基于因子图和联合消息传递的无线网络协作定位算法

发布时间：2018-01-19 19:23

本文关键词： 近似贝叶斯推理因子图置信传播平均场方法无线传感器网络协作定位　出处：《计算机应用》2017年05期 　论文类型：期刊论文

【摘要】：针对现有基于消息传递算法的无线网络节点定位算法复杂度和通信开销过高的问题,提出一种基于测距的、低复杂度低协作开销的联合消息传递节点定位算法。所提算法考虑参考节点位置的不确定性以减少误差累积,并将消息约束为高斯函数以降低通信开销。首先,根据系统的概率模型和因子分解设计因子图;然后,根据状态转移模型和测距模型的特点,分别使用置信传播和平均场方法计算预测消息和协作消息;最后,在每次迭代过程中,通过非线性项的泰勒展开将非高斯置信消息近似为高斯函数。仿真分析表明,所提算法的定位性能与基于粒子的SPAWN算法接近,但节点间传输的信息由大量粒子变为均值向量和协方差矩阵,同时计算复杂度也大幅降低。
[Abstract]:In order to solve the problem of high complexity and communication overhead of the existing wireless network node location algorithm based on message passing algorithm, a new method based on ranging is proposed. A joint messaging node location algorithm with low complexity and low collaboration overhead. The proposed algorithm considers the uncertainty of the reference node position to reduce the error accumulation and restricts the message to Gao Si function to reduce the communication overhead. According to the probability model of the system and factor decomposition design factor graph; Then, according to the characteristics of the state transition model and the ranging model, the predictive message and the cooperative message are calculated by using the confidence propagation method and the mean field method, respectively. Finally, in each iteration process, the non-#china_person0# confidence message is approximated to Gao Si function through the Taylor expansion of the nonlinear term. The simulation results show that. The localization performance of the proposed algorithm is similar to that of the Particle based SPAWN algorithm, but the information transmitted between nodes is changed from a large number of particles to mean vector and covariance matrix, and the computational complexity is also greatly reduced.
【作者单位】：国家数字交换系统工程技术研究中心;洛阳师范学院物理与电子信息学院;郑州大学信息工程学院;
【基金】：国家自然科学基金资助项目(61571402,61401401)~~
【分类号】：TN929.5;TP212.9
【正文快照】： 0引言在基于无线传感器网络(Wireless Sensor Network,WSN)的应用中,传感器节点检测到的信息若没有准确的位置信息将变得毫无价值[1]。但考虑到成本和能量限制,一般只有少数参考节点的位置是已知的,其他大部分节点(称为待定位节点)通过邻近的参考节点的位置和与其之间的距离等

【相似文献】