【摘要】 无线物联网感知层的多源传感数据量较多，差异化较大，在数据共享阶段，由于大差异性导致存在异常数据的问题，提出了针对无线物联网感知层的多源传感数据实时共享方法。结合滑动窗口、ARIMA模型和K-means算法剔除多源传感数据中的异常数据，采用改进支持度函数加权算法融合数据后，依据共享条件引入ERDQN算法实现数据实时共享。仿真结果表明，所提方法能量消耗低于0.302 J、数据延迟均值低于0.388 s、数据到达率高于0.585、节点生存数目高于71个，解决了无线物联网感知层多源传感数据共享实时性较差的问题。更多还原

【Abstract】 Because of the large amount and great difference of multi-source sensing data in the sensing layer of wireless Internet of things, in the data sharing stage, there are abnormal data due to large difference. In view of this issue, a real-time sharing method of multi-source sensing data for the sensing layer of wireless Internet of things is proposed. Combined with sliding window, ARIMA model and K-means algorithm, the abnormal data in multi-source sensor data are eliminated. After the data is fused by using improved support function weighting algorithm, ERDQN algorithm is introduced according to the sharing conditions to realize real-time data sharing. The simulation results show that the energy consumption of the proposed method is less than 0.302 J,the average data delay is less than 0.388 s, the data arrival rate is higher than 0.585,and the number of nodes is higher than 71,which solves the problem of poor real-time sharing of multi-source sensing data in the sensing layer of wireless Internet of things.更多还原