Mihui Kim

Prof. Mihui Kim

orcid

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 November 2016 in Sensors
Verified
Journal Article
Open access
PDF file
MIMason Itkin
Mason Itkin
MKMihui Kim
Mihui Kim
Younghee ParkYounghee Park
Younghee Park
Unmanned aerial vehicles (UAVs) are an emerging technology with the potential to revolutionize commercial industries and the public domain outside of the military. UAVs would be able to speed up rescue and recovery operations from natural disasters and can be used for autonomous delivery systems (e.g., Amazon Prime Air). An increase in the number of active UAV systems in dense urban areas is attributed to an influx of UAV hobbyists and commercial multi-UAV systems. As airspace for UAV flight becomes more limited, it is important to monitor and manage many UAV systems using modern collision avoidance techniques. In this paper, we propose a cloud-based web application that provides real-time flight monitoring and management for UAVs. For each connected UAV, detailed UAV sensor readings from the accelerometer, GPS sensor, ultrasonic sensor and visual position cameras are provided along with status reports from the smaller internal components of UAVs (i.e., motor and battery). The dynamic map overlay visualizes active flight paths and current UAV locations, allowing the user to monitor all aircrafts easily. Our system detects and prevents potential collisions by automatically adjusting UAV flight paths and then alerting users to the change. We develop our proposed system and demonstrate its feasibility and performances through simulation.

5 November 2021 in Sensors
Verified
Journal Article
Open access
PDF file
GLGihun Lee
Gihun Lee
MKMihui Kim
Mihui Kim
Recently, artificial intelligence has been successfully used in fields, such as computer vision, voice, and big data analysis. However, various problems, such as security, privacy, and ethics, also occur owing to the development of artificial intelligence. One such problem are deepfakes. Deepfake is a compound word for deep learning and fake. It refers to a fake video created using artificial intelligence technology or the production process itself. Deepfakes can be exploited for political abuse, pornography, and fake information. This paper proposes a method to determine integrity by analyzing the computer vision features of digital content. The proposed method extracts the rate of change in the computer vision features of adjacent frames and then checks whether the video is manipulated. The test demonstrated the highest detection rate of 97% compared to the existing method or machine learning method. It also maintained the highest detection rate of 96%, even for the test that manipulates the matrix of the image to avoid the convolutional neural network detection method.

19 October 2012 in Energies
Verified
Journal Article
Open access
PDF file
KCKyung Choi
Kyung Choi
XCXinyi Chen
Xinyi Chen
SLShi Li
Shi Li
MKMihui Kim
Mihui Kim
KCKijoon Chae
Kijoon Chae
JNJungChan Na
JungChan Na
In this paper, we analyze the Network and System Management (NSM) requirements and NSM data objects for the intrusion detection of power systems; NSM is an IEC 62351-7 standard. We analyze a SYN flood attack and a buffer overflow attack to cause the Denial of Service (DoS) attack described in NSM. After mounting the attack in our attack testbed, we collect a data set, which is based on attributes for the attack. We then run several data mining methods with the data set using the Waikato Environment for Knowledge Analysis (WEKA). In the results, we select the decision tree algorithms with high detection rates, and choose key attributes in high level components of the trees. When we run several data mining methods again with the data set of chosen key attributes, the detection rates of most data mining methods are higher than before. We prove that our selected attack attributes, and the proposed detection process, are efficient and suitable for intrusion detection in the smart grid environment.

1 January 2016 in Journal of Computer Networks and Communications
Verified
Journal Article
Open access
PDF file
JHJune S. Hong
June S. Hong
MKMihui Kim
Mihui Kim
Small power plants and buildings with renewable power generation capability have recently been added to traditional central power plants. Through these facilities, prosumers appear to have a concurrent role in both energy production and consumption. Based on bidirectional power transfers by large numbers of prosumers, a smart microgrid has become an important factor in efficiently controlling the microgrids used in power markets and in conducting effective power trades among grids. In this paper, we present an approach utilizing the game theory for effective and efficient energy routing, which is a novel and challenging procedure for a smart microgrid network. First, we propose strategies for choosing the desired transaction price for both electricity surpluses and shortages to maximize profits through energy transactions. An optimization scheme is utilized to search for an energy route with minimum cost using the solving method used in a traditional transportation problem by treating the sale and purchase quantities as transportation supply and demand, respectively. To evaluate the effect of the proposed decision strategies, we simulated our mechanism, and the results proved that our mechanism yields results pursued by each strategy. Our proposed strategies will contribute to spreading a smart microgrid for enhancing the utilization of microgrids.

16 May 2020 in Future Internet
Verified
Journal Article
Open access
PDF file
MKMihui Kim
Mihui Kim
YKYoungmin Kim
Youngmin Kim
As a representative example for the construction of a smart city, a smart parking system has been developed in past research and implemented through IoT and cloud technologies. However, the initial installation cost of IoT sensor devices is preventing the spread of this technology, and thus as an alternative, a crowdsensing-based system, operating through data from publicly owned mobile devices, has been proposed. In this paper, we propose a multi-blockchain structure (i.e., constructed with public chain and private chain) in a crowdsensing-based smart parking system. In this structure, many sensing data contributors participate through the opened public blockchain, to transparently provide sensing information and to claim corresponding rewards. The private blockchain provides an environment for sharing the collected information among service providers in real time and for providing parking information to users. The bridge node performs an information relay role between the two blockchains. Performance analysis and security analysis on the implemented proposed system show the feasibility of our proposed system.

28 June 2024 in Electronics
Verified
Journal Article
Open access
PDF file
HKhaesoo kim
haesoo kim
MKMihui Kim
Mihui Kim
For malicious purposes, attackers hide malware in the software used by their victims. New malware is continuously being shared on the Internet, which differs both in terms of the type of malware and method of damage. When new malware is discovered, it is possible to check whether there has been similar malware in the past and to use the old malware to counteract the new malware; however, it is difficult to check the maliciousness and similarity of all software. Thus, deep learning technology can be used to efficiently detect and classify malware. This study improves this technology’s accuracy by converting static features, which are binary data, into images and by converting time-series data, such as API call sequences, which are dynamic data with different lengths for each datum, into data with fixed lengths. We propose a system that combines AI-based malware detection and classification systems trained on both static and dynamic features. The experimental results showed a detection accuracy of 99.34%, a classification accuracy of 95.1%, and a prediction speed of approximately 0.1 s.