Dynamic Utility-Based Service Discovery among Socially-enhanced Digital Twins

The rapid proliferation of connected devices in Internet of Things (IoT) ecosystems has created significant challenges for efficient service discovery, with traditional approaches suffering from scalability limitations and excessive resource consumption. This paper presents a novel utility-driven service discovery model for Social Digital Twins (SDTs) environments that optimises resource utilization while maintaining discovery effectiveness. Our proposed model introduces a dynamic utility function that evaluates the contribution of each social relationship among digital twins to the discovery process, enabling nodes to make intelligent forwarding decisions with only local knowledge. By balancing exploitation of high-utility paths with exploration of potentially valuable alternatives, the system adapts to network conditions and service distribution patterns. Extensive simulations on a large-scale SDT dataset demonstrate that our service-specific utility approach achieves 41.7% efficiency, outperforming flooding-based techniques (2.54% efficiency) by a factor of 16.4 while maintaining high service discovery rates. The model shows particular effectiveness in dynamic environments where service availability fluctuates, making it suitable for next-generation IoT deployments where scalability and resource efficiency are critical concerns.