非高斯环境噪声下的高效自适应无线传感器网络定位算法研究

项目名称： 非高斯环境噪声下的高效自适应无线传感器网络定位算法研究

项目编号： No.61261034

项目类型： 地区科学基金项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 闫坤

作者单位： 桂林电子科技大学

项目金额： 40万元

中文摘要： 真实环境中出现的信号并不总是满足高斯统计特性。但由于技术局限和系统计算量的要求，对非高斯统计特性的研究应用常常被有意的忽视。因此，在数据量少，精度要求高的估计问题中，现有系统的高斯假设与信号统计模型间的误差成为无法克服的困难。本项目将针对这一问题探索能够高效准确分析非高斯噪声环境并进行处理的方法，并在这种思想指导下设计在复杂环境中稳定准确工作的定位算法。 本课题旨在通过对统计特征的分析自适应调节估计算法各项参数，从而使优化算法在不同噪声环境中稳定运行。这一研究成果将有效提高检测估计算法的鲁棒性，有深刻的理论指导意义。同时，本课题将综合考虑计算复杂度和算法精度，以保证算法精度、降低计算复杂度为指导目标，设计平衡系统参数的手段，对实际系统设计也有广泛的应用价值。 基于这种方法论，通过检测信号统计特性，本项目将设计非高斯噪声环境中的高效无线传感器定位算法。

中文关键词： 非高斯噪声；定位算法；鲁棒性分析；高斯度测量；压缩感知

英文摘要： In detection and estimaion, Gaussian model might not always fit the real data. However, gaussian model is always employed in practice for its simplicity. Consequently, it is quite difficult to get a better accuracy when the data is sparse. Non-Gaussian noise become the bottle-neck for detection and estimation. This work is proposed to tackle this difficulty. Then the corresponding localization algorithm will be designed thereupon. To make the optimized localization algorithm effective and efficient in different circumstances, this proposal tried some novel adaptive schemes to adjust the parameters in algorithm. By promoting the robustness of optimization algorithm, this research will be quite useful in different algorithm design problems.Moreover, considering the tradeoff between the accuracy and the computational burden, the novel estimation scheme will be designed with an optimized tradeoff. Consequently, this research project provides a promising approach for solving practical problems. The localization in wireless sensor network will be effectively solved by our proposed scheme. Non-Gaussian noise can be efficiently dealt with. The parameters in localization algorithm are adjusted according to the statistical features of the data.

英文关键词： Non-Gaussian Noise；Localization Algorithm；Robustness Analysis；Gaussianity Test；Compressive Sensing