Access Permission Delegation for Supply Chain of Social Things
Olfa Dallel, Souheil Ben Ayed, Jamel Bel Hadj Taher · 2024
The Supply Chain of Things (SCoT) integrates Internet of Things (IoT) objects into supply chain operations to enable reliable traceability and effective quality verification. However, connection corruption and object mobility are the main challenges to handle. Furthermore, contributors at every stage of the supply chain need to have access to product information in order to update it as needed and make decisions. In practice, predefined access and delegation policies are the common approaches for controlling access and sharing access permissions with others, respectively. However, they cannot be practical solutions in a highly scalable and dynamic environment such as the SCoT. In this paper, we suggest integrating the social network concept into the SCoT in order to facilitate cooperation and ensure ubiquitous access. To control access to product information, we leverage trust Friend of A Friend (FoAF) relationships established between social IoT objects to propose multi-step delegation based on a clustering-Long Short-Term Memory (LSTM) model. This model automates the multi-step delegation operation by predicting the next delegatee. Finally, we train our suggested clustering-LSTM prediction model, and simulation results show that with 125 clusters, our model achieves an accuracy of 88% for training and 77% for validation.