Research on Cooperative Localization Method in Dynamic Networks

Cooperative localization technology can improve the overall positioning accuracy of the system and nodes through spatial and temporal cooperation among multiple nodes, and it has a broad application demand in the areas that cannot be covered by the inherent navigation and positioning facilities. However, in the actual application process, the nodes in the system are always in motion, and often face the problem of scenario switching. This may lead to changes in roles between anchor nodes and ordinary nodes, the disappearance or emergence of ranging values between nodes, as well as nodes randomly accessing or exiting the network. These circumstances can result in a decrease in the accuracy of cooperative localization algorithms or even cause them to malfunction. In order to address the above problems, this paper carries out the research on cooperative localization method, and proposes the filtering method of adaptive adjustment of observation error covariance matrix and the fusion of the federated filtering method, which enables the cooperative localization system to work normally in the case of the change of anchor nodes and ordinary nodes among mobile nodes, the disappearance or emergence of ranging values between nodes, and the random access and exit of nodes, etc. Simulation results show that the cooperative localization system composed of five ordinary nodes achieves a localization accuracy of 3. 6 m, which represents a 197. 2% improvement compared to the dead reckoning localization accuracy of 10. 702 meters. , and verifies the validity of the proposed dynamic cooperative localization method through the simulation setting.