Abstract
The drones industry has witnessed great progress, and its systems have many important applications. The free autonomous movement of drones is considered a double-edged sword; it enables a tremendous use cases, at the same time, it makes the design of the communication network among drones, especially the routing protocol, a very delicate matter. Therefore, the research is heading toward achieving joint design that controls the movement in favor of communication. The current work is based on the idea of exploiting the use of drones in conveying data for building digital twin in building digital twin of the drones system itself such that the joint design can be realized. The decision support of the network digital twin is provided by model-based reinforcement learning using dynamic programming and policy iteration algorithm. The digital twin model allows the reinforcement learning model to learn, offline plan, and online re-plan through observing the outcomes of the real environment. This paper describes and implements the proposed solution and compares it to a standard Ad-hoc routing protocol and a model-free reinforcement learning-based routing protocol. The simulation results showed that the proposed solution greatly improves the overall network Quality of Service (QoS).
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call