非视距环境下无线定位系统的位置欺骗问题研究

There has been a growing concern about location spoofing on today's wireless localization systems. To deal with the problem, existing studies typically make use of a consistency analysis which explores the inconsistency between localization results and localization signals caused by location spoofing. A drawback is the requirement of good channel conditions. In case of non-line-of sight (NLOS) channel conditions, these studies will face challenges from two aspects: (1) incapable of differentiating NLOS errors from biases caused by location spoofing, and (2) misidentifying some location biases when used outside of the convex hull area. In this project, we will address the problem of location spoofing in NLOS conditions, with a focus on the above two challenges. Our project will consist of three parts. First, we will investigate the misidentification problem including how to bound the area of misidentification and how to develop a new method to ensure accurate identification in the area. Then, we will dive into the physical layer and network layer of a localization system, to design a signal based identification method and a distributed voting based identification method for location spoofing. Finally, a time series analysis will be carried out to identify anomaly trajectories caused by location spoofing. The outcome of this project is expected to guide the policy and practice standard development regarding wireless positioning in our country.

随着无线定技术的广泛应用，位置欺骗问题开始引起广泛关注。现有解决方案普遍采用一类“一致性”分析，通过发掘定位信号和定位结果间的不一致实现对位置欺骗的鉴别。然而，这些方案在复杂无线环境，尤其是非视距环境下都存在严重缺陷，主要包括两个方面(1)无法区分非视距传播误差和位置欺骗造成的偏差，以及(2)在凸包外围存在误差漏判问题。本项目拟从理论出发，结合实验验证，系统性的研究非视距环境下的误差发现和位置欺骗鉴别问题。研究包含三个层面。首先，研究一致性分析在凸包外围的漏判问题，确定失效区域，设计新的误差侦测方案。然后，从定位系统的物理层和网络层入手，发掘空间相似性，设计基于信号物理特征的位置鉴别方法和基于投票的位置鉴别方法。最后，利用运动用户轨迹的连续性，设计高效的时间序列算法，发现位置欺骗造成的轨迹冲突。本项目将为无线定服务的安全保障提供理论依据和数据支持。

随着无线定位技术在智能手机和物联网领域的广泛应用，无线定位系统的安全和可靠性问题开始受到广泛关注。其中，以位置欺骗这类安全威胁最为严重。早在2003年，美国交通部的报告就曾指出恶意攻击者可以通过伪造信号欺骗GPS定位用户，导致定位结果异常，甚至造成一定区域内GPS功能彻底瘫痪。近年来，越来越多的研究开始关注这一问题，主要研究方向包括如何发现恶意用户造成的定位结果异常以及如何消除这类异常等。造成定位结果异常的另一原因是无线信号的非视距传播。现有无线定位系统在非视距环境下定位误差可达589米，而一般的误差仅为数十米。如何侦测非视距传播及消除它的影响一直以来都无线定位系统研究的热点之一。本项目通过对位置欺骗信号研究和非视距信号研究，设计了一套异常误差检测机制，解决了非视距环境下定位欺骗鉴别问题。研究成果为无线定位异常误差检测和鉴别提供了理论依据和技术支持，并为相关部门制定行业标准和法规政策提供了依据。