Asharul Islam
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Asharul Islam has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Abdulrahman Albarrak
Department of Software Engineering, College of Computer Science and Engineering, University of Ha’il, Ha’il, Saudi Arabia
1 shared publication - Anandhavalli Muniasamy
King Khalid University, Saudi Arabia
1 shared publication - Anoop Kumar
Department of Computer Science, College of Computing and Mathematics, Banasthali Vidyapith, Rajasthan, India
1 shared publication - Khaled M. Alalayah
Computer Science Department, Applied College, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia
1 shared publication - Mohamed A. G. Hazber
Department of Information and Computer Science, University of Ha’il, Ha’il, Saudi Arabia
1 shared publication - Mohammad Ashique E Rasool
Research Scholar, College of Computing and Mathematics, Banasthali Vidyapith, Rajasthan, Postal Code 304022, India
1 shared publication
Latest Publications
- Open Access
- 19 September 2025
- Published by Springer Nature in Cluster Computing
- Vol. 28 (13) , 1-23
A blockchain-enabled edge computing framework leveraging artificial neural network and aquila optimization to enhance security and scalability of cloud-based IoT platforms
This work proposes a robust convergence-based approach that integrates blockchain and edge computing, supporting a proactive and adaptive secure-optimized framework to fortify security measures and mitigate unauthorized access in IoT systems. In the developed framework the blockchain technology enforces proactive security by ensuring data integrity, decentralized authentication, and resilient threat prevention. While the fusion of an Artificial Neural Network (ANN) and Aquila Optimizer (AO), enables the framework to enhance the accuracy of threat detection by fine-tuning parameters and optimizing the network’s architecture. The ANN adeptly captures intricate patterns and features from input data, while the AO intelligently adjusts critical hyperparameters to foster swift convergence and achieve peak network performance. This integration empowers a robust and accurate security detection system, enabling effective identification of complex and multifaceted security threats. Through extensive experimental analyses, we validate the effectiveness of the proposed ANN-AO framework, demonstrating a reduction in latency by 587 ms, 1187 ms, 1079 ms, and 460 ms, while improving security detection by 17.00%, 41.00%, 11.00%, and 8.00%, when compared to cutting-edge security methods. Moreover, the results are further verified and validated through statistical and computational analyses.
- 19 September 2025
- Published by Springer Nature in Cluster Computing
- Vol. 28 (13) , 1-23
Open AccessA blockchain-enabled edge computing framework leveraging artificial neural network and aquila optimization to enhance security and scalability of cloud-based IoT platforms
This work proposes a robust convergence-based approach that integrates blockchain and edge computing, supporting a proactive and adaptive secure-optimized framework to fortify security measures and mitigate unauthorized access in IoT systems. In the developed framework the blockchain technology enforces proactive security by ensuring data integrity, decentralized authentication, and resilient threat prevention. While the fusion of an Artificial Neural Network (ANN) and Aquila Optimizer (AO), enables the framework to enhance the accuracy of threat detection by fine-tuning parameters and optimizing the network’s architecture. The ANN adeptly captures intricate patterns and features from input data, while the AO intelligently adjusts critical hyperparameters to foster swift convergence and achieve peak network performance. This integration empowers a robust and accurate security detection system, enabling effective identification of complex and multifaceted security threats. Through extensive experimental analyses, we validate the effectiveness of the proposed ANN-AO framework, demonstrating a reduction in latency by 587 ms, 1187 ms, 1079 ms, and 460 ms, while improving security detection by 17.00%, 41.00%, 11.00%, and 8.00%, when compared to cutting-edge security methods. Moreover, the results are further verified and validated through statistical and computational analyses.
- Open Access
- 11 June 2025
- Published by Slovenian Association Informatika in Informatica
- Vol. 49 (21)
A Systematic Review of Optimization Techniques for Computation Offloading in Mobile Edge Computing: Lyapunov, Convex, Heuristic, Game-Theoretic, and Machine Learning Approaches
Mobile Edge Computing (MEC) has become a key paradigm for reducing latency, energy consumption, and system overhead in computation-intensive applications by enabling task offloading to edge servers. In this paper, we conduct a Systematic Literature Review (SLR) of 70 peer-reviewed studies, with 30 papers coded in detail across five main optimization approaches: Lyapunov, convex, heuristic, game-theoretic, and machine learning methods. The review systematically compares these paradigms in terms of optimization objectives, task assumptions, and evaluation setups. Our analysis reveals that while Lyapunov and game-theoretic approaches rely entirely on simulation (100%), heuristic studies exhibit stronger practical grounding, with 33% including real-world validation. Convex methods demonstrated 10–18% energy savings in heterogeneous MEC scenarios, whereas machine learning approaches showed adaptability under uncertain conditions but lacked real-world testing. Beyond numerical comparison, the review highlights the methodological evolution of MEC optimization, from formal mathematical programming to learning-driven and hybrid approaches. The coding of 30 representative papers provides a quantitative foundation that exposes common assumptions—such as independent task models and simplified simulators—that limit real-world applicability. At the same time, the synthesis identifies promising trends, including reinforcement learning for adaptive decision-making and hybrid frameworks that combine convex optimization with predictive models. These insights underscore the need for future MEC strategies that balance computational efficiency, adaptability, and scalability. By synthesizing results, quantifying trends, and framing open challenges, this SLR provides researchers and practitioners with a comprehensive understanding of the strengths, limitations, and opportunities in MEC computation offloading.
- 11 June 2025
- Published by Slovenian Association Informatika in Informatica
- Vol. 49 (21)
Open AccessA Systematic Review of Optimization Techniques for Computation Offloading in Mobile Edge Computing: Lyapunov, Convex, Heuristic, Game-Theoretic, and Machine Learning Approaches
Mobile Edge Computing (MEC) has become a key paradigm for reducing latency, energy consumption, and system overhead in computation-intensive applications by enabling task offloading to edge servers. In this paper, we conduct a Systematic Literature Review (SLR) of 70 peer-reviewed studies, with 30 papers coded in detail across five main optimization approaches: Lyapunov, convex, heuristic, game-theoretic, and machine learning methods. The review systematically compares these paradigms in terms of optimization objectives, task assumptions, and evaluation setups. Our analysis reveals that while Lyapunov and game-theoretic approaches rely entirely on simulation (100%), heuristic studies exhibit stronger practical grounding, with 33% including real-world validation. Convex methods demonstrated 10–18% energy savings in heterogeneous MEC scenarios, whereas machine learning approaches showed adaptability under uncertain conditions but lacked real-world testing. Beyond numerical comparison, the review highlights the methodological evolution of MEC optimization, from formal mathematical programming to learning-driven and hybrid approaches. The coding of 30 representative papers provides a quantitative foundation that exposes common assumptions—such as independent task models and simplified simulators—that limit real-world applicability. At the same time, the synthesis identifies promising trends, including reinforcement learning for adaptive decision-making and hybrid frameworks that combine convex optimization with predictive models. These insights underscore the need for future MEC strategies that balance computational efficiency, adaptability, and scalability. By synthesizing results, quantifying trends, and framing open challenges, this SLR provides researchers and practitioners with a comprehensive understanding of the strengths, limitations, and opportunities in MEC computation offloading.