在非视距环境下利用无线网络对目标进行定位具有重要的理论和实际意义,其中完全未知或者已知部分非视距误差的统计信息以及对非视距路径判决误差稳健的定位算法是未来研究的两个热点问题。为提高非视距误差统计信息匮乏时的定位性能，我们拟建立关于目标位置以及非视距误差的联合优化问题，并采用凸优化松弛方法实现抑制非视距误差并提高定位精度。对于非视距路径判决时出现的漏检和虚警的情况，我们拟采用凸优化方法中的稳健性处理方法降低判决误差对定位性能的影响，减小对判决机制可靠性的依赖。最后，我们将研究两个问题中定位算法的性能下界并通过仿真验证和分析算法的性能。本项目对上述两个问题的研究是开创性、探索性的。由于该研究的前提条件更符合实际情况，因此它对算法从理论应用于实际将起到良好的推动作用，具有重要的理论和实际意义。

Localization by a wireless network in Non-Line-of-Sight (NLOS) environments is of great importance since it finds wide applications in practice. The future research mainly focuses on the cases that the statistical information of NLOS errors is completely unknown or partly known, and the detection to the NLOS paths is subject to errors. To improve the localization performance without knowing much statistical information of NLOS errors, we intend to establish an optimization problem with respect to the source location and the NLOS errors, and apply the popular convex relaxation technique to jointly estimate them. When there exist detection errors to NLOS paths, we intend to employ the robust processing methods to improve the robustness of the localization methods to the detection errors. This work is a pioneering and innovative work, which is not found in current research. Since this work releases some impractical assumptions, the research will facilitate the localization methods to their applications in practice, and hence, this work is of great importance in both theory and practice.