• Journal of Information Processing Systems
• /
• v.20 no.3
• /
• pp.375-390
• /
• 2024

Edge computing architecture has effectively alleviated the computing pressure on cloud platforms, reduced network bandwidth consumption, and improved the quality of service for user experience; however, it has also introduced new security issues. Existing anomaly detection methods in big data scenarios with cloud-edge computing collaboration face several challenges, such as sample imbalance, difficulty in dealing with complex network traffic attacks, and difficulty in effectively training large-scale data or overly complex deep-learning network models. A lightweight deep-learning model was proposed to address these challenges. First, normalization on the user side was used to preprocess the traffic data. On the edge side, a trained Wasserstein generative adversarial network (WGAN) was used to supplement the data samples, which effectively alleviates the imbalance issue of a few types of samples while occupying a small amount of edge-computing resources. Finally, a trained lightweight deep learning network model is deployed on the edge side, and the preprocessed and expanded local data are used to fine-tune the trained model. This ensures that the data of each edge node are more consistent with the local characteristics, effectively improving the system's detection ability. In the designed lightweight deep learning network model, two sets of convolutional pooling layers of convolutional neural networks (CNN) were used to extract spatial features. The bidirectional long short-term memory network (BiLSTM) was used to collect time sequence features, and the weight of traffic features was adjusted through the attention mechanism, improving the model's ability to identify abnormal traffic features. The proposed model was experimentally demonstrated using the NSL-KDD, UNSW-NB15, and CIC-ISD2018 datasets. The accuracies of the proposed model on the three datasets were as high as 0.974, 0.925, and 0.953, respectively, showing superior accuracy to other comparative models. The proposed lightweight deep learning network model has good application prospects for anomaly traffic detection in cloud-edge collaborative computing architectures.

• Journal of Korean Society of Transportation
• /
• v.36 no.2
• /
• pp.67-85
• /
• 2018

Vehicle platooning through wireless communication and automated driving technology has become realized. Platooning is a technique in which several vehicles travel at regular intervals while maintaining a minimum safety distance. Truck platooning is of keen interest because it contributes to preventing truck crashes and reducing vehicle emissions, in addition to the increase in truck flow capacity. However, it should be noted that interactions between vehicle platoons and adjacent manually-driven vehicles (MV) significantly give an impact on the performance of traffic flow. In particular, when vehicles entering from on-ramp attempt to merge into the mainstream of freeway, proper interactions by adjusting platoon size and inter-platoon spacing are required to maximize traffic performance. This study developed a methodology for establishing operational strategies for truck platoonings on freeway on-ramp areas. Average speed and conflict rate were used as measure of effectiveness (MOE) to evaluate operational efficiency and safety. Microscopic traffic simulation experiments using VISSIM were conducted to evaluate the effectiveness of various platooning scenarios. A decision making process for selecting better platoon operations to satisfy operations and safety requirements was proposed. It was revealed that a platoon operating scenario with 50m inter-platoon spacing and the platoon consisting of 6 vehicles outperformed other scenarios. The proposed methodology would effectively support the realization of novel traffic management concepts in the era of automated driving environments.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.23 no.1
• /
• pp.134-145
• /
• 2024

The purpose of this study was to propose a framework for developing test scenarios for issuance of conditional driver's licenses. The framework was composed of five stages. Initially, we reviewed the literature on traffic crash characteristics in terms of accident frequency and severity regarding the main factors of crashes caused by older drivers. In the second stage, the characteristics of crashes attributed to non-elderly, early elderly, and late elderly drivers were analyzed using data obtained from the Traffic Accident Analysis System (TAAS), and crash types for elderly drivers were derived. In the third stage, black box videos of high-risk crash types were analyzed to derive crash stories that described the circumstances in which crashes occurred. In the fourth step, crash situations were classified by rating the types of crash stories derived to develop various scenarios. Step 5 involved creating a scenario by applying the PEGASUS 5-Layer format, which has recently been used to develop test scenarios for autonomous vehicles. The results of this study are expected to be used as a basis for developing driving ability evaluation scenarios for the issuance of conditional driver's licenses.

• International Journal of Aeronautical and Space Sciences
• /
• v.13 no.3
• /
• pp.377-385
• /
• 2012

The Automatic Dependent Surveillance Broadcast (ADS-B) system is a key component of CNS/ATM recommended by the International Civil Aviation Organization (ICAO) as the next generation air traffic control system. ADS-B broadcasts identification, positional data, and operation information of an aircraft to other aircraft, ground vehicles and ground stations in the nearby region. This paper explores the ADS-B based trajectory prediction and the conflict detection algorithm. The multiple-model based trajectory prediction algorithm leads accurate predicted conflict probability at a future forecast time. We propose an efficient and accurate algorithm to calculate conflict probability based on approximation of the conflict zone by a set of blocks. The performance of proposed algorithms is demonstrated by a numerical simulation of two aircraft encounter scenarios.

• Journal of Korean Society of Transportation
• /
• v.18 no.1
• /
• pp.87-97
• /
• 2000

Signal optimization model is divided bandwidth-maximizing model and delay-minimizing model. Bandwidth-maximizing model express model formulation as MILP(Mixed Integer Linear Programming) and delay-minimizing model like TRANSYT-7F use "hill climbing" a1gorithm to optimize signal times. This study Proposed optimization model using genetic algorithm one of evolution algorithm breaking from existing optimization model This Proposed model were tested by several scenarios and evaluated through NETSIM with TRANSYT-7F\`s outputs. The result showed capability that can obtain superior solution.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.8
• /
• pp.674-682
• /
• 2013

Along with increasing of its subscribers and traffic for high-speed LTE service, it is widely expected to bring about the 2nd data explosion and results in any inconvenience of users. This paper proposes a smart traffic management scheme and its system design based on user's participating with the mobile application, in order to provide a fast and stable service for end users as well as to resolve the network overhead with increasing traffic. By making use of a time shift by users who want to reduce data usage amount and communication cost, the proposed scheme tries to reduce the peak traffic by shifting the delay-tolerant services, such as file sharing and software download, into the other time line. It also reduces the high volumed video traffics by transforming them into a user customized one. In addition, we suggest a mechanism to select video size and service time, then proposes an overall system design and its user service scenarios.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.2
• /
• pp.27-35
• /
• 2009

The roundabout was developed in the United Kingdom to rectify problems associated with these traffic circles. In 1966, the United Kingdom adopted a mandatory "give-way" rule at alt circular intersections, which required entering traffic to give way, or yield, to circulating traffic. However, the roundabout used in United States or Europe do not common use in Korea. The purposes of the study is to comparatively analyze the effectiveness of roundabout. In pursuing the above, this study designs the scenarios which can reflect the directional traffic volumes, and analyzes the average control delay of roundabout and 4-leg signalized intersection by SIDRA. The main study results are as follows. First, when the entering traffic volume are less than 2,000pcph, the single-lane roundabout is analyzed to be more effective than 4-leg signalized intersection. Second, when the entering traffic volume are less than 3,200pcph, the double-lane roundabout is evaluated to be more effective than 4-leg signalized intersection. Third, which is installed left and right turn lane at intersection, area for the roundabout is analyzed to be less than that for general intersection area.

• Journal of Korean Society of Transportation
• /
• v.27 no.5
• /
• pp.51-62
• /
• 2009

Variable message signs (VMS) are a useful tool for providing real-time traffic information to drivers. In particular, effective warning information provision leading to safer driving would be an important countermeasure to prevent traffic accidents. The purpose of this study was to identify users' preferences for traffic safety warning information formats. A variety of warning information scenarios using text and pictograms were devised and investigated for the purpose of selecting more effective methods to provide warning information. A portable variable message sign (PVMS) was used to evaluate users' preferences. The results of this study can be used for designing better warning information for the enhancement of traffic safety.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.19 no.6
• /
• pp.30-36
• /
• 2011

The current test methods are insufficient to evaluate and ensure the safety and reliability of vehicle system for all possible dynamic situation including the worst case such as rollover, spin-out and so on. Although the known NHTSA Sine with dwell steering maneuvers are applied for the vehicle performance assessment, they aren't enough to estimate other possible worst case scenarios. Therefore, it is crucial for us to verify the various worst cases including the existing severe steering maneuvers. This paper includes useful worst case based upon the existing worst case scenarios mentioned above and worst case evaluation for vehicle dynamic controller in simulation basis and UCC HILS. The only human steering angle is selected as a design parameter here and optimized to maximize the index function to be expressed in terms of both yaw rate and side slip angle. The obtained scenarios were enough to generate the worst case to meet NHTSA worst case definition. It has been concluded that the new procedure in this paper is adequate to create other feasible worst case scenarios for a vehicle dynamic control system.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.16 no.6
• /
• pp.1-9
• /
• 2008

The current test methods are insufficient to evaluate and ensure the safety and reliability of vehicle system for all possible dynamic situation including the worst case such as rollover, spin-out and so on. Although the known NHTSA J-turn and Fish-hook steering maneuvers are applied for the vehicle performance assessment, they aren't enough to estimate other possible worst case scenarios. Therefore, it is crucial for us to verify the various worst cases including the existing severe steering maneuvers. This paper includes the procedure to search for other useful worst case based upon the existing worst case scenarios mentioned above and its application in simulation basis. The only human steering angle is selected as a design parameter here and optimized to maximize the index function to be expressed in terms of either roll angle or yaw rate. The obtained scenarios were enough to generate the worst case to meet NHTSA worst case definition (ex.2-inch wheel lift). Additionally, as an application, the worst case steering maneuver is acquired for the vehicle to operate with a simple ESP system. It has been concluded that the new procedure in this paper is adequate to create other feasible worst case scenarios for a vehicle system both with an intelligent safety control system and without it.