基于深度强化学习的应急通信网规划方法

Emergency communication network planning method based on deep reinforcement learning

应急通信具有较强的突发性和不确定性,为满足应急通信网规划中灵活快速组网的要求,根据不同层次网络特点,进行网络拓扑结构分层建模描述,应用深度强化学习算法,实现拓扑结构生成,并通过算法优化其生成效率,依据业务特点,按策略分配应急通信网业务资源,实现完整的应急通信网规划,最后通过样例仿真,验证了应急通信网模型及算法的科学性和高效性,为应急通信网的规划提供参考。

Emergency communication has the characteristics of strong sudden and uncertainty, to meet the requirements of flexible network planning, on the basis of different characteristics of diverse layered network to accomplish the modeling description, apply the deep reinforcement learning algorithm to implement network topology planning task, and enhance the generation efficiency through the algorithm optimization. According to the business characteristics and strategies to allocate the emergency communications network business resources, then achieve the complete emergency communications network planning, finally the emergency communications network planning model and the method is verified with high rationality and efficiency by a sample simulation, which provides a certain reference for the emergency communications network planning, which provides a certain reference for the emergency communications network planning.