基于双层预测机制的传感器网络目标跟踪算法

Dual Prediction-based Algorithm for Target Tracking in Sensor Networks

针对层次型传感器网络的目标跟踪问题,提出了一种基于双层预测机制的目标跟踪算法,由传感器节点执行Bayes估计算法进行目标位置的预测与更新(微观预测);由簇头进行基于曲线运动方程的目标位置预测(宏观预测),并用预测结果对传感器节点所得到的目标位置进行更新,提高了目标位置计算精度。仿真结果表明,算法具有较好的跟踪精度和能量效率。

Aiming at the issue of target tracking in hierarchical sensor networks,an algorithm based on dual prediction mechanism was proposed, which is expected to enhance the tracking accuracy. The sensor nodes perform a Bayes filtering algorithm to predict and update a target position （micro prediction）, and the cluster headers employ a curvilinear movement equation to predict the target position （macro prediction）, which is used to update the target position measured by the sensor nodes. Experiment results demonstrate that the proposed algorithm has better tracking accuracy and energy efficiency.