eCAR: edge-assisted Collaborative Augmented Reality Framework

Abstract In collaborative augmented reality (CAR), data communication that synchronizes virtual objects results in large network traffic and high network latency. CAR applications without continuous data communication for coordinate system alignment exhibit virtual object inconsistencies. However, as the number of collaborative devices increases, In addition, synchronization messages for online virtual object updates experience high latency. To solve this problem, we propose a novel edge-assisted multiuser CAR framework, eCAR, for a large indoor environment. The framework utilizes edge computing to match the coordinate system of a device with less network traffic continuously. Further, we extend the co-visibility graph of the edge server for maintaining the spatial–temporal consistency of virtual objects in neighboring devices by synchronizing a local graph. We evaluated the proposed system quantitatively and qualitatively using a public dataset and physical indoor environment. The eCAR communicates data for coordinate system alignment between the edge server and devices with less network traffic and latency. The proposed system continuously aligns coordinate systems with multiple devices in a large indoor environment and shares augmented reality content. Our system enables users to interact with virtual objects and share augmented reality experiences with neighboring users.