RescueNet: Reinforcement-learning-based communication framework for emergency networking

Abstract A new paradigm for emergency networking is envisioned to enable reliable and high data-rate wireless multimedia communication among public safety agencies in licensed spectrum while causing only acceptable levels of disruption to incumbent network communication. The novel concept of mission policies , which specify the Quality of Service (QoS) requirements of the incumbent networks as well as of the emergency networks involved in rescue and recovery missions, is introduced. The use of mission policies, which vary over time and space, enables graceful degradation in the QoS of the incumbent networks (only when necessary) based on mission policy specifications. A Multi-Agent Reinforcement Learning (MARL)-based cross-layer communication framework, “RescueNet,” is proposed for self-adaptation of nodes in emergency networks based on this new paradigm. In addition to addressing the research challenges posed by the non-stationarity of the problem, the novel idea of knowledge sharing among the agents of different ages (either bootstrapping or selective exploration strategies or both) is introduced to improve significantly the performance of the proposed solution in terms of convergence time and conformance to the mission policies.