The fast pace of Internet of Things (IoT) ecosystems has come with a major problem of secure task offloading and efficient routing, especially when the network is dynamic and its resources are limited. Traditional optimization and learning methods are usually limited by premature convergence, poor trust management, slowness and excessive energy usage, which limit their use in large-scale IoT systems. The solution to these problems is important in facilitating scalable and reliable IoT infrastructures. The paper will provide a secure and energy-efficient IoT offloading and routing framework that combines blockchain-enabled trust management and a genetic-inspired deep neural network optimization approach. The given strategy focuses on the compromise in the decision-making process by collectively looking at reliability, network lifetime, and communication efficiency, but remains flexible to the heterogeneous IoT settings. Mechanisms of blockchains are used to increase trust, transparency, and integrity of data in offloading processes whereas diversity-sensitive genetic optimization assists stable and effective learning behaviour across different network states. The effectiveness of the suggested framework is tested based on large-scale experiments involving simulations and compared to different existing models of IoT offloading and routing. Numerical data show that the main indicators of evaluation are significantly improved. The proposed scheme has a precision of 94%, accuracy of 93%, recall of 92% and F1-score of 93% as well as a stable extended network life and shorter latency with larger network sizes. These results report better robustness and reliability of decisions compared to baseline approaches. In general, the findings support the conclusion that the proposed framework can provide a useful and scalable solution to secure IoT offloading and routing. Through a capable integration of trust-conscious blockchain functionalities with adaptive learning-based optimization, the model would overcome the most significant constraints of the current protocols and will enable the needs of the IoT systems of the next generation.
Sindhuja et al. (Sun,) studied this question.