异步协作定位的误差下限推导与最优化资源分配

发布时间：2018-07-03 05:12

本文选题：异步协作定位 + 性能下限　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：无线技术快速发展的今天,位置信息越来越受重视。在实际应用中,考虑到成本的需求和无线网络节点分布的不确定性,通过增加协作机制能够有效提升网络的定位性能。由于节点间晶振很难做到精确相同,节点间的时钟同步难以实现,因此异步定位能更好地贴近现实情况。在资源受限(功率与带宽)的无线定位网络中合理分配节点间的系统资源使定位精度达到最优是本文需要研究的重要课题。本文以无线异步协作定位网络误差下限推导为目标讨论了资源最优化分配的几个问题:给出衡量定位算法的标准,本文建立了无线异步协作定位的系统模型,在此基础上给出了联合位置估计与时钟偏差估计的克拉美-罗下界,并通过该下界的舒尔补分别给出关于位置估计、时钟偏差估计的克拉美-罗下界。根据得到的联合位置估计与时钟偏差估计的克拉美-罗下界,提取下界矩阵特征值参数,本文给出了三维空间中的几何模型,并解释了它的物理意义。通过降维讨论法,本文又给出了基本定位算法的几何模型及物理意义的解释。本文随之对资源受限的无线异步协作定位网络进行研究,以联合位置估计与时钟偏差估计的克拉美-罗下界为目标函数,通过泰勒展开与迭代法使优化问题变为凸问题,建立框架模型给出了网络中功率与带宽一定的情况下两者合理分配与均匀分配所带来的定位精度的差异对比。本文同时对无线异步非协作定位与无线异步协作定位在最优化资源分配的前提下进行比较,证明协作在提高定位精度时发挥的作用。本文通过协作定位类比给出协作同步的概念,同时与协作定位进行比较,给出了定位与同步之间相互促进的关系,同时对它的几何特征进行分析,得出可以把同步看作是一维的定位问题,定位为同步的理解提供了基础。本文首次给出了定位与同步具体的关系表达式,为进一步探索同步与定位的关系提供了理论支撑。
[Abstract]:With the rapid development of wireless technology, more and more attention is paid to location information. In practical application, considering the demand of cost and the uncertainty of node distribution in wireless network, the location performance of the network can be effectively improved by adding cooperative mechanism. Because the crystal oscillator between nodes is difficult to achieve the exact same and the clock synchronization between nodes is difficult to achieve, asynchronous positioning can be more close to the reality. It is an important task for this paper to allocate the system resources between nodes to optimize the location accuracy in wireless location networks with limited resources (power and bandwidth). In this paper, several problems of optimal resource allocation are discussed in order to deduce the lower limit of error of wireless asynchronous cooperative location network. The criteria for measuring location algorithm are given, and the system model of wireless asynchronous cooperative location is established in this paper. On this basis, the Clame-Luo lower bound of the joint position estimation and the clock deviation estimate is given, and by the Schulcomplement of the lower bound, the Crame-Luo lower bound of the joint position estimation and the clock deviation estimate is given respectively. Based on the lower bound of the joint position estimation and clock deviation estimation, the eigenvalue parameters of the lower bound matrix are extracted. In this paper, the geometric model in three dimensional space is given and its physical meaning is explained. The geometric model of the basic localization algorithm and the explanation of its physical meaning are given by the method of dimension reduction. In this paper, the resource constrained wireless asynchronous cooperative localization network is studied. With the lower bound of joint position estimation and clock deviation estimation as the objective function, the optimization problem becomes convex by Taylor expansion and iteration. The frame model is established and the difference of location accuracy between the two reasonable allocation and uniform allocation is given under the condition of certain power and bandwidth in the network. At the same time, this paper compares wireless asynchronous non-cooperative location with wireless asynchronous cooperative location on the premise of optimizing resource allocation, and proves the role of cooperation in improving positioning accuracy. In this paper, the concept of cooperative synchronization is given through the analogy of cooperative location, and the relationship between location and synchronization is given, and its geometric characteristics are analyzed. It is concluded that synchronization can be regarded as a one-dimensional localization problem, which provides the basis for understanding synchronization. In this paper, the expression of the relationship between location and synchronization is presented for the first time, which provides theoretical support for further exploring the relationship between synchronization and positioning.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92

【参考文献】