An Advanced Trust Model in Social Internet of Things: Simulating Multidimensional Relationships Using Watts–Strogatz Random Graphs

Abstract

Trust is recognized as a vital component in ensuring secure and stable interactions within Social Internet of Things (SIoT) networks. This study introduces an advanced model for simulating multidimensional trust relationships in SIoT, which is designed based on Watts–Strogatz (WS) random graphs and successfully reproduces the real topological features of SIoT networks with high fidelity. The proposed model incorporates a variety of relationship types such as Co-Location-Based Relationships (CLOR), Ownership-Based Relationships (OOR), Social Relationships (SOR), and Popularity-Based Relationships (POR), and integrates key attributes including spatial density, interaction frequency, owner reliability, and co-presence time to deliver a flexible and scalable structure. Analysis of the results indicates that the model performs significantly well in replicating topological metrics such as average path length, clustering coefficient, and average degree. For instance, in OOR and SOR graphs, the clustering coefficients reached values of 0.9 and 0.7 respectively, and in the CLOR graph, the average path length was limited to 2.4. Furthermore, in the POR graph, the average degree was consistently maintained at a stable value of 120. A comparison between the proposed model and traditional models such as Erdős–Rényi (ER) and Barabási–Albert (BA) graphs reveals that the use of advanced random graphs alongside the integration of additional trust-related features significantly enhances the accuracy, flexibility, and analytical capability of SIoT network behavior. In addition, the application of gradient descent-based optimization algorithms for fine-tuning model parameters ensures the efficiency and structural balance of the model, thereby positioning it as an effective and scalable solution for the analysis and development of Social Internet of Things networks.