A Novel Opportunistic Social Network Routing Method Based on Social-Aware Mechanism and Multiobjective Optimization
Huan Jia, Peng Li · IEEE Sensors Journal · 2025
Opportunistic network is classified as mobile networks that do not rely on traditional network infrastructure. It utilizes random encounters among mobile nodes to facilitate message transmission and employ social network attributes derived from the nodes’ characteristics to assist in message forwarding. Current routing algorithms in Opportunistic networks focus on community clustering and the hierarchical structure of nodes, which tends to ignore the social attributes of the nodes themselves. This paper utilizes the social profiles of nodes, in conjunction with encounter information during actual mobility, to establish a dynamic set of neighboring nodes and to compute the similarity of their social attributes. Furthermore, these parameters are integrated into a multi-objective optimization algorithm, with appropriate improvements made to the traditional optimization algorithm. This enables the periodic updating of the set of optimal and suboptimal relays for each message during the node movement process. To maximize the potential of each encounter opportunity, we develop forwarding strategies specifically designed for interactions with various node types. Simulation experiments utilizing the INFOCOM2006 dataset illustrate that the proposed forwarding strategy substantially improves the message delivery success rate while maintaining low net overhead and latency, particularly in environments characterized by limited network resources. Compared with the state-of-the-art routing algorithms MMFD and NCTA, our proposed algorithm demonstrates substantial performance improvements across various experimental conditions. Specifically, as cache capacity increases, our approach enhances the average message delivery success rate by 51.80% and 13.75% respectively, while simultaneously reducing the average message drop ratio by 13.5% and 20%. Furthermore, the algorithm maintains superior performance in terms of network overhead and message delivery latency metrics.