• ETRI Journal
• /
• v.44 no.2
• /
• pp.220-231
• /
• 2022

Building smart transportation services in urban cities has become a worldwide problem owing to the rapidly increasing global population and the development of Internet-of-Things applications. Traffic congestion and environmental concerns can be alleviated by sharing mobility, which reduces the number of vehicles on the road network. The taxi-parcel sharing problem has been considered as an efficient planning model for people and goods flows. In this paper, we enhance the functionality of a current people-parcel taxi sharing model. The adapted model analyzes the historical request data and predicts the current service demands. We then propose two novel online routing algorithms that construct optimal routes in real-time. The objectives are to maximize (as far as possible) both the parcel delivery requests and ride requests while minimizing the idle time and travel distance of the taxis. The proposed online routing algorithms are evaluated on instances adapted from real Cabspotting datasets. After implementing our routing algorithms, the total idle travel distance per day was 9.64% to 12.76% lower than that of the existing taxi-parcel sharing method. Our online routing algorithms can be incorporated into an efficient smart shared taxi system.

• Journal of Internet Computing and Services
• /
• v.23 no.2
• /
• pp.1-7
• /
• 2022

Machine learning (ML) algorithms have been intended to seamlessly collaborate for enabling intelligent networking in terms of massive service differentiation, prediction, and provides high-accuracy recommendation systems. Mobile edge computing (MEC) servers are located close to the edge networks to overcome the responsibility for massive requests from user devices and perform local service offloading. Moreover, there are required lightweight methods for handling real-time Internet of Things (IoT) communication perspectives, especially for ultra-reliable low-latency communication (URLLC) and optimal resource utilization. To overcome the abovementioned issues, this paper proposed an intelligent scheme for traffic steering based on the integration of MEC and lightweight ML, namely support vector machine (SVM) for effectively routing for lightweight and resource constraint networks. The scheme provides dynamic resource handling for the real-time IoT user systems based on the awareness of obvious network statues. The system evaluations were conducted by utillizing computer software simulations, and the proposed approach is remarkably outperformed the conventional schemes in terms of significant QoS metrics, including communication latency, reliability, and communication throughput.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.142-142
• /
• 2021

An arbitrary representation of an urban drainage sewer system was devised using a geographic information system (GIS) tool in order to calculate the surface and subsurface flow interaction for simulating urban flood. The proposed methodology is a mean to supplement the unavailability of systematized drainage system using high-resolution digital elevation(DEM) data in under-developed countries. A modified DEM was also developed to represent the flood propagation through buildings and road system from digital surface models (DSM) and barely visible streams in digital terrain models (DTM). The manhole, sewer pipe and storm drain parameters are obtained through field validation and followed the guidelines from the Plumbing law of the Philippines. The flow discharge from surface to the devised sewer pipes through the storm drains are calculated. The resulting flood simulation using the modified DEM was validated using the observed flood inundation during a rainfall event. The proposed methodology for constructing a hypothetical drainage system allows parameter adjustments such as size, elevation, location, slope, etc. which permits the flood depth prediction for variable factors the Plumbing law. The research can therefore be employed to simulate urban flood forecasts that can be utilized from traffic advisories to early warning procedures during extreme rainfall events.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.144-146
• /
• 2022

차량 환경에서 발생하는 계산 집약적인 태스크가 증가하면서 모바일 엣지 컴퓨팅(MEC, Mobile Edge Computing)의 필요성이 높아지고 있다. 하지만 지상에 존재하는 MEC 서버는 출퇴근 시간과 같이 태스크가 일시적으로 급증하는 상황에 유동적으로 대처할 수 없으며, 이러한 상황을 대비하기 위해 지상 MEC 서버를 추가로 설치하는 것은 자원의 낭비를 불러온다. 최근 이 문제를 해결하기 위해 UAV(Unmanned Aerial Vehicle)기반 MEC 서버를 추가로 사용해 엣지 서비스를 제공하는 연구가 진행되고 있다. 그러나 UAV MEC 서버는 지상 MEC 서버와 달리 한정적인 배터리 용량으로 인해 서버 간 로드밸런싱을 통해 에너지 사용량을 최소화 하는 것이 필요하다. 본 논문에서는 UAV MEC 서버의 에너지 사용량을 고려한 마이그레이션 기법을 제안한다. 또한 GRU(Gated Recurrent Unit) 모델을 활용한 트래픽 예측을 바탕으로 한 마이그레이션을 통해 지연시간을 최소화할 수 있도록 한다. 제안 시스템의 성능을 평가하기 위해 MEC의 마이그레이션 시점을 결정하는 기준점와 차량의 밀도에 따라 실험을 진행하고, 서버의 로드 편차, UAV MEC 서버의 에너지 사용량 그리고 평균 지연 시간 측면에서 성능을 분석한다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.4
• /
• pp.86-98
• /
• 2018

Since 2015, the Korea Expressway Corporation has provided predicted travel time information, which is reproduced from DSRC systems over the extended expressway network in Korea. When it is open for public information, it helps travelers decide optimal routes while minimizing traffic congestions and travel cost. Although, sutiable evaluations to investigate the reliability of travel time forecast information have not been conducted so far. First of all, this study seeks to find out a measure of effectiveness to evaluate the reliability of travel time forecast via various literatures. Secondly, using the performance measurement, this study evaluates concurrent travel time forecast information in highway quantitatively and examines the forecast error by exploratory data analysis. It appears that most of highway lines provided reliable forecast information. However, we found significant over/under-forecast on a few links within several long lines and it turns out that such minor errors reduce overall reliability in travel time forecast of the corresponding highway lines. This study would help to build a priority for quality control of the travel time forecast information system, and highlight the importance of performing periodic and sustainable management for travel time forecast information.

• Journal of Korean Society of Transportation
• /
• v.17 no.1
• /
• pp.173-185
• /
• 1999

The purpose of this study is to develope a forecasting model to implement short-term Congestion Management Program (CMP) based on TDM strategies in Seoul. The CMP is composed of three elements: 1) setting a goal of short-term traffic management. 2) developing a model to forecast the impacts of TDM alternatives, and 3) finding TDM measures to achieve the goal To Predict the impacts of TDM alternatives, a model called SECOMM (SEoul COngestion Management Model) is developed. The model assumes that trip generation and distribution are not changing in a short term, and that only mode split and traffic assignment are affected by TDM. The model includes the parameter values calibrated by a discrete mode choice model, and roadway and transit networks with 1,020 zones. As a TDM measure implement, it affects mode choice behavior first and then the speeds of roadway network. The chanced speed again affects the mode choice behavior and the roadway speeds. These steps continue until the network is equilibrated. The study recommends that CMP be introduced in Seoul, and that road way conditions be monitored regularly to secure the prediction accuracy of SECOMM. Also, TDM should be the major Policy tools in removing short-term congestion problems in a big city.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.49 no.7
• /
• pp.378-388
• /
• 2000

In this paper, we propose Fuzzy Polynomial Neural Networks(FPNN) based on Polynomial Neural Networks(PNN) and Fuzzy Neural Networks(FNN) for model identification of complex and nonlinear systems. The proposed FPNN is generated from the mutually combined structure of both FNN and PNN. The one and the other are considered as the premise part and consequence part of FPNN structure respectively. As the consequence part of FPNN, PNN is based on Group Method of Data Handling(GMDH) method and its structure is similar to Neural Networks. But the structure of PNN is not fixed like in conventional Neural Networks and self-organizing networks that can be generated. FPNN is available effectively for multi-input variables and high-order polynomial according to the combination of FNN with PNN. Accordingly it is possible to consider the nonlinearity characteristics of process and to get better output performance with superb predictive ability. As the premise part of FPNN, FNN uses both the simplified fuzzy inference as fuzzy inference method and error back-propagation algorithm as learning rule. The parameters such as parameters of membership functions, learning rates and momentum coefficients are adjusted using genetic algorithms. And we use two kinds of FNN structure according to the division method of fuzzy space of input variables. One is basic FNN structure and uses fuzzy input space divided by each separated input variable, the other is modified FNN structure and uses fuzzy input space divided by mutually combined input variables. In order to evaluate the performance of proposed models, we use the nonlinear function and traffic route choice process. The results show that the proposed FPNN can produce the model with higher accuracy and more robustness than any other method presented previously. And also performance index related to the approximation and prediction capabilities of model is evaluated and discussed.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.38 no.1
• /
• pp.1-13
• /
• 2001

In this paper, we propose a new dynamic bandwidth allocation and call admission control for VBR video sources with QoS constraints to support an efficient resource management and at the same Lime to satisfy the user's quality or service requirements. For the dynamic bandwidth allocation, first the next amount of traffic is predicted using a modified adaptive linear prediction method that considers abrupt scene change effects. And then, we dynamically allocate the necessary bandwidth to each connection based on the currently provided quality degree by the network with respect to the user's QoS requirements in terms of average delay and loss ratio. For the admission control, we determine the acceptance or rejection or a new connection based on the quality satisfaction degrees of the existing connections. Simulation results show that our proposed dynamic schemes are able to provide a stable service, which well meets the user's quality requirements.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.3
• /
• pp.361-367
• /
• 2021

Information search in video streaming services such as YouTube is replacing traditional information search services. To find desired detailed information in such a video, users should repeatedly navigate several points in the video, resulting in a waste of time and network traffic. In this paper, we propose a method to assist users in searching for information in a video by using DBSCAN clustering and LSTM. Our LSTM model is trained with a dataset that consists of user search sequences and their final target points categorized by DBSCAN clustering algorithm. Then, our proposed method utilizes the trained model to suggest an expected category for the user's desired target point based on a partial search sequence that can be collected at the beginning of the search. Our experiment results show that the proposed method successfully finds user destination points with 98% accuracy and 7s of the time difference by average.

• Journal of Cadastre & Land InformatiX
• /
• v.51 no.2
• /
• pp.141-150
• /
• 2021

Fine dust is a substance that greatly affects human health, and various studies have been conducted in this regard. Due to the human influence of particulate matter, various studies are being conducted to predict particulate matter grade using past data measured in the monitoring network of Seoul city. In this paper, predictive model have focused on particulate matter concentration in May, 2019, Seoul. The air pollutant variables were used to training such as SO₂, CO, NO₂, O₃. The predictive model based on Adaboost, and training model was dividing PM₁₀ and PM_2.5. As a result of the prediction performance comparison through confusion matrix, the Adaboost model was more conformable for predicting the particulate matter concentration grade. Although air pollutant variables have a higher correlation with PM2.5, training model need to train a lot of data and to use additional variables such as traffic volume to predict more effective PM₁₀ and PM_2.5 distribution grade.