ID 原文 译文
45606 它由大量具有移动能力且会长时间进入睡眠状态的节点自组织而成，可以部署在恶劣环境中执行长期的监测任务，在国防、工业、农业等领域具有广泛的应用前景。 Such networks have wide application prospects in national defense, industry, agriculture and other fields that need long term monitoring in severe environments.
45607 但是，节点的移动和睡眠导致网络拓扑不断发生改变，使节点很难以较少的能耗快速发现其全部的邻居， However, the movement and the sleeping features of nodes lead to constantly change of network topology, which makes the nodes difficult to discover their neighbors quickly.
45608 导致节点无法获得最优的分布式决策结果，影响网络应用的效果。 Therefore, the nodes cannot achieve optimal distribution decisions.
45609 为了解决这个难题，提出一种新的主动式邻居发现算法。 In order to solve this problem, a new proactive neighbor discovery algorithm was proposed.
45610 该算法使网络中的节点在苏醒时主动寻找自己的邻居，避免传统被动式邻居发现中长时间等待所产生的时延。 This algorithm made the nodes in the network take the initiative to find their neighbors when they woke up,and avoided the delay caused by long time waiting in the traditional passive neighbor discovery.
45611 此外，通过对邻居移动速度及距离的预测，快速确定未来下一时刻的邻居集合，在进一步减少时延的同时获得更准确的邻居发现结果。 In addition, by predicting the movement speed and distance of neighbors, the neighbor set at the next moment can be quickly determined, which can further reduce the delay and obtain more accurate neighbor discovery results.
45612 理论分析和实验结果表明，与已有算法相比，所提算法能够在 MLDC-WSN 中以更小的能耗、更低的时延发现全部的邻居。 Theoretical analysis and experimental results show that compared with the existing algorithms, the algorithm can find all the neighbors in MLDC-WSN with less energy consumption and lower delay.
45613 针对无线网络中的数据传输问题，提出一种基于深度 Q 学习（QL, Q learning）的传输调度方案。 To cope with the problem of data transmission in wireless networks, a deep Q learning based transmission scheduling scheme was proposed.
45614 该方案通过建立马尔可夫决策过程（MDP, Markov decision process）系统模型来描述系统的状态转移情况； The Markov decision process system model was formulated to describe the state transition of the system.
45615 使用 Q 学习算法在系统状态转移概率未知的情况下学习和探索系统的状态转移信息，以获取调度节点的近似最优策略。 The Q learning algorithm was adopted to learn and explore the system states transition information in the case of unknown system states transition probability to obtain the approximate optimal strategy of the schedule node.