应用消息传递和卡尔曼模型的无线传感网定位跟踪算法

针对无线传感网系统中传感节点的定位跟踪问题，提出了一种基于消息传递和卡尔曼模型的定位跟踪算法。将目标物与多个锚节点之间的测距整理为线性混合模型，通过贝叶斯公式进行因式分解，建立因子图。假定目标物位置随时间连续移动，相邻的定位坐标之间存在潜在的缓变关系，并利用卡尔曼模型挖掘该关系，即将测距、定位和卡尔曼模型整体进行优化。在因子图中选择合适的消息更新规则，对每条边上的消息进行更新计算，形成定位跟踪算法。最后，建立虚拟和实测环境对所提算法进行数值验证。由于将测距、跟踪和卡尔曼模型融合为一个整体进行优化，相比文献中已有方法，所提定位跟踪算法在仿真环境中表现出超过2 dB的性能优势，在实测场景下也具有接近1 dB的性能增益，证明了其有效性。

A Positioning and Tracking Algorithm Based on Message Passing and Kalman Model in Wireless Sensor Network

To solve the problem of target positioning and tracking problem in wireless sensor network(WSN) system,a positioning and tracking algorithm based on message passing and Kalman model is proposed.The distance measurement between the target and multiple anchor nodes is sorted into a linear mixing model,and the factor graph is established by factoring the Bayes formula.It is assumed that the position of the object moves continuously with time,and there is a potential slow change relationship between the adjacent positioning coordinates,the Kalman model is used to explore the relationship,that is,the Kalman model is constructed into a whole for optimization.The appropriate message update rule is selected in the factor graph,the edge of each message is updated and calculated to form a positioning and tracking algorithm.Finally,a virtual and measured environment is established to verify the proposed algorithm numerically.Due to the integration of ranging,tracking and Kalman models into a whole for optimization,compared with the existing methods in the literature,the proposed algorithm proposed shows a performance advantage of more than 2 dB in the simulation environment,and also has a performance gain of close to 1 dB in the actual scene,which verifies its effectiveness.