Volume 15, Issue 1, February-March 2024
Future Landscape, Challenges, and The Role of the Internet of Things (IoT) in Empowering Sustainable Development Goals (SDGs) in Pakistan |
Pages: 1-10 (10) | [Full Text] PDF (752K) |
M Hasnain Abbas Khan, Farhan Abid, Junaid Arshad |
Department of Computer Science, University of Engineering and Technology (UET), Lahore, Pakistan |
Abstract - The fast development and adoption of smart IoT (Internet of Things)-based technologies has created new opportunities for technological growth in many areas of life. The major purpose of IoT technologies is to simplify procedures in numerous industries, boost system efficiency, save time and money, and, eventually, improve quality of life. Today, digitization has a significant impact on the future of humanity, industries, and everything else around us. As a result, digital technologies have gained popular attention. This problem impacts even emerging countries like Pakistan. This article will look at how digital technologies, such as IoT, may assist poor nations meet their Sustainable Development Goals. This notion is explained with instances from Pakistan. As a developing country, Pakistan confronts various hurdles in its quest for technical advancement and sustainability. To fully capitalize on the potential of sophisticated technologies that are meant to bring about real sustainability, Pakistan would need to adopt operational strategies and solutions to cope with these difficulties while minimizing risks and maximizing benefits. |
Index Terms - IoT Future, Challenges, Digital Technologies, Sustainability, Sustainable Development Goals, Artificial Intelligence and Pakistan’s Digital Transformation |
Applications and Services for Next-Generation Industrial Internet of Things |
Pages: 11-20 (10) | [Full Text] PDF (491K) |
Chou Kung Yi, Lea Yue Wang |
Department of Computer Science and Information Technology (CS&IT), Korea |
Abstract - As a complex cyber-physical system, IoT integrates various devices equipped with sensing, identification, processing, communication, and networking capabilities. In particular, sensors and actuators are getting increasingly powerful, less expensive and smaller, which makes their use ubiquitous. Industries have strong interest in deploying IoT devices to develop industrial applications such as automated monitoring, control, management, and maintenance. Due to the rapid advances in technology and industrial infrastructure, IoT is expected to be widely applied to industries. For example, the food industry is integrating WSN and RFID to build automated systems for tracking, monitoring, and tracing food quality along the food supply chain in order to improve food quality. This paper reviews the recent researches on IoT from the industrial perspective. Firstly, we introduce the background and SOA models of IoT and then discuss the fundamental technologies that might be used in IoT. Next, we introduce some key industrial applications of IoT. Afterward, we analyzed the research challenges and future trends associated with IoT. Different from other IoT survey papers, a main contribution of this review paper is that it focuses on industrial IoT applications and highlights the challenges and possible research opportunities for future industrial researchers. |
Index Terms - Information and Communication Technology (ICT), Industrial System, Applications, Services and Industrial Internet of Things-IIoT |
Mitigating Malicious URLs Using Machine Learning Techniques |
Pages: 21-28 (8) | [Full Text] PDF (1.29M) |
MeharunNisa, Ahmad Raza, M. Junaid Arshad |
Department of Computer Science, University of Engineering and Technology, Lahore, Pakistan |
Abstract - This paper provides a review of recent studies on malign URL detection using models from Machine Learning, with a focus on their application in the context of the what dataset being used, what number of dataset being included, if feature engineering has been applied and if yes then what type of features are being used, also the comprehensive view of classifiers used at time in all those papers under discussion with. Moreover it has analyzed each study’s methodology, outcomes, and conclusions, and discuss the strengths and limitations with mentioned future work in each. While the tabular comparisons are for the very latest publications as of 2023 and 2024. The analysis shows that while existing metrics can be useful for identifying areas of improvement, they may not always provide a complete picture of software quality. The goal of this study is to determine the finest factors helping in getting the effective technique for spotting a phishing URL in large datasets. When employing machine learning algorithms to identify phishing URLs, users encounter numerous difficulties. It must be done to defend users against phishing attempts if you want people to continue having trust with online platforms and services. In order to ensure the security of user details as well as adhere with industry standards and data protection requirements, phishing URL detection must be reliable. |
Index Terms - Machine Learning, Intrusion Detection, Cyber Attack and Phishing Detection |
Advancements in Automated Penetration Testing for IoT Security by Leveraging Reinforcement Learning |
Pages: 29-34 (6) | [Full Text] PDF (317K) |
Abdul Samad, Saad Altaf |
Department of Computer Science, University of Engineering and Technology, Lahore-Pakistan |
Abstract - Penetration testing, commonly referred to as pentesting or PT, is a prevalent method for actively evaluating the security measures of a computer network. This involves planning and executing various attacks to identify and exploit existing vulnerabilities. Despite the continuous evolution of tools, current penetration testing methods are becoming increasingly non-standard, intricate, and resource-intensive. In this paper, we propose and assess an innovative AI-driven pen-testing system named the Intelligent Automated Penetration Testing System (IAPTS). This system utilizes machine learning techniques, specifically Reinforcement Learning (RL), to comprehend and replicate both average and complex pen-testing activities. IAPTS comprises a module that seamlessly integrates with established PT frameworks, allowing it to capture information, learn from experiences, and replicate tests in subsequent similar testing scenarios. The primary objective of IAPTS is to optimize human resources while delivering significantly improved results in terms of time efficiency, reliability, and testing frequency. The approach taken by IAPTS involves modeling PT environments and tasks as a partially observed Markov decision process (POMDP) problem, which is effectively solved by a POMDP solver. Although the focus of this paper is limited to PT planning for network infrastructures and not the entire practice, the findings strongly support the hypothesis that RL can elevate PT capabilities beyond those of any human PT expert, particularly in terms of time efficiency, coverage of attack vectors, and the accuracy and reliability of outputs. Furthermore, this research addresses the intricate challenge of capturing and reusing expertise by empowering the IAPTS learning module to store and reuse PT policies. This mimics the learning process of a human PT expert but in a more efficient manner |
Index Terms - Machine Learning, Software Security, Automated Penetration Testing, Attack Tree, Deep Reinforcement Learning and Deep Learning Networks |
A Comprehensive Review of Artificial Intelligence and Machine Learning to Overcome Globalwarming Impacts |
Pages: 35-42 (8) | [Full Text] PDF (226K) |
Sajjad Sohail, Sharjeel Tariq, M. Junaid Arshad |
Department of Computer Science, University of Engineering and Technology, Lahore |
Abstract - Advances in technology have transformed the way we understand and respond to weather patterns, leading to significant improvements in forecasting and climate change mitigation. Historically, weather prediction played a crucial role in agriculture and human migration, but it has faced numerous challenges due to data ambiguity and the limitations of traditional numerical weather prediction (NWP) models. However, with the emergence of machine learning (ML), deep learning (DL), and the Internet of Things (IoT), new opportunities for more accurate and timely weather predictions have arisen. These advancements hold the potential to protect crops, alert farmers to adverse conditions, and contribute to climate change mitigation strategies. Machine learning and artificial intelligence (AI) methods have been integrated into weather forecasting to enhance prediction accuracy. Techniques such as neural networks, motion detection, and computer vision are applied to analyze vast datasets, providing more precise environmental monitoring. Innovations in IoT, along with machine learning models, offer the ability to detect changes in weather conditions in real-time, leading to more responsive forecasting. In this comprehensive review, we explore various deep learning, machine learning, and IoT-based approaches that are employed to improve weather prediction and analyze their comparative effectiveness. By examining these advanced methods, we aim to highlight their role in combating the impacts of global warming and supporting sustainable practices. The review also underscores how AI and ML techniques contribute to mitigating the consequences of climate change, enabling a proactive approach to safeguarding the environment and addressing global challenges. |
Index Terms - Machine Learning, Deep Learning, Weather Prediction, Weather Forecast, IoT, Python and Artificial Neural Network |
Analyzing Internet Traffic Dynamics for Enhanced Emergency Response in IoT Environments |
Pages: 43-49 (7) | [Full Text] PDF (375K) |
Nimra Latif |
Department of Computer Science and Engineering, University of Engineering and Technology, LHR |
Abstract - In this research, we address the flow of internet traffic patterns and the safety of message exchange in this work, with M2M (machine-to-machine) communication technology. By analyzing the whole dataset from different IoT machines and also mimicked attacks, we wish to uncover traffic flows, detect abnormality and prescribe security approaches to improve the emergency communication networks defense. Data are subjected to advanced analysis methods by us in order show the weak areas as well as ethical considerations. Our objective is also to help to avoid the risks to data privacy and security. We believe our discoveries to be a basic step toward coping strategies improvements that put in place reliable emergency and situation communication infrastructure. |
Index Terms - Internet Traffic, Emergency Response, IoT Environment and Strategies |