Research Briefs on Information and Communication Technology Evolution

Malicious Packet Detection Technology Using Reinforcement Learning

2024-04-25T09:14:38+00:00

In the present day, the advancement of 5G and Internet of Things (IoT) technology has resulted in the interconnectedness of everyday objects through networks. However, attempts to exploit networked computers for malicious purposes continue to rise while the attacks utilizing malicious codes to compromise user information's confidentiality and integrity are becoming increasingly sophisticated and intelligent. To counter these evolving threats, researches have been conducted on a method to identify malicious network packets using a combination of security control systems and Artificial Intelligence (AI) technology, especially supervised learning. Unfortunately, the current cybersecurity control systems suffer from inefficiencies in terms of both manpower and cost. Moreover, the surge in remote work has created challenges in responding swiftly to security incidents. Furthermore, the existing AI technology based on supervised learning has limitations, particularly in detecting new variants of malicious code, and its accuracy in identifying malicious code depends heavily on the quantity and quality of available data. In light of these challenges, this study, reinforcement learning is employed to overcome the limitations of the original supervised learning-based malicious packet detection system, such as high dependency on training data and the failure to detect variant malicious packets. This research proposes a malicious packet detection technology capable of addressing new malicious packets or variant types effectively.

Implementation and Evaluation of Mutual Authentication Protocol for Artificial Pancreas System

2024-05-27T21:28:02+00:00

Bluetooth Low Energy (BLE) is an international standard protocol widely used in IoT (Internet of Things) environments due to its low power consumption. However, when BLE is used in an Artificial Pancreas System (APS), it faces the threat of Man-In-The-Middle (MITM) attacks. These attacks can lead to eavesdropping on sensitive medical information or manipulation of insulin injection commands, potentially endangering the patient's life. In this paper, we introduce a protocol that mitigates the vulnerability of BLE to MITM attacks and enhances resilience to sudden situations that may occur in APS environments. Finally, we evaluate the implementation suitability by deploying this protocol in a test environment.

Predictive Analytics in Law Enforcement: Unveiling Patterns in NYPD Crime through Machine Learning and Data Mining

2024-06-05T18:27:53+00:00

Urban crime poses multifaceted challenges to cities' socio-economic structures. This study employs machine learning and data mining to bolster predictive policing in New York City. Using a comprehensive NYPD crime dataset spanning 2006 to 2017, the analysis identifies historical patterns and forecasts future crime trends. Rigorous methodologies ensure data fidelity, with algorithms like Random Forest and K-Means clustering parsing the intricate spatio-temporal crime dynamics. Results pinpoint crime hotspots and track criminal activity evolution, informing strategic law enforcement resource allocation and community involvement. Ethical considerations, including data privacy and algorithmic biases, are scrutinized alongside their impacts on community-police relations. The study recommends operational improvements and advocates for ongoing innovation in data-driven public safety strategies, advocating for the integration of new data sources and analytical methods in advancing smart city infrastructures.

Implementation of the Drowsy Driving Prevention System using AI-Based Conversations to Maintain Driver Arousal

2024-06-20T07:05:04+00:00

Since car accidents caused by drowsy driving have been one of significant issues for Intelligent Transport Systems, many previous studies have focused on developing safe driving support systems. However, most of these studies rely on sensors to detect drowsiness, which may result in delayed prevention due to the fast movement of the car. Therefore, this study proposes a drowsy driving prevention system that utilizes a smart speaker to maintain the driver's arousal. The proposed method involves two types of conversations controlled by cloud-based AI modules with the drivers. The first type is lightweight, consisting of selectable questions, while the second type is heavyweight, involving natural conversations with the AI module. This paper reports on experiments conducted with a implemented prototype system of the proposed method. The results show that the proposed system effectively maintains the driver's arousal state. However, natural conversations may induce a slightly annoyed state compared to selectable conversations.

Analyzing the Impact of Smoking and Drinking on Health Metrics in Korea: A Data-driven Predictive Approach

2024-06-20T07:11:35+00:00

This investigation thoroughly explores the effects of smoking and alcohol consumption on health measures within the population of South Korea, using a comprehensive dataset from the National Health Insurance Service of Korea. The study aims to build and validate predictive models by applying logistic regression methodologies, outlier detection, and cross-validation techniques. These models are designed to identify patterns in smoking and drinking behaviors, offering predictive accuracy. The analysis clarifies the relationship between lifestyle choices and key health indicators such as liver enzyme activity, blood fat levels, and heart and blood vessel measurements. The outcomes of this study are ready to significantly contribute to the development of targeted public health strategies, aiming to lessen the risks associated with lifestyle-induced health conditions.

Implementation and Evaluation of Pull-type Rationed Goods Request Framework

2024-07-31T14:55:03+00:00

In this study, we implemented and evaluated a pull-type rationed goods request framework. The proposed framework comprises three component systems. The first system is the individual rationed goods request system for disaster victims, which allows disaster victims to request personalized rationed goods. The second system is the individual rationed goods request system for informationally disadvantaged disaster victims, which allows informationally disadvantaged disaster victims, e.g., the elderly, to request personalized rationed goods using tablet terminals installed at evacuation centers. The third system is the individual rationed goods request management system for system administrators. This system allows system administrators to centrally manage information on rationed goods requested from the individual rationed goods request system for disaster victims and the individual rationed goods request system for informationally disadvantaged disaster victims. The proposed pull-type rationed goods request framework was evaluated using the system usability scale evaluation with 30 participants. In addition, the 30 subjects evaluated the operability, readability, functionality, and relevance of the proposed framework . The results demonstrate that the proposed framework received high marks in many items.

Survey on Security for Non-Terrestrial Networks

2025-01-09T07:16:10+00:00

Non-terrestrial networks (NTN) are studied in a way that goes beyond the environmental limitations of existing terrestrial networks. In particular, it ensures connectivity in remote areas, rural areas, and areas where communication infrastructure is insufficient.
NTN leverages state-of-the-art space and aviation platforms to provide highly versatile services in domains such as IoT and disaster management. These NTNs are the cornerstone of next-generation communication systems, and at the same time, they are likely to address gaps in the connectivity of communications. In this paper, we examine the latest trends related to the security of NTN, and analyze its network-specific characteristics, methodologies, and challenges. Integrating NTN with the ground infrastructure creates security challenges such as seamless handover management between satellites and ground systems, cyber threats from quantum computers, and development of encryption technologies for IoT environments with limited resources. To address these challenges, this study categorizes NTN security requirements into security requirements that apply to all network operations and security requirements for the inherent vulnerabilities of NTN. According to the analysis of NTN, it highlights the need for Post Quantum encryption algorithm, robust handover security protocol, and lightweight encryption method suitable for the dynamic and resource-limited nature of NTN. Advanced technologies such as AI-based threat detection, blockchain-based mechanisms for data integrity, and adaptive security frameworks are showing the possibility of secure and scalable NTN communication. This paper raises recommendations for the secure integration of NTN with ground network based on the results of a recent comprehensive study. This survey provides direction to aid in the development of standardized security protocol and at the same time provide NTN future oriented communication solution. Addressing these challenges will show the potential of NTN as a key component of the global communication infrastructure in the coming era.

Development of Scene Segmentation to Improve Work Efficiency of Learner Monitoring

2025-01-09T07:18:59+00:00

Since the outbreak of COVID-19, many universities have introduced video conference systems and e-learning systems, but these issues still make it difficult to obtain actual learning time. However, during operation of these systems, the participants or learner may not watch the video because they have also other tasks. For this reason, we have developed some prototype systems for monitoring learner behavior at watching learning content. In this paper, we describe the development of video scene segmentation based on a video correlation matrix, which reflects learner behavior, aiming to enhance the efficiency of learner monitoring. From our evaluation, we have been able to segment scenes from videos capturing learners based on the difference in correlation values between adjacent frames.

Security Analysis of Network Slicing-based Mission-Critical Services with Formal Verification Tool

2025-01-09T07:28:59+00:00

Mission Critical Service is a service requiring high reliability and low latency, utilized in various fields such as automotive safety systems, disaster safety communications, and medical services. In particular, group communication is essential in disaster scenarios, where the secure delivery of information and user reliability are critical factors. This paper focuses on the user authentication and service authorization procedures for Mission Critical X (MCPTT, MCVideo, MCData), the core components of MCS. The OpenID Connect 1.0 protocol is employed as a key framework for authentication and identity verification, while the TLS protocol ensures the confidentiality and integrity of data. The application procedures of these protocols are elaborated, and the security of the TLS (Transport Layer Security) protocol used in Mission Critical X within the Mission Critical Service environment is analyzed using the formal verification tool BAN-Logic.

Construction of a Cybersecurity Behavior Knowledge Base for Malicious Behavior Analysis

2025-01-09T07:31:16+00:00

Facing the surge in malicious behaviors in the network environment, the existing cybersecurity knowledge graph suffers from fragmented security knowledge and limited application scenarios, making it challenging to collaborative malicious behavior analysis. To address this, we propose a cybersecurity behavior knowledge base (CSBKB) framework for comprehensive malicious behavior analysis. Based on knowledge of user behavior, attack traffic, and attack paths, we construct six types of knowledge graphs to characterize malicious behavior, including user behavior perception, user behavior mapping, malicious behavior association, malicious behavior category, domain attack, and malicious behavior path traceability graph. These graphs characterize malicious behaviors and form a comprehensive security behavior knowledge base. To fully utilize the graph structure information, we design a reasoning module based on the graph neural network further to explore the relationship between entities in the graph. Using DDoS attacks as a case study, we demonstrate this framework's construction and knowledge-reasoning capabilities. Experimental results demonstrate that the proposed CSBKB framework effectively realizes a comprehensive malicious behavior analysis mechanism encompassing "malicious user behavior monitoring, malicious behavior type detection, and malicious behavior path tracing." It can effectively analyze malicious behaviors, with an accuracy of more than 0.97 in detecting abnormal users, more than 0.97 in inferring DDoS attack types, and an identification rate of more than 0.92 for malicious behavior paths.