Research on the Design of a Mimetic Defense Architecture for Distributed Situation Awareness Devices Based on Dynamic Heterogeneous Redundancy

To enhance the robustness and security of distributed situation awareness systems in complex adversarial environments, this paper proposes a mimetic defense architecture based on dynamic heterogeneous redundancy. The architecture integrates heterogeneous hardware platform deployment, deep deterministic policy gradient (DDPG) scheduling, PBFT consensus mechanism, and blockchain-based control auditing to construct a three-layer defense system comprising perception, control, and execution. By deploying redundant nodes on various processors such as ARM, x86, and RISC-V, and dynamically scheduling task paths through reinforcement learning, the system effectively disrupts the attacker’s modeling logic while improving node fault tolerance and task recovery capabilities. Experiments conducted using real attack traffic validate the system’s performance, demonstrating superior task accuracy, response latency, and resilience to attacks compared to traditional homogeneous redundancy systems. The results highlight the architecture’s strong dynamic adaptability and practical application potential.