The technology of dynamic map fusion among networked vehicles has been developed to enlarge sensing ranges and improve sensing accuracies for individual vehicles. This paper proposes a federated learning (FL) based dynamic map fusion framework to achieve high map quality despite unknown numbers of objects in fields of view (FoVs), various sensing and model uncertainties, and missing data labels for online learning. The novelty of this work is threefold: (1) developing a three-stage fusion scheme to predict the number of objects effectively and to fuse multiple local maps with fidelity scores; (2) developing an FL algorithm which fine-tunes feature models (i.e., representation learning networks for feature extraction) distributively by aggregating model parameters; (3) developing a knowledge distillation method to generate FL training labels when data labels are unavailable. The proposed framework is implemented in the Car Learning to Act (CARLA) simulation platform. Extensive experimental results are provided to verify the superior performance and robustness of the developed map fusion and FL schemes.
I. INTRODUCTIONThe intelligent networked vehicle system (INVS) is an emerging vehicle-edge-cloud system that accomplishes cooperative perception, map management, planning and maneuvering tasks via vehicle-to-everything (V2X) communication [1]-[3]. Among all the tasks, distributed map management aims to enlarge sensing ranges and improve sensing accuracies for individual vehicles and plays a central role in . While static maps describe stationary objects (e.g., roads, buildings, and trees), dynamic maps emphasize updating information of mobile objects (e.g., pedestrians, cars, and animals) in real time. Fig. 1 shows the architecture of an intelligent networked vehicle system. There are three main steps for dynamic map fusion: (1) local sensing and perception, (2) local map fusion and uploading, and (3) global map fusion and broadcasting.