项目名称: 多项式相位信号参数估计的迭代算法研究
项目编号: No.61302138
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 无线电电子学、电信技术
项目作者: 边家文
作者单位: 中国地质大学(武汉)
项目金额: 25万元
中文摘要: 噪声中多项式相位信号的参数估计是非平稳信号处理中的一个基本问题,也是统计信号处理领域中的一个重要的研究内容,它在信号处理的多个领域有着广泛的应用。本项目研究多项式相位信号参数估计问题,针对乘性和加性平稳以及非平稳噪声情形,振幅均值为常数以及时变情形下的信号模型,在深入分析信号模型统计特征基础上分别构造基于观测信号以及基于待估计参数的优化目标函数两类基于统计量的迭代算法,同时研究相应复杂噪声背景下具有理论上的最优估计性能的LSE、MLE以及M-估计,并从算法的计算量、估计的无偏性以及信噪比门限等方面评价算法的估计性能。本项目的完成不仅可以弥补传统算法估计精度不足以及由此引起的误差传播较大的问题,为多项式相位信号的参数估计提供与LSE或MLE等效且计算简便、稳定、适应性强的算法,并应用于多项式相位信号参数的实时在线估计,而且对于改变目前对于迭代算法理论的认识,发展新的迭代算法理论具有重要意义。
中文关键词: 迭代算法;多项式相位信号;最小二乘估计;CRLB;统计量
英文摘要: The parameter estimation of polynomial phase signal in noise is a fundamental problem in non-stationary signal processing and also an important part in statistical signal processing. It has many applications in several fields of signal processing. This project is focused on the parameter estimation of polynomial phase signal model in condition of multiplicative and additive stationary noise as well as non-stationary noise respectively, and also in condition of constant and time varying mean of amplitude respectively. Two statistics based iterative algorithms which are based on the observation of signals and the objective function regarding the parameters respectively, will be put forward by deeply analyzing the statistical character of the above model. The research of LSE, MLE and M-estimator for the corresponding model in condition of complex noise will also be conducted at the same time. The performance of the iterative algorithms will be evaluated by the computation load, the unbiasedness and the SNR threshold etc. The completion of this project can not only make up the insufficiency in estimation accuracy and reduce the error propagation, but also provide computationally efficient, stable, robust and LSE or MLE equivalent algorithm for the parameter estimation of polynomial phase signal. Moreover, the iterat
英文关键词: Iterative algorithm;Polynomial phase signal;Least Squares Estimator;CRLB;Statistics