Pre-Analyzed Resource and Time Provisioning in Distributed Real-Time Systems: An Application to Mobile Robotics
The work is motivated by mobile robotic applications where a team of autonomous robots cooperate in achieving a goal, e.g., using sensor feeds to locate trapped humans in a building on fire. To attain the goal, kinematic Line-Of-Sight (LOS) constraints must be enforced to build interference free communication paths, and precedence, communication and location constrained tasks must be allocated and scheduled across the processing entities in a predictable and scalable fashion. To maintain LOS constraints, symmetric task models are derived and coordinated. To improve the schedulability of these tasks and to make the predictable scalability analysis possible in the presence of precedence, communication and location constraints, both task dependencies and workloads are examined with the objective of minimizing communication costs and workload of each processor.