Volume 16, Issue 2, May 2025

Integration of Serverless Computing in IoT: A Systematic Understanding of Architecture and Technical Challenges

Farzain Ud Din Kirmani, M.J. Arshad, Bareeha Muqadas, Ihsan Ali

Department of Computer Science, University of Engineering & Technology, Lahore, Pakistan
Department of Physics, NFC Institute of Engineering & Technology, Multan, Pakistan
Imperial College of Business Studies, Lahore, Pakistan

Abstract - IoT (internet of Things) is an innovative paradigm which paves the way of new services by enabling the seamless transformation of physical world objects to virtual but intelligent objects. Recent advancements of cloud computing in terms of efficiently managing the processing, storage and networking has broaden its domain to IoT as well. However, there are some managing tasks regarding servers and its traffic set some limits in terms of administration. Therefore, serverless computing plays it significant role in automating the server management and providing continuous serverless environment to the IoT devices. It has been observed that there is an absence of clear and concise architecture which integrates serverless computing with IoT. Moreover, with the advent of integrating serverless computing and IoT, there are number of benefits and applications which participates in the whole paradigm but there are some technical as well as framework challenges which also show their inclusion in the integration. This research paper presents an insight of integrating architecture and technical challenges which are related to the integrated domain of serverless computing and IoT.

Index Terms - Serverless Computing, Internet of Things (IoT), Function as a Service, Technical Challenges and Integrated Architecture

Impact of Accrual Basis of Accounting System on the Quality of Financial Reporting: Accountant’s Perspective

Pages: 7-22 (16) | [Full Text] PDF (455K)

Janet A. Chumba

Finance Department, Rongo University, P.O. Box 103-40404 Rongo, Kenya

Abstract - Background: While some researchers dispute the significance of accrual-based accounting, the growing popularity of this accounting approach is premised on the inherent strengths in terms of quality of financial reporting. In the course of implementing accrual accounting, there is need to confront inherent challenges such as such as lack of clear policy framework, inadequate professional accountants as well as inaccessibility to relevant ICT tools. The motivation this review is therefore to bring to the fore both the strengths as well as the gaps in the implementation of accrual accounting across different entities.Broad Objective: To systematically map the growing adoption of accrual accounting in financial reporting from the global, regional and local perspectives. Methodology: Systematic literature review of articles on accrual accounting. From the initial two hundred and fifty-seven (257) articles, the review on narrow down to (129) articles. The inclusion/ exclusion criteria were informed by the relevance of the available articles in terms of their focus on thematic area of accrual accounting. The review narrows down to bibliographic databases comprising Research Gate, Emerald Journals as well as Google Scholar. Eventually, 129 articles published between the years 1999 and 2024 were included in the review. Findings: Accrual accounting significantly improves the quality of financial reporting by enhancing transparency, accountability, and decision-making. It enables better asset-liability management and performance measurement, particularly in public sector institutions. However, challenges such as technological constraints, subjective judgments, and resistance to change hinder its implementation. Developing countries face additional barriers, including inadequate funding and lack of skilled personnel, despite progress in countries like Kenya, Nigeria, and Rwanda.....

Index Terms - Financial Reporting, Financial Statements, Accrual-Based Accounting, Public Sector and Accounting Standards

A Comprehensive Review of Energy-Efficient Communication Strategies for IoT Networks Across Smart Cities

Pages: 23-31 (9) | [Full Text] PDF (614K)

Ammara Ejaz

Data Science Department, University of Engineering and Technology, Lahore

Abstract - The development of smart cities relies heavily on advanced communication networks that are energy-efficient and reliable. This paper delivers a complete overview of energy-efficient communication protocols for IoT applications within smart cities, focusing on technologies such as ZigBee, LoRaWAN, BLE, 5G, NB-IoT, and LTE-M. It explores wireless energy transfer, federated learning, edge and fog computing, reconfigurable intelligent surfaces, and blockchain systems. The study aims to enhance energy efficiency while ensuring low latency and interoperability for urban services like Smart Grid, Intelligent Transportation, environmental monitoring, and healthcare. A comparative analysis of the challenges faced, including scalability and security versus energy trade-offs, is provided. Additionally, future trends such as AI-driven energy optimization and 6G communication paradigms are discussed to outline pathways for developing sustainable smart cities. This research serves as a valuable resource for those interested in designing efficient IoT communication frameworks for urban environments.

Index Terms - Internet of Things (IoT), Smart Cities, Energy Efficient Communication, Wireless Sensors Networks (WSNs), 5G Networks, 6G Networks, ZigBee, LoRaWAN, Bluetooth Low Energy (BLE), Fog Computing, Edge Computing, Federated Learning and Wireless Power Transfer (WPT)

Role of Artifical Intelligence and Machine Learning in Network Secuirty- ReviewRole of Artifical Intelligence and Machine Learning in Network Secuirty- Review

Pages: 32-37 (6) | [Full Text] PDF (487K)

Muhammad Furqan, M. Junaid

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

Abstract - With threats of cybersecurity becoming increasingly complex and scaling up in their nature, Artificial Intelligence (AI) and Machine Learning (ML) have been adopted as key technologies to reinforce the existing network security mechanisms. By using AI and ML, security systems can identify, evaluate, and respond to threat in real time at a higher level of accuracy and speed than traditional rule based methods. In this review, we delve into the dynamic contribution of AI and ML to securing the modern networks through their usage in the application of intrusion detection, anomaly detection, malware classification, and predictive threat intelligence. Using massive datasets and constantly soaking up attacks’ patterns, AI security solutions enable organizations to stay ahead of APTs and zero days. Although these intelligent systems are increasingly being integrated, there are challenges to models explain ability, adversarial attacks, and data privacy. Additionally, this article discusses the emerging trends such as automated incident response, AI driven security orchestration, and ethical AI in the context of cybersecurity. We synthesize the current research and use case of AI & ML employed in industry for enhancing network security, resilience and safeguarding critical digital assets.

Index Terms - AI in Network Security, Machine Learning Cyber Defense, Intrusion Detection Systems, Anomaly Detection and AI-Driven Threat Intelligence

Vision-Based Human Activity Recognition Uses A Deep Learning Approach

Pages: 38-43 (6) | [Full Text] PDF (532K)

Pranta Kumar Sarkar, Moskura Hoque, Mostofa Kamal Nasir

Computer Science and Engineering, Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj, Bangladesh
Department of Computer Science and Engineering Mawlana Bhashani Science and Technology University Santosh, Tangail-1902, Bangladesh

Abstract - In today's world, daily life increasingly depends on vision-based advanced technologies, which enhance the reliability and convenience of human lifestyles. Among these technologies, vision-based Human Activity Recognition (HAR) stands out as a comprehensive and challenging field of study, with broad exploration and practical applications. HAR systems are designed to identify diverse human actions under varying environmental conditions.Vision-based activity recognition plays a crucial role in a wide range of applications, including user interface design, robot learning, security surveillance, healthcare, video searching, abnormal activity detection, and human-computer interaction. This study focuses on recognizing various human activities in real-world settings, highlighting the importance of consistency and credibility in the results.To achieve this, data was collected from multiple sources and processed using three distinct models—Convolutional Neural Network (CNN), VGG-16, and ResNet50—to identify the most effective approach for activity recognition. Among these, a highly optimized CNN model was further evaluated for its ability to capture human activity features in specific video sequences.The training, validation, and testing phases utilized a comprehensive dataset comprising 56,690 images. Remarkably, the proposed system achieved an impressive accuracy of 96.23%, demonstrating its effectiveness in human activity recognition.

Index Terms - Computer vision, Activity Recognition, Accuracy, Deep learning and High Performance

A Review of IoT Security Challenges in Transport and Network Layers: Technologies, Limitations, and Future Directions

Pages: 44-49 (6) | [Full Text] PDF (438K)

Haider Ali Adeel, Sehar Hafeez

Institute of Data Science, University of Engineering and Technology, Lahore

Abstract - The Internet of Things (IoT) is transforming industries by enabling seamless communication among smart devices. However, securing the transport and network layers remains a pressing concern due to vulnerabilities that can lead to unauthorized access, data breaches, and energy inefficiencies. This review highlights the primary challenges in securing these layers—such as scalability, energy consumption, and real-time threat detection—and evaluates the effectiveness of emerging technologies including machine learning (ML), blockchain, and artificial intelligence (AI)-based frameworks. Despite promising advances, issues like dataset bias, insufficient hardware security, and limited real-world implementations continue to hinder progress. The paper concludes by outlining future research directions, emphasizing the need for quantum-resistant cryptographic methods, standardized datasets, and optimized deep learning approaches to build resilient IoT infrastructures.

Index Terms - IoT Security, Transport Layer, Network Layer, Machine Learning, Blockchain and Quantum Cryptography

Machine Learning-Based Routing in Ad-Hoc Networks: A Comparison with AODV and DSR

Pages: 50-55 (6) | [Full Text] PDF (412K)

Fahad Amer

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

Abstract - Ad-hoc networks do not require any infrastructure and update their layout continually. Toughening the routing process results from moving nodes, limited energy and variable link quality. Many systems rely on the common AODV and DSR reactive protocols, but their performance is low when it comes to adapting, saving energy and scaling up. All ML-based routing protocols developed in the years 2020 to 2025 are presented in this paper and assessed using AODV and DSR using measures like packet delivery, delay, energy consumption and scaling. Furthermore, the paper explores the use of deep reinforcement learning, federated learning and graph neural networks in predictive and adaptive routing. In addition, it summarizes the problems faced when putting ML-driven routing into practice and lists future approaches and topics worth exploring in large real-time ad hoc networks.

Index Terms - Internet of Thing (IoT), Ad-Hoc Networks, Machine Learning, Routing Protocols, AODV, DSR, Deep Reinforcement Learning, Federated Learning, GNN, MANET and Wireless Networks

Volume 16, Issue 2, May 2025

IJCST Managemet

Current Issue

Recent Issues

Author Guidelines

IJCST RSS Feeds

News & Updates