.34 no.6
• /
• pp.87-99
• /
• 2019

Reinforcement learning is a type of machine learning paradigm that forces agents to repeat the observation-action-reward process to assess and predict the values of possible future action sequences. This allows the agents to incrementally reinforce the desired behavior for a given observation. Thanks to the recent advancements of deep learning, reinforcement learning has evolved into deep reinforcement learning that introduces promising results in various control and optimization domains, such as games, robotics, autonomous vehicles, computing, industrial control, and so on. In addition to this trend, a number of programming libraries have been developed for importing deep reinforcement learning into a variety of applications. In this article, we briefly review and summarize 10 representative deep reinforcement learning libraries and compare them from a development project perspective.

• Electronics and Telecommunications Trends
• /
• v.36 no.4
• /
• pp.61-71
• /
• 2021

Navigation satellite systems (GPS, GLONASS etc.) provide three main services, i.e., positioning for location based services, navigation for multi-modal transportation services, and timing for communication and critical infrastructure services. They were started as military systems but were extended to civil service. Navigation satellite navigation system began with GPS in the USA and GLONASS in Russia at nearly the same time. Indian NavIC and Chines BDS announced their FOCs in 2016 and 2020, respectively and European Galileo and Japanese QZSS are catching up others. In these days, Navigation Satellite System, Positioning, Navigation, and Timing services are part of our daily life very closely. They are required for autonomous driving car, Unmanned vehicles like UAV, UGV, and UMV, 5G/6G telecommunications, world financial system, power system, survey, agriculture, and so on. The services among navigation satellite systems are very competitive and also cooperative one another. This article describes the status of these systems and evolution in the technical and service senses, which may be helpful for planning korea positioning system(KPS).

• Journal of Drive and Control
• /
• v.17 no.4
• /
• pp.72-79
• /
• 2020

To ensure the safety of Advanced Driver Assistance Systems (ADAS) or autonomous vehicles, it is important to recognize the vehicle position, and specifically, the increased accuracy of the lateral position of the vehicle is required. In recent years, the quality of GPS signals has improved a lot and the price has decreased significantly, but extreme urban environments such as tunnels still pose a critical challenge. In this study, we proposed stable and precise lane recognition and tracking methods to solve these two issues via fusion of AVM images and vehicle sensor data using an extended Kalman filter. In addition, the vehicle's lateral position recognition and the abnormal state of RTK GPS were determined using this approach. The proposed method was validated via actual vehicle experiments in urban areas reporting multipath and signal disconnections.

• The Journal of Korea Robotics Society
• /
• v.17 no.2
• /
• pp.110-117
• /
• 2022

This paper describes a localization algorithm utilizing relative observations for multiple autonomous underwater vehicles (Multiple-AUVs). In order to maximize the efficiency of operation and mission accomplishment and to prevent problems such as collision and interference, the locations and directions of Multiple-AUVs must be precisely estimated. To estimate the locations and directions, we designed a localization algorithm utilizing relative observations and verified it with simulations based on sensor data sets acquired through real sea experiments. Also, an optimal combination of relative observation information for efficient localization is figured out through combining various relative observations. The proposed method shows improved localization results compared to those only using the navigation algorithm. The performance of localization is improved up to 58% depending on the combination of relative observations.

• Journal of the Korean Society of Industry Convergence
• /
• v.25 no.4_2
• /
• pp.645-653
• /
• 2022

Recently, deep reinforcement learning has been applied to a variety of industries, such as games, robotics, autonomous vehicles, and data cooling systems. An algorithm called reinforcement learning allows for automated asset allocation without the requirement for ongoing monitoring. It is free to choose its own policies. The purpose of this paper is to carry out an empirical analysis of the performance of asset allocation strategies. Among the strategies considered were the conventional Mean- Variance Optimization (MVO) and the Proximal Policy Optimization (PPO). According to the findings, the PPO outperformed both its benchmark index and the MVO. This paper demonstrates how dynamic asset allocation can benefit from the development of a reinforcement learning algorithm.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.330-331
• /
• 2022

자율주행 차량의 운전자는 현재 레벨3에서 탑승하는 운전자에부터 최종적으로 레벨 5단계에서 탑승자로 변화하게 된다. 관련하여 자율주행차량이 운행하는 동안 탑승자는 무엇을 하는지가 중요한 이슈로 대두될 여지가 있다. 탑승자는 뉴스를 읽거나 노래를 부르거나 주변 환경을 감상할 수 있고, 또는 탑승자는 다른 탑승자와 게임을 하거나 대화를 하거나 회의와 의사결정을 내릴 수도 있다. 자율주행차량은 이용자의 활용에 따라 오락공간, 휴식공간, 회의공간으로 트랜스포메이션되는 셈이다. 본 논문은 자율주행차량에서 블록체인 기술 중 하나인 NFT를 활용하여 차량의 탑승자에게 소유권이 있는 생산 데이터에 대해 스마트 계약을 구현하는 방법에 대하여 연구하였다. 자율주행 차량 내에서의 소유권을 표식한 스마트 계약 체결과 향후 적용 운용환경을 연구.개발하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.171-174
• /
• 2022

자동차의 전기차로의 전환과 더불어 IT 기술이 접목된 자율주행기술이 빠르게 발전하고 있으며, 현대기아자동차 그룹을 비롯한 글로벌 완성차 제조사 및 애플, LG 등과 같은 IT 제조사간에 완성도 높은 자율주행차 개발에 치열한 경쟁이 이루어지고 있다. 특히 사람의 개입이 거의 없는 4단계 이상의 자율주행 기술이 적용된 자율주행차는 운전석이 없는 등 기존과는 매우 다른 실내 구조를 가질 것으로 예상되며, 사람에게 움직이는 생활공간을 제공할 것으로 기대된다. 이와 같은 자율주행차 내의 미디어 소비는 고화질 미디어를 대형 화면으로 볼 것으로 예상되며, 지상파 TV 방송은 미디어를 단방향으로 대용량 전송하는데 유리하여 자율주행차를 대상으로 미디어 서비스에 적용성이 높을 것으로 기대된다. 본 논문은 이러한 미래 자율주행차가 제공하는 움직이는 생활공간에서 TV를 시청하는 '이동TV'를 염두에 두고, 현재의 4K-UHDTV 시대 및 미래의 8K/Post-8K 시대의 비용 효과적인 지상파 TV 방송망 기술을 소개한다.

• Journal of Advanced Information Technology and Convergence
• /
• v.9 no.2
• /
• pp.89-97
• /
• 2019

Recently, there is an increasing demand for applications utilizing maps and locations such as autonomous vehicles and location-based services. Since these applications are developed based on spatial data, interest in spatial data processing is increasing and various studies are being conducted. In this paper, I propose a parallel mining algorithm using the CUDA library to efficiently analyze large spatial data. Spatial data includes both geometric (spatial) and non-spatial (aspatial) attributes. The proposed parallel spatial data mining algorithm analyzes both the geometric and non-spatial relationships between two layers. The experiment was performed on graphics cards containing CUDA cores based on TIGER/Line data, which is the actual spatial data for the US census. Experimental results show that the proposed parallel algorithm using CUDA greatly improves spatial data mining performance.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.607-609
• /
• 2023

본 논문은 인공지능과 자율 주행차의 현재 상황과 향후 전망에 대해 조사한 결과를 제시한다. 자율 주행차의 기술적 발전과 인공지능의 개발이 상호보완적으로 진행되며, 운전의 안전성과 효율성을 향상시키는 가능성이 크다. 본 연구는 자율 주행차와 인공지능의 상호작용을 탐구하고, 향후 연구 및 개발 방향을 제안한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.116-118
• /
• 2023

자율주행차가 제공하는 새로운 시장과 경쟁력, 인력 및 시간 절약, 교통 체증 문제 해결 등의 장점을 다루고, UN 사이버 보안 법률에 따른 자율주행차의 기술적인 요구사항을 준수해야 한다. 하지만 자율주행차에 대한 기술적인 요구사항을 준수하는 것으로는 모든 사이버 공격에 대해서 막을 수 없다. 자율주행차의 법적 요구사항과 사이버 보안 위협에 대처하는 방법을 다룬다. 특히 RTOS(Real Time OS)와 같은 실시간 시스템에 매우 위험할 수 있는 DRAM(Dynamic Random Access Memory)에 대한 로우해머링 공격 기법에 대해 분석하고 로우해머링에 대한 보안 방법을 제시한다. 그리고 자율 주행 시스템의 안전과 신뢰성을 보장하기 위해 하드웨어 기반 또는 소프트웨어 기반 방어 기술을 소개하고 있다.