• Annual Conference of KIPS
• /
• 2022.05a
• /
• pp.110-113
• /
• 2022

인터넷을 통해 주위 사물과 연결된 차량은 사용자에게 편리성을 제공하기 위해 다양한 콘텐츠를 요구하는데 클라우드로부터 가져오는 시간이 비교적 오래 걸리기 때문에 차량과 물리적으로 가까운 위치에 캐싱하는 기법들이 등장하고 있다. 본 논문에서는 기반 시설이 밀집하게 설치된 도시 환경에서 maximum distance separable(MDS) 코딩을 사용해 road side unit(RSU)에 캐싱하는 방법에 대해 연구하였다. RSU의 중복된 서비스 커버리지 지역을 고려하여 차량의 콘텐츠 요구에 대한 RSU hit ratio를 높이기 위해 deep Q-learning(DQN)를 사용하였다. 실험 결과 비교 알고리즘보다 hit raito 측면에서 더 높은 성능을 보이는 것을 증명하였다.

• The Journal of Bigdata
• /
• v.8 no.2
• /
• pp.97-107
• /
• 2023

Small business owners are relatively likely to be alienated from various benefits caused by the change to a big data/AI-based society. To support them, we would like to detect a hot place based on the floating population to support small business owners' decision-making in the start-up area. Through various studies, it is known that the population size of the region has an important effect on the sales of small business owners. In this study, inland regions were extracted from the Incheon floating population data from January 2019 to June 2022. the Data is consisted of a grid of 50m intervals, central coordinates and the population for each grid are presented, made image structure through imputation to maintain spatial information. Spatial outliers were removed and imputated using LOF and GAM, and temporal outliers were removed and imputated through LOESS. We used ConvLSTM which can take both temporal and spatial characteristics into account as a predictive model, and used AutoEncoder structure, which performs outliers detection based on reconstruction error to define an area with high MAPE as a hot place.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.4B
• /
• pp.345-352
• /
• 2002

In this paper, in order to evaluate the performance of IS-95 system reverse link in white gaussian noise and rayleigh fading environment, we suggest epochal proposal to improve computer run-time and its efficiency is verified in terms of the number of samples. MC(Monte Carlo) simulation is the most popular simulation technique lately, but MC simulation requires a number of samples at low bit error rate. Therefore, MC cannot avoid the limit of computer run-time. To alleviate these problems, we apply the suggested method called central moment technique to the reverse link of the IS-95 system and can obtain discrete probability mass functions from Nth order central moments of the less number of received signal samples than those required in MC. Continuous cumulative probability distribution function can be accurately estimated by using interpolation and the improvement effect for the number of samples is proven.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.7 no.4
• /
• pp.234-237
• /
• 1974

Electro-chemical aspects of stress corrosion on the mild steels when immersed in the $5\%\;H_2SO_4$ solution and charged with 100mV and 100mA were discussed. The main results of the experiment are follows; 1. The weight loss by corrosion was concerned with the applied stress. and the larger the applied stress, the greater the weight loss. 2. Reduction of corrosion stress was a factor of inverse proportion to the applied stress. 3. Corrosion began at first on the parts of impurities concentrated and the grain boundaries, and gradually developed and spreaded out. 4. The materials of unsteady structure deformed of space lattice by the high stress or work-hardening showed less reduction of corrosion stress.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.112-113
• /
• 2021

하천으로 유입되는 오염물질은 유수의 흐름에 따라 이송되며 혼합된다. 이러한 오염물질의 해석을 위해서는 확산 또는 분산계수 산정이 필요하다. 오염물질의 거동과 관련된 실험적 연구는 방사선 동위원소와 형광성 물질을 이용하여 수행되어 왔으나, 추적자 실험은 많은 비용 및 인력을 요하며, 고정식으로 설치한 계측장비로부터 수집한 시계열 농도자료만을 이용하여 분석하기 때문에 공간적 분포에 대한 자료 취득은 어렵다는 한계가 있다. 하천의 오염물질을 모니터링하기 위해서는 공간을 이동하는 입자의 관점에서 물리량을 표현하는 Lagrangian 방식보다 특정 위치에서 물리량 변화를 표현하는 Eulerian 방식이 적합하다. 그러나 드론을 활용한 하천원격탐사 연구의 대부분은 이동식 플랫폼으로 활용되어 특정 시간에 공간적 분광특성의 분포 파악이 한정적이며, 동일 지점에서 분광특성의 시간적 변화를 파악할 수 없다. 따라서 본 연구에서는 기존의 면단위를 측정하는 이동식 플랫폼의 한계를 극복하고 하천 모니터링에 적합한 Eulerian 방식을 적용하기 위하여 CCTV 형태의 고정식 초분광촬영 플랫폼을 도입하였다. 본 연구에서는 하천으로 유입되는 오염물질의 거동을 분석하기 위하여 자연하천에서 형광성물질인 Rhodamine WT를 이용하여 추적자 실험을 수행하였으며, 접촉식 센서를 활용한 농도측정과 동시에 드론과 초분광센서를 활용하여 CCTV 형태의 고정식 초분광영상을 획득하였다. 실험결과 도출된 전통적인 방식의 분산계수 산정과 시공간 초분광영상을 활용한 분산계수 산정을 비교하여 오염물질 거동 분석에 초분광영상 활용의 가능성을 검토하였다. 본 연구에서 제시한 드론기반 시공간 초분광영상 기법을 교량이나 기타 하천구조물에 초분광 센서를 설치하여 CCTV형식으로 활용할 경우, 공단이나 하·폐수 처리장 등의 점오염원이 밀집해 있는 지역에 직접 설치하여 화학사고의 감지 및 오염물질의 유출 확인 및 조류, 부유사 등의 다양한 수질항목의 농도 변화 감지가 가능하고, 수심변화 감지로 장기적으로 활용할 경우 특정 지점에서의 하상변동 조사가 가능하다. 또한, 오염물질의 유출 사고 발생 등의 사람이 직접 접근이 불가능한 지역에 드론을 활용하여 초분광센서를 이용한 오염물질 감지가 가능할 것으로 판단된다.

• Fire Science and Engineering
• /
• v.32 no.4
• /
• pp.122-130
• /
• 2018

Recent urbanization, population densification, and the impact of global climate change are causing disasters to become larger and more complex. Meanwhile, in Korea, there is an emphasis on preventing, restoring, and recovering from disasters. However, disaster medical care, which is absolutely necessary to maintain life in a disaster, is being ignored. Therefore, in this study, we selected Seoul as the study area where flood damage is frequent and underground housing is densely populated. Assuming underground housing in the immersion history area as the emergency patient site, transfer distance and transfer time were analyzed. This study considered both accessibility to emergency medical facilities and disaster sites and accessibility from emergency services to disaster sites. Therefore, it seems to be meaningful as basic data for the improvement of emergency medical services.

• Journal of the Korean association of regional geographers
• /
• v.9 no.3
• /
• pp.262-275
• /
• 2003

In this study, the characteristics of locational distribution of the eminent people show their socially and economically stabilized standing in Daegu, taking the aforesaid situation as a background of the study, and the process of this study is as follows ; to examine the forming process of residential areas in the city as a theoretical supporting, to put in order on classical interpretation to formation of residential areas, and general type modem residential areas formation, and economic decision factor of land use. Therefore, this study aims to examine growth and development of eminent persons' residential areas and, at the same time, extract locational characteristics through the pattern of eminent persons' location and predict changes in the future. Furthermore, a long-term development plan of Daegu is expected to contribute to completing formation of sub-centers as well as expanding metropolis areas. Recent efforts by the city government such as the ones to attract prestigious high schools in the upper city areas are also expected to bring out balanced development of the city.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.28 no.4D
• /
• pp.459-467
• /
• 2008

Because of the insufficiency of ground space, the utilization of underground is getting more and more in these days. Moreover, underground space is being used not only buildings but multipurpose space for movement, storage and shopping. However, ground space has vital weakness for fire compared to ground space. Especially in case of underground shopping center, there are various stuffs to burn and poisonous gas can be exposed on this count when the space is burned. A large number of casualties can be also occurred from conflagration as underground space has closed structures that prevent rapid evacuation and access. Therefore, this research proposes the guidance methodology for evacuation from conflagration in large-sized underground facilities. In addition, suggested methodology uses high technology wireless sensor information from up-to-date ubiquitous sensor networks. Fire information collected by sensors is integrated with existing underground facilities information and this is sent to guidance systems by inducing process. In the end, this information is used for minimum time paths finding algorithm considering the passageway capacity and distance. Also, usefulness and inadequacies of proposed methodology is verified by a case study.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.355-364
• /
• 2023

As advanced cyber threats continue to increase in recent years, it is difficult to detect new types of cyber attacks with existing pattern or signature-based intrusion detection method. Therefore, research on anomaly detection methods using data learning-based artificial intelligence technology is increasing. In addition, supervised learning-based anomaly detection methods are difficult to use in real environments because they require sufficient labeled data for learning. Research on an unsupervised learning-based method that learns from normal data and detects an anomaly by finding a pattern in the data itself has been actively conducted. Therefore, this study aims to extract a latent vector that preserves useful sequence information from sequence log data and develop an anomaly detection learning model using the extracted latent vector. Word2Vec was used to create a dense vector representation corresponding to the characteristics of each sequence, and an unsupervised autoencoder was developed to extract latent vectors from sequence data expressed as dense vectors. The developed autoencoder model is a recurrent neural network GRU (Gated Recurrent Unit) based denoising autoencoder suitable for sequence data, a one-dimensional convolutional neural network-based autoencoder to solve the limited short-term memory problem that GRU can have, and an autoencoder combining GRU and one-dimensional convolution was used. The data used in the experiment is time-series-based NGIDS (Next Generation IDS Dataset) data, and as a result of the experiment, an autoencoder that combines GRU and one-dimensional convolution is better than a model using a GRU-based autoencoder or a one-dimensional convolution-based autoencoder. It was efficient in terms of learning time for extracting useful latent patterns from training data, and showed stable performance with smaller fluctuations in anomaly detection performance.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.06a
• /
• pp.5-8
• /
• 2016

키 프레임 추출 기법은 2차원 비오 영상을 3차원으로 재구성하기 위해 꼭 필요한 프레임을 선택하는 방법이다. 본 논문에서는 비디오에서 빠르게 프레임을 검사하며 최적의 키 프레임을 선택하는 기법을 제안한다. 제안하는 기법은 3차원 재구성을 위한 전처리 과정에 초점을 둔 것으로 프레임 간 대응점 비율 검사를 통해 프레임의 도약 강도를 결정하고 기하 모델 추정이 원활한 프레임을 선택한다. 이로부터 3차원 복원 후처리 과정을 통해 최종적인 3차원 점군(point cloud) 데이터를 획득한다. 실험을 통해 다른 기법과 성능을 비교했을 때, 제안하는 기법이 복원 소요 시간도 적게 들고 보다 밀집된 3차원 데이터를 얻을 수 있었다.