Network localization is capable of providing accurate and ubiquitous position information for numerous wireless applications. This paper studies the accuracy of cooperative network localization in large-scale wireless networks. Based on a decomposition of the equivalent Fisher information matrix (EFIM), we develop a random-walk-inspired approach for the analysis of EFIM, and propose a position information routing interpretation of cooperative network localization. Using this approach, we show that in large lattice and stochastic geometric networks, when anchors are uniformly distributed, the average localization error of agents grows logarithmically with the reciprocal of anchor density in an asymptotic regime. The results are further illustrated using numerical examples.
翻译:网络本地化能够为众多无线应用程序提供准确和无处不在的定位信息。本文研究合作网络在大型无线网络中本地化的准确性。根据同等的渔业信息矩阵的分解,我们开发了一种随机行走驱动法来分析渔业信息矩阵,并提出了合作网络本地化的定位信息路由解释。我们采用这种方法表明,在大型的固定和随机几何网络中,当锚分布一致时,代理物的平均本地化错误会随着固定密度的对等在无防系统中的对等而增加。我们用数字实例进一步说明了结果。