A Time-Sensitive Attribute-Oriented Multi-Message Dynamic Scheduling Method in ROS2-RTPS

Cheng Hui Ji, Chuang Ye Hu, Zhandong Liu, Nan Ding

Abstract: With the development of the era of robot 4.0, the application scale and scope of cloud-edge-end fusion robots are expanding, and the robots that transmit data through the network have higher real-time performance and reliability requirements. ROS2, which uses Data Distribution Service (DDS) as the communication middleware, ensures low latency and high reliability when robotic applications interact with large amounts of data in real-time over the network. However, the optimization of ROS2 for the simultaneous transmission of diverse data, multi-modal data, and data with different priorities is not perfect. Therefore, based on the theory of hybrid switching system and the transport protocol-RTPS underlying DDS, this paper designs the ROS2-RTPS multi-type message scheduling model as a hybrid switching system scheduling model including a priority-based subsystem and a time-based subsystem and proposes a time-sensitive attribute-oriented multi-message dynamic scheduling method for ROS2-RTPS. The proposed method considers both the priority and time of different types of messages in ROS2 applications. Finally, the simulation results show that compared with the existing methods, the proposed method can meet the requirements of high-priority data packet transmission, reduce the packet loss rate, and improve the real-time performance and reliability in the scenarios of discrete data, continuous data, and different priority data transmission.