Deep Learning and Optimization Approaches in Trusted Cloud-Enabled IoT Networks Using Blockchain and Siamese Heterogeneous Convolutional Neural Networks: A Review

Abstract

The rapid expansion of cloud-enabled Internet of Things (IoT) ecosystems has introduced unprecedented opportunities for real-time data processing, automation, and intelligent decision-making. However, the integration of heterogeneous devices, distributed architectures, and large-scale data flows also exposes IoT systems to critical security, privacy, and trust-related challenges. Recent advancements in deep learning and blockchain technologies have emerged as promising solutions to address these issues. This review paper presents a comprehensive analysis of deep learning and optimization approaches for building trusted cloud-enabled IoT networks, with a particular focus on blockchain integration and Siamese heterogeneous convolutional neural networks (SHCNNs). Blockchain enhances data integrity, decentralization, and transparency, while deep learning models enable intelligent threat detection, anomaly recognition, and adaptive decision-making.

Furthermore, Siamese architectures provide similarity-based learning capabilities that improve detection accuracy in dynamic and unknown attack scenarios. The paper systematically reviews recent literature (2020–2023), compares existing methodologies, and identifies research gaps. Key challenges such as scalability, computational overhead, energy efficiency, and explainability are also discussed. The findings highlight that hybrid frameworks combining blockchain with deep learning significantly improve trust, robustness, and security in IoT-cloud environments. Finally, future research directions are proposed to advance intelligent, scalable, and secure IoT infrastructures.