三边测量法的结果稳定性研究  

Research on the Stability of Trilateration Algorithms

作　　者：王小平[1] 罗军[1] 沈昌祥[2] 

机构地区：[1]国防科学技术大学计算机学院,湖南长沙410073 [2]海军计算技术研究所,北京100841

出　　处：《计算机工程与科学》2012年第6期12-17,共6页Computer Engineering & Science

基　　金：国家自然科学基金资助项目(61170261;60903224)

摘　　要：定位技术作为网络应用和服务的支撑技术,是无线传感器网络研究的热点问题之一。其中,三边测量法是定位技术的基础,是构建复杂定位系统的基本模块。本文针对三边测量法最优化过程对迭代初值敏感的问题,对结果的稳定性展开研究,提出一种新的三边测量法的迭代最优化的初值选择算法GICC。对比多种三边测量法的实现技术,GICC算法能够自适应地寻找全局最优解,使得三边测量法通过简单下降法即可得到全局最优解,从而避免陷入局部极小解。通过仿真实验证明该算法能有效保证三边测量法结果的稳定性,即保证三边测量法的结果误差一直保持在较低的水平上,从而显著减小全网定位的误差。Localization is one of the major supportive techniques for the application and services, making it a research focus in wireless sensor networks. Trilateration is the basis of localization tech- niques, which is a primary building block of many complicated localization systems. This paper studies the stability of the optimization of trilateration, which is very sensitive to the initial value. We propose a new algorithm, named GICC, to conquer such difficulties. Compared with the existing implementation schemes of trilateration, GICC allows the optimization process of trilateration to obtain the global mini- mum by a simple gradient-based descent method. We conduct extensive simulations to evaluate the effec- tiveness of the proposed method. The results show that GICC can effectively avoid local minima, and di- minish the errors of the localization result significantly.

关 键 词：定位 三边测量法 无线网络 最优化 最小二乘 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]