IDMO: A Multi-Stage Optimized Deep Learning Framework for Efficient and Scalable IoT Big Data Analytics

International Journal of Electronics and Communication Engineering

© 2025 by SSRG - IJECE Journal

Volume 12 Issue 11

Year of Publication : 2025

Authors : Ch.Ellaji, R.S. Ponmagal, V. Saritha

10.14445/23488549/IJECE-V12I11P102

How to Cite?

Ch.Ellaji, R.S. Ponmagal, V. Saritha, "IDMO: A Multi-Stage Optimized Deep Learning Framework for Efficient and Scalable IoT Big Data Analytics," SSRG International Journal of Electronics and Communication Engineering, vol. 12,  no. 11, pp. 8-20, 2025. Crossref, https://doi.org/10.14445/23488549/IJECE-V12I11P102

Abstract:

The fast growth of Internet of Things (IoT) ecosystems created massive data volumes, which show both diverse characteristics and sporadic connectivity while maintaining strict privacy rules. The current technological environment requires instantaneous analytics processing at the network boundary using conventional deep learning systems alongside federated learning solutions, which usually encounter processing limitations, model complexity limitations, and intense data transfer downtime. The research presents IDMO as a novel three-stage framework that optimizes deep neural networks in IoT-FL systems through model compression integration with communication efficiency techniques. The IDMO pipeline comprises three main components: (i) a Bi-Level Utility (BLU)-guided structured pruning method that adaptively eliminates redundant filters while maintaining crucial feature representations; (ii) a Selective-Importance-driven Joint Fine-tuned Optimisation (SI-JFO) quantisation approach that employs metaheuristically guided, non-uniform encoding according to weight significance and gradient sensitivity; and (iii) a Niblack-Adaptive Thresholding (NA-T)- based selective update mechanism that reduces communication costs by exclusively transmitting significant local parameter changes in federated environments. Tests conducted on CIFAR-10 data sets confirm IDMO achieves 91.43% accuracy and reduces model size to 5.20 MB from 1.49 MB with a 71.3% decrease, while FLOPs drop to 53.5% compared to standard FL protocols. This leads to 65% lower communication expenses. Merging these improvements does not impact inference performance, yet makes IDMO operational in edge settings with limited resources. The research outcomes demonstrate that IDMO technology can transform IoT analytics capabilities because it delivers efficient edge-processing of adaptive deep learning models that protect user privacy.

Keywords:

IoT, Deep Learning, Pruning, Quantisation, Federated Learning, Maxout Networks, Edge Computing.

References:

[1] Shahnawaz Ahmad et al., “Deep Learning Models for Cloud, Edge, Fog, and IoT Computing Paradigms: Survey, Recent Advances, and Future Directions,” Computer Science Review, vol. 49, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[2] Saif Saad Ahmed et al., “Intelligent Decision Making in IoT-based Enterprise Management through Fusion Optimisation with Deep Learning Models,” Fusion: Practice and Applications, vol. 11, no. 2, pp. 8-20, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[3] Omair Rashed Abdulwareth Almanifi et al., “Communication and Computation Efficiency in Federated Learning: A Survey,” Internet of Things, vol. 22, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[4] Ummar Bibi et al., “Advances in Pruning and Quantization for Natural Language Processing,” IEEE Access, vol. 12, pp. 139113-139128, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[5] Xing Chang et al., “Efficient Federated Learning via Adaptive Model Pruning for Internet of Vehicles with a Constrained Latency,” IEEE Transactions on Sustainable Computing, vol. 10, no. 2, pp. 300-316, 2005.
[CrossRef] [Google Scholar] [Publisher Link]
[6] Rui Chen, Xiaoyu Chen, and Jing Zhao, “Private and Utility Enhanced Intrusion Detection based on Attack Behavior Analysis with Local Differential Privacy on IoV,” Computer Networks, vol. 250, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[7] Zhixiong Chen et al., “Adaptive Model Pruning for Communication and Computation Efficient Wireless Federated Learning,” IEEE Transactions on Wireless Communications, vol. 23, no. 7, pp. 7582-7598, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[8] Pengzhou Cheng et al., “LSF-IDM: Deep Learning-based Lightweight Semantic Fusion Intrusion Detection Model for Automotive,” Peer-to-Peer Networking and Applications, vol. 17, pp. 2884-2905, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[9] Mohamed Amine Ferrag et al., “Federated Deep Learning for Cyber Security in the Internet of Things: Concepts, Applications, and Experimental Analysis,” IEEE Access, vol. 9, pp. 138509-138542, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[10] MohammadNoor Injadat et al., “Multi-Stage Optimized Machine Learning Framework for Network Intrusion Detection,” IEEE Transactions on Network and Service Management, vol. 18, no. 2, pp. 1803-1816, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[11] Yu Ji, and Lan Chen, “FedQNN: A Computation–Communication-Efficient Federated Learning Framework for IoT with Low-Bitwidth Neural Network Quantization,” IEEE Internet of Things Journal, vol. 10, no. 3, pp. 2494-2507, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[12] Fazal Muhammad Ali Khan et al., “Advancing IIoT with Over-the-Air Federated Learning: The Role of Iterative Magnitude Pruning,” IEEE Internet of Things Magazine, vol. 7, no. 5, pp. 46-52, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[13] Mohamad Khayat et al., “Empowering Security Operation Center with Artificial Intelligence and Machine Learning-A Systematic Literature Review,” IEEE Access, vol. 13, pp. 19162-19197, 2005.
[CrossRef] [Google Scholar] [Publisher Link]
[14] Rudrani Maity et al., “Explainable AI based Automated Segmentation and Multi-Stage Classification of Gastroesophageal Reflux using Machine Learning Techniques,” Biomedical Physics & Engineering Express, vol. 10, no. 4, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[15] Debasmita Mishra et al., “Light Gradient Boosting Machine with Optimized Hyperparameters for Identification of Malicious Access in IoT Network,” Digital Communications and Networks, vol. 9, no. 1, pp. 125-137, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[16] Ammar Nasif, Zulaiha Ali Othman, and Nor Samsiah Sani, “The Deep Learning Solutions on Lossless Compression Methods for Alleviating Data Load on IoT Nodes in Smart Cities,” Sensors, vol. 21, no. 12, pp. 1-27, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[17] Sukanya Pechetti, and Battula Srinivasa Rao, “Optimized MobileNetV3: A Deep Learning-based Parkinson’s Disease Classification using Fused Images,” PeerJ Computer Science, vol. 9, pp. 1-24, 2003.
[CrossRef] [Google Scholar] [Publisher Link]
[18] Pavana Prakash et al., “IoT Device Friendly and Communication-Efficient Federated Learning via Joint Model Pruning and Quantization,” IEEE Internet of Things Journal, vol. 9, no. 15, pp. 13638-13650, 2022.
[CrossRef] [Google Scholar] [Publisher Link]
[19] Mahmoud Ragab et al., “Robust DDoS Attack Detection using Piecewise Harris Hawks Optimizer with Deep Learning for a Secure Internet of things Environment,” Mathematics, vol. 11, no. 21, pp. 1-18, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[20] Parjanay Sharma et al., “Role of Machine Learning and Deep Learning in Securing 5G-Driven Industrial IoT Applications,” Ad Hoc Networks, vol. 123, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[21] Boyu Wang et al., “AI-Enhanced Multi-Stage Learning-to-Learning Approach for Secure Smart Cities Load Management in IoT Networks,” Ad Hoc Networks, vol. 164, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[22] Yifu Wu et al., “DDLPF: A Practical Decentralized Deep Learning Paradigm for Internet-of-Things Applications,” IEEE Internet of Things Journal, vol. 8, no. 12, pp. 9740-9752, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[23] Wenyuan Xu et al., “Accelerating Federated Learning for IoT in Big Data Analytics with Pruning, Quantization and Selective Updating,” IEEE Access, vol. 9, pp. 38457-38466, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[24] Jingren Zhou, Xin Hong, and Peiquan Jin, “Information Fusion for Multi-Source Material Data: Progress and Challenges,” Applied Sciences, vol. 9, no. 17, pp. 1-18, 2019.
[CrossRef] [Google Scholar] [Publisher Link]