Multi-Agent Based Collaborative Agent Architecture for Battlefield Situation Awareness

The proposed architecture consists of real-time battlefield recognition technology that supports decision-making by fusing individually processed multimodal knowledge information from collaborative neuro-agent. We design an architecture and generate dataset for training the core network. It provides collaborative intelligence-based real-time battlefield situation recognition technologies, which is expected to be applicable to actual battlefield environments or combat training simulator. Simulation results are provided to demonstrate the performance of a multi-agent based collaborative agent architecture for battlefield situation awareness.