DDS: Dynamic Data Synchronization for Digital Twins in SAGIVN

The digital twins (DT) empowered space-air-ground integrated vehicular networks (SAGIVN) can efficiently manage data of nodes and provide cross time interactive decisions for nodes, thus significantly enhancing the network service capability. In DT empowered SAGIVN (DT-SAGIVN), vehicles need to comprehensively consider their data synchronization requirements and the resource status of heterogeneous networks to update new parameters to their DTs deployed in the cloud. Therefore, how to continuously access a group of optimal networks in the process of driving from the origin to the destination to complete data synchronization while maximizing the utility of each vehicle has become a key challenge. To solve this problem, we propose a dynamic data synchronization (DDS) scheme in DT-SAGIVN. In this scheme, we first establish a DT empowered network model and a communication model in heterogeneous networks. Then, by considering the dynamic wireless networks, the mobility of vehicles and the characteristics of heterogeneous links, the vehicle data synchronization optimization problem is modeled as a Markov decision process (MDP). After that, based on the vehicle's personal preference for cost and delay, we propose a DQN-based method to solve the MDP problem, so as to make the optimal decision for each vehicle to complete data synchronization. Simulation results show that the proposed scheme can bring the highest utilities for vehicles compared with the traditional schemes.