Distributed Combat System of Systems Communication Planning Method based on Multi-Agent Deep Reinforcing Learning

In order to solve the problem of communication planning in combat System of Systems (SoS), a distributed communication planning decision model is designed. The model is based on multi-agent deep reinforcement learning algorithm, which can provide decision support for operational nodes and communication nodes in combat SoS, so that each operational node can get long-term stable and efficient communication services. Different from the previous methods, the proposed method models each operational node or communication node as independent agent and uses a distributed mode to make decisions. The agents can make rapid decisions based on their own observation information. The experimental results show that the proposed method can assist the operational nodes and communication nodes making decisions effectively.