The Node Optimization Strategy for Sensor-Rich Vehicle Aided Cooperative Positioning

The sensor-rich vehicle (SRV) aided cooperative positioning (CP) method, which introduces several SRVs into a CP system to aid most common vehicles (CoVs), could further improve the CP performance for V2X applications. However, sometimes part of the CoVs may lose their connections with SRVs in complex urban environments, leading to the degradation of their positioning performance. In this work, we propose a two-step SRV aided CP method based on a node optimization strategy, aiming to fully use the high accuracy of the SRVs in unbalanced node connection scenarios. A node optimization strategy is formulated which firstly uses the SRVs to enhance the positioning of important vehicles (ImVs), whose performance would affect most of the vehicles in the system, by constructing an important sub-cluster (ImC) from the system. Then, the corrected position of the ImVs are used to enhance CP for rest CoVs. Based on the node optimization strategy, a two-step positioning algorithm is designed which firstly uses the centralized framework for ImC to fully improve the positioning performance of ImVs, and then uses the distributed framework for other CoVs. Simulations are conducted, except for one single vehicle, the cluster CP statistics are defined to evaluate the performance of the whole system. Results validate that the proposed method outperforms the existing methods, especially for the CoVs far away from the SRVs. In particular, the performance of the whole system improves more significantly when the SRVs are on the edge of the system or there are less SRVs in the system.