Phantom Guard: An MDP-Based Anti-Jamming Defense Strategy for Military 5G Networks

The 5G technology marks a significant milestone in the evolution of mobile networks, offering enhanced capabilities and performance compared to its predecessors. It is also a promising technology for military applications, potentially enabling highly reliable and low-latency operations of autonomous agents. However, commercial 5G radio resource schedulers struggle against radio jamming attacks in critical military 5G scenarios, impacting operational integrity. Current anti-jamming techniques—though effective—are computationally expensive due to utilizing Deep Reinforcement Learning (DRL) strategies that scale with the problem space. Therefore, we propose Phantom Guard, a novel jamming-aware resource allocation strategy based on the Markov Decision Process (MDP), enabling quick adaptation to reactive jamming with limited computation requirements. The proposed MDP-based scheduler has a fixed state and action spaces, learns the jamming pattern from the network, and adjusts resource allocation in real-time —enhancing overall efficiency and resilience against jamming. Mathematical analysis and extensive simulation results validate the performance of the proposed approach and advent a new direction on jamming evasion under limited computation capabilities.