基于非参数化动态噪声模型的自适应室内定位算法研究

项目名称： 基于非参数化动态噪声模型的自适应室内定位算法研究

项目编号： No.61501443

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 无线电电子学、电信技术

项目作者： 赵毓斌

作者单位： 中国科学院深圳先进技术研究院

项目金额： 19万元

中文摘要： 室内定位技术能够在未来为移动互联网和物联网带来巨大的商业和应用价值，然而由于缺少对室内复杂环境的了解和计算复杂度高的缘故，目前仍然没有一种普适性较高的算法能够推广室内定位的大规模应用。针对上述问题，本项目面向环境未知的非确定性室内场景，以无线定位系统为研究内容，研究能够适用于多种通信系统的通用定位算法。研究的主要目标包括：（1）针对无线信号在室内传输的复杂性，转变传统的概率拟合和信道建模的方法，提出面向非确定性场景下的信号噪声非参数化动态建模方法，研究动态模型调节机制和参数估计方法，为算法设计提供模型依据；（2）研究面向多种无线通信系统的通用自适应算法架构，结合非线性滤波和最优化方法，开发能够依据环境变化自适应调节参数的定位算法，实现无线信号的融合和定位算法性能的优化，保证准确定位；（3）理论上分析推导基于克拉美-罗下限的自适应算法在室内非确定性场景下的最优性能，为算法的改进提供参考。

中文关键词： 室内定位；非线性滤波；位置估计；克拉美-罗下界

英文摘要： Indoor localization techniques can bring potential commercial business and services for the future mobile internet and the internet of things. However, due to the lack of knowledge of the complicated indoor environment and high computational complexity, a general algorithm, which can benefit the wide application of indoor localization systems, still does not exist. To solve this problem, this project will research the design method of the general localization algorithm for multiple wireless communication systems in the unknown and uncertain indoor environment. In summary, the main works of this project can be categorized as follows:.(1).Due to the complicated propagation feature in the indoor uncertain environment, a non-parametric dynamic modeling method for the indoor measurement noise is proposed, which is different from the classical distributed fitting and channel modeling method..(2).The architecture of adaptive localization algorithms for multiple wireless system is proposed. By applying the nonlinear filters and optimization method, the localization algorithm which applies nonlinear filters for the fusion of wireless signals and can self-adapt according to the change of the indoor environment is developed. To guarantee the estimation accuracy, the performance is improved by using the convex optimization method..(3).Deriving the Cramér-Rao lower bound in the uncertain indoor environment to indicate the optimal performance for the proposed adaptive algorithm. The analysis can indicate the improvement of the proposed algorithm.

英文关键词： Indoor Localization;Nonlinear Filter;Position Estimation;Cramér-Rao Lower Bound