Title:Communication-aware Motion Planning for Multi-agent Systems from Signal Temporal Logic Specifications

Abstract:We propose a mathematical framework for synthesizing motion plans for multi-agent systems that fulfill complex, high-level and formal local specifications in the presence of inter-agent communication. The proposed synthesis framework consists of desired motion specifications in temporal logic (STL) formulas and a local motion controller that ensures the underlying agent not only to accomplish the local specifications but also to avoid collisions with other agents or possible obstacles, while maintaining an optimized communication quality of service (QoS) among the agents. Utilizing a Gaussian fading model for wireless communication channels, the framework synthesizes the desired motion controller by solving a joint optimization problem on motion planning and wireless communication, in which both the STL specifications and the wireless communication conditions are encoded as mixed integer-linear constraints on the variables of the agents' dynamical states and communication channel status. The overall framework is demonstrated by a case study of communication-aware multi-robot motion planning and the effectiveness of the framework is validated by simulation results.