The lightweight and trust-sharing mechanism rectifies the issues of data islands, data leakages, and high costs of trust in IoT-based systems. The lightweight and trust-sharing models ensure the trust of data sharing while preserving the privacy level of the entities. With the development of smart cities and malicious device installation, the main problem associated with IoT communication is data sharing and fusion.
In order to avoid data leakage without increasing the cost of creating a trusted IoT environment, a Lightweight and Trusted Sharing Mechanism (LTSM) is developed. It utilizes the advantages of blockchain and federated learning to realize data sharing. Federated learning is a machine learning technique that trains an algorithm by maintaining the local data samples without exchanging them. It ensures data privacy as well as security during data transmission. The nodes are selected and evaluated using credit value and smart contract algorithms, respectively. Moreover, it can improve the quality of federated algorithms.