当前位置:主页 > 科技论文 > 信息工程论文 >

无线网络中基于变分消息传递的分布式协作定位算法

发布时间:2018-04-25 09:32

  本文选题:无线网络 + 协作定位 ; 参考:《信号处理》2017年05期


【摘要】:针对基于消息传递算法的节点定位方法复杂度和通信开销较高的问题,提出一种适用于节点可移动网络的低复杂度低协作开销的节点自定位算法。为降低通信负载,该算法将消息约束为高斯型函数,网络中只需传输各消息的均值和方差,并采用适用于指数模型的变分消息传递(VMP)算法以降低计算复杂度。首先,根据节点的历史轨迹对节点位置进行预测,得到当前时刻的先验信息。然后,在因子图上按照VMP消息更新规则、通过迭代近似求解节点位置变量的后验分布。在消息更新中,对于非线性测距模型引起的非高斯置信,通过非线性项的二阶泰勒级数展开将其近似为高斯型函数。最后,根据最大后验估计准则得到位置估计。仿真结果表明,该算法的定位精度与基于非参数化置信传播的SPAWN(Sum-Product Algorithm over a Wireless Network)接近,但计算复杂度和通信负载均显著降低。
[Abstract]:In order to solve the problem of high complexity and communication overhead of node localization based on message passing algorithm, a low complexity and low cooperative node localization algorithm for node mobile networks is proposed. In order to reduce the communication load, the algorithm restricts the message to Gao Si type function, only needs to transmit the mean value and variance of each message in the network, and adopts the variational message passing (VMP) algorithm suitable for exponential model to reduce the computational complexity. First, the node position is predicted according to the historical track of the node, and the prior information of the current moment is obtained. Then, according to the VMP message update rule, the posteriori distribution of node location variables is approximately solved by iterating on the factor graph. In message updating, for the non- confidence caused by the nonlinear ranging model, the nonlinear term is approximated to a Gao Si type function by the second-order Taylor series expansion of the nonlinear term. Finally, the position estimation is obtained according to the maximum posteriori estimation criterion. Simulation results show that the location accuracy of the algorithm is close to that of SPAWN(Sum-Product Algorithm over a Wireless Network based on nonparametric confidence propagation, but the computational complexity and communication load are significantly 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) 河南省科技攻关项目(152102210333)
【分类号】:TN929.5;TP212.9

【相似文献】

相关博士学位论文 前1条

1 罗咏R,

本文编号:1800771


资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/1800771.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户f76b7***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com