Abstract: Wireless sensor network is an important technique that is used in industry and agriculture as a monitoring network.In large-scale monitoring applying condition，wireless sensor network has the characteristics of ultra-large-scale，ultra-low-cost and complex topological changing which results in its limited transmission.It brings the network protocol rigorous requirements.The existing protocol algorithms，such as CNS，SPT and GIT，don’t consider these characteristics roundly，so they can’t be applied directly.An equally effective routing algorithm based on fine-grain gradient（FGLRA） is proposed.Furthermore，based on routing theoretical analysis and simulation，the simulation of the FGLRA is completed from small-scale to ultra-large-scale.Meanwhile，FGLRA is compared with aggregation tree routing and location-assisted routing.Simulation shows，FGLRA routing network can effectively reduce energy consumption，uniform network load and extend network life cycle，and it is especially suitable for large-scale monitoring data transmission.

摘要： 针对大规模无线传感器网络在其传输受限的应用环境下，提出一种基于精细化梯度层次场的有效路由算法（FGLRA）。该算法通过邻居节点集中低层次节点数来刻画节点在层次内的边界位置，实现精细化梯度层次场要求。在此基础上，描述了路由算法的总体框架，并对边界位置值确定、下一跳节点集选择、路由转发机制及数据回传4个关键部分进行了详细分析设计。最后，通过不同规模环境下的仿真模拟，并与融合树、位置辅助等路由算法进行了相关比较。结果表明，该路由算法能有效降低网络能耗、均匀网络负载、延长网络生命周期。