【摘要】 传感器节点能量受限是制约无线传感器网络使用寿命的关键因素,为了节约传感器网络的能量,提出了灰色模型的动态功耗管理(DPM)方法。该方法利用传感器节点上的历史数据应用灰色模型预测未来值,预测过程中可以动态调整预测参数,实现自适应预测,和小波自回归预测算法相比,提高了预测的准确性。基本思想是根据S ink节点上的数据来决定整个传感器网络的工作模式,在下个周期内若传感器节点的观测值不超过预定的阀值则不向S ink节点发送数据,通过缩短传感器节点的工作时间,降低节点间数据传输量来减少传感器网络的能量消耗。理论分析和实验结果表明本文提出的方法无论在预测准确性方面,还是在节约能量方面都是有效的。更多还原

【Abstract】 The energy constraint for sensor nodes is the key factor that limit the life of wireless sensor networks.So an effective method of dynamic power management that based on grey model is proposed to save energy of wireless sensor networks.Historical data gathered of sensor node is used to forecast the future value with the method of grey model,meanwhile the parameters are adjusted automatically in the process of forecast so as to achieve the adaptive forecast.Compared with the algorithm of wavelet and AR,the accuracy of forecasting for grey model is improved.The basic idea is to decide the working pattern of the entire sensor networks by Sink,and in the next period sensor nodes do not send back their results if their observed values are not out of threshold.Reducing energy consumption of the entire sensor networks is by shortening the working hours and reducing transmitted messages between the nodes.Theory analysis and experiment result show that it is effective not only in the forecasting accuracy but also in the energy efficiency.更多还原