Volume 16, Issue 4, August 2025

Optimizing IoT Ecosystem through Scalable, Secure and Efficient Network Management

Muhammad Aleem Subhani, Muhammad Junaid Arshad

Department of Computer Science, University of Engineering and Technology, Lahore, Pakistan

Abstract - The rapid growth of the Internet of Things (IoT) has transformed numerous industries by enabling seamless communication among billions of devices, sensors, and applications. However, traditional network infrastructures struggle to meet the scalability, security, and efficiency demands of these dynamic and expanding ecosystems. Limitations such as increased latency, inefficient resource allocation, and vulnerability to cyber threats hinder the performance and reliability of IoT networks. This study proposes a comprehensive framework to optimize IoT ecosystems by integrating advanced network management techniques. Scalability is addressed through virtualization, network slicing, and dynamic resource allocation, enabling adaptive responses to device proliferation. To strengthen security, the framework incorporates firewalls, Access Control Lists (ACLs), and Intrusion Detection Systems (IDS) for proactive monitoring and mitigation of network threats. Efficiency is improved using load balancing and algorithmic optimization to ensure effective bandwidth usage and resource distribution. The proposed solution is validated using Cisco Packet Tracer simulations, where traditional and virtualized network models are implemented and analyzed under identical conditions. Key performance metrics latency, throughput, packet loss, and resource utilization are measured to evaluate and compare network behavior. Results demonstrate significant improvements across all metrics, confirming the effectiveness of the virtualized approach. This research offers a practical and scalable model for modern IoT applications, including smart cities, healthcare systems, and industrial automation. By addressing the core limitations of existing network infrastructures, the proposed framework facilitates the development of resilient, secure, and energy-efficient IoT environments.

Index Terms - Internet of things (IoT), Software Defined Networks (SDN), NFV (Network Function Virtualization), Network Management, Security, Scalability and Efficiency

Security Threat Landscape of Wireless Communication in IoT-Driven Healthcare Platforms

Pages: 9-15 (7) | [Full Text] PDF (413K)

Babatunde Damilola, Chinara Adebayo

Department of Information Technology, University of Science and Technology, Nigeria

Abstract - The Internet of Things (IoT) has numerous application areas, with healthcare being one of the most prominent. The growing adoption of wireless sensors and identification technologies within the healthcare sector has drawn significant attention to an emerging approach known as Wireless Medical Sensor Networks (WMSNs). Like any evolving technology, IoT—particularly in the healthcare domain utilizing WMSNs—faces several challenges, with security being one of the most critical concerns. This paper examines the technological advancements in IoT-based healthcare systems and analyzes the security requirements and potential threats associated with wireless sensor networks (WSNs) in healthcare applications. Additionally, it presents a comprehensive literature review of the existing solutions proposed to secure healthcare systems and highlights unresolved issues that need further discussion. These concerns are particularly relevant to improving the quality of life for vulnerable groups such as children, the elderly, and patients with chronic illnesses. The primary aim is to provide readers with an in-depth understanding of the progress made (in terms of protocols and solutions) and to identify the security challenges that still need to be addressed in this rapidly evolving field.

Index Terms - Healthcare Systems, Wireless Sensor Network (WSN), Applications, Evolving Areas and Security Issues

Emerging Trends in Computer Architecture: Quantum, Neuromorphic, and Reconfigurable Systems

Pages: 16-22 (7) | [Full Text] PDF (302K)

Sofia Munawar

Data Science Department, University of Engineering and Technology, Lahore

Abstract - Computer architecture is undergoing a major change as the demand for intelligent, scalable systems that save a lot of energy keeps increasing. Because standard silicon computers are running into physical and performance limitations, new computing methods have to be found. This paper presents a detailed study of quantum computing, neuromorphic systems and reconfigurable structures which are currently developing fast in computer architecture. Using new research and advances in technology, the paper describes the key features, types of design and important ways each paradigm is operated. As well, it tackles the different use cases—such as cryptography, optimization, sensory processing and adaptive computing—that offer significant advantages for these structures. This report discusses how these systems might come together, with a main focus on their ability to support edge computing and design better, more efficient and durable computing solutions. Comparisons and detailed research are used to highlight what makes each technology different and how they can be combined, giving a complete view of future computer architecture.

Index Terms - Quantum Computing, Neuromorphic Systems, Reconfigurable Architectures, Edge Computing, CMOS Scaling, FPGAs and AI-Driven Architecture

Face Recognition Based Attendance System with Presence Duration

Pages: 23-32 (10) | [Full Text] PDF (809K)

Muhammad Jafar, Usman Ali, Muhammad Junaid Arshad

Department of Computer Science, University of Engineering and Technology, Lahore, Pakistan
Department of Information Sciences, University of Education, Lahore, Pakistan

Abstract - In an era where educational institutions struggle with unreliable manual attendance systems and superficial participation tracking, our research addresses the critical need for technology that not only automates presence verification but also promotes meaningful classroom engagement by continuously monitoring student attendance throughout entire sessions. Unlike traditional methods, our system continuously monitors multiple faces simultaneously during live video capture, with attendance finalized when instructors close the camera. The technology employs a time-based presence threshold, marking students as present only when they attend more than 75% of the total class duration, effectively preventing superficial attendance and encouraging meaningful participation. Leveraging Dlib’s robust recognition library alongside advanced deep learning models, our methodology achieves impressive accuracy metrics during train- test splitting on our dataset: 92% face detection accuracy using Histogram of Oriented Gradients (HOG) and 96% recognition precision with our optimized K-Nearest Neighbors (KNN) model implementation when testing on single face images. The multi- face processing algorithm enables simultaneous identification of all students in the same classroom context, achieving 88% overall attendance marking accuracy in real-world implementation. University attendance management now benefits from instant statistics functionality through an easy-to-use interface which advances institutions by converting administrative work into quality student engagement tools.

Index Terms - Face Recognition, Attendance System, Real-Time Monitoring, Face Detection, OpenCV and Dlib

Leveraging Edge Intelligence Integration into Software for Real-Time Decision-Making and Data Analysis, while Mitigating the Utilization of Cloud Processing

Pages: 33-40 (8) | [Full Text] PDF (438K)

Muneeb ur Rehman Khan, Rehman Kabir

Department of Data Science, University of Engineering and Technology, Lahore, Pakistan

Abstract - The integration of edge computing with artificial intelligence facilitates real-time data processing and decisionmaking at the network’s edge. This review offers an in-depth examination of the fundamental concepts and architecture of edge intelligence, its applications across various sectors, and the associated challenges. Additionally, it outlines potential future research avenues that include the integration of 5G technology, federated learning, and energy-efficient methodologies. For example, edge intelligence has the potential to transform industries such as healthcare, automotive, smart cities, and industrial IoT by bridging cloud computing with local processing.

Index Terms - Component, Formatting, Style, Styling and Insert

Advancements in Iot Security: A Comprehensive Review of Machine Learning and Deep Learning-Based Anomaly Detection with Real-Time Constraint

Pages: 41-48 (8) | [Full Text] PDF (417K)

Irfan Ahmed

Department of Data Science, University of Engineering and Technology (UET), Lahore, Pakistan

Abstract - The exponential growth of the Internet of Things (IoT) has significantly improved smart environments, yet it has also introduced critical cybersecurity challenges due to the limited computational resources of IoT devices and their susceptibility to sophisticated attacks. Conventional security approaches often fail to address real-time and evolving threats. This paper provides a comprehensive review of machine learning (ML) and deep learning (DL) techniques for anomaly detection in IoT networks, with a particular focus on methods suitable for real-time applications. We evaluate the effectiveness of diverse models including CNNs, RNNs, LSTMs, and hybrid approaches, alongside novel strategies such as federated learning, reinforcement learning, and lightweight AI models. Despite advances, challenges persist in achieving low-latency detection, model interpretability, and robustness against adversarial attacks. We discuss optimization strategies, scalability issues, and the integration of blockchain and explainable AI for enhancing trust and resilience. This work aims to guide researchers and practitioners toward developing efficient, scalable, and real-time anomaly detection systems tailored for IoT environments.

Index Terms - Internet of Things (IoT), Anomaly Detection, Machine Learning (ML), Deep Learning (DL), RealTime Security, Federated Learning, Reinforcement Learning and Lightweight AI Models

Volume 16, Issue 4, August 2025

IJCST Managemet

Current Issue

Recent Issues

Author Guidelines

IJCST RSS Feeds

News & Updates