An aggregation method for wireless mobile sensing network data

The invention provides an aggregation method for wireless mobile sensing network data. A neural network model is trained by utilizing historical data, time required for performing data aggregation ona node and a surrounding communicable node is predicted according to the model and real surrounding data, and the node with shortest time is selected as a next hop node. Aggregation is performed on data by utilizing the method through each mobile sensing node, only a next hop node where data aggregate is fastest at current time point is selected. The disadvantage that the mobility of the sensing node is not considered in the traditional data aggregation method is overcome, the influence brought from the surrounding environment is also comprehensively considered, and thereby data aggregation time can be effectively saved. 本发明提供种无线移动感知网络数据聚集方法。利用历史数据训练神经网络模型，通过模型和实际周围环境数据来预测节点与周围可通信节点进行数据聚集需要的时间，并选择时间最短的作为下跳节点。每个移动感知节点都采用这种方法对数据进行聚集，都只选择当前时刻数据聚集最快的下跳节点。该方法既克服了传统数据聚集方法中未考虑感知节点移动性的缺点，也综合考虑了周围环境带来的影响，从而有效节省数据聚集时间。