基于指数平滑预测模型的移动节点定位算法
A Localization Algorithm Based on Index Smoothing Prediction Model for Mobile Nodes

针对无线传感器网络移动节点定位精度不足的问题，提出了一种基于指数平滑预测模型的移动节点定位算法，该算法利用指数平滑预测模型获得节点的速度信息和方向信息，然后根据节点的历史轨迹时间序列，应用H指数进一步精确采样区域，从而获得更接近于未知节点位置的高质量样本点． 此外，通过高质量样本点构建“有效样本点矩形盒子”，减少了节点采样次数． 仿真结果表明，改进算法在不同的运动速度、锚节点密度等条件下，均提高了定位精度，表现出了较好的性能．
： To solve the problem of low location accuracy of mobile nodes in Wireless Sensor Network, an algorithm based on index smoothing prediction model is proposed in this paper. By applying the Smoothing Prediction Model in the algorithm, the information of nodes’ speed and direction is obtained, and according to the nodes’ historical trajectory the accuracy of sampling area through the application of Hurst Exponent is further improved. By this way, the above information to predict the unknown node’s next state is used and the “high quality sample” is obtained. Moreover, by making use of the “high quality sample” the “valid sample rectangular box” is constructed, which can decrease the number of sampling. The simulation result shows that the proposed algorithm has improved the location accuracy in the context of different time, velocity, radius and anchor density, showing good performance