A method with error estimates for band-limited signal extrapolation from inaccurate data
A method with error estimates for band-limited signal extrapolation from inaccurate data
复制标题
DOI：
10.1088/0266-5611/13/6/015
复制
发表时间：
1997-12
期刊：
影响因子：
2.1
通讯作者：
X. Xia;M. Nashed
中科院分区：
文献类型：
--
作者：
X. Xia;M. Nashed
在这篇文章中，我们考虑了从固定区间外的带限信号的(近似值或污染值)外推该区间的问题。我们提出了一种新的外推方法来估计外推值与真实值之间的误差，并解决了问题的不适定性问题。这种方法被称为改进的最小范数解(MMNS)方法。研究了连续MMN及其离散化问题。当数据误差的最大幅度已知时，误差估计对于某些类别的带限信号是成立的。这些类型的带限信号也被刻画出来。
In this paper, we consider the problem of extrapolation of a band-limited signal outside a fixed interval from its (approximate or contaminated) values in that interval. We propose a new extrapolation method that estimates the error between the extrapolated and true values, and which also resolves the ill-posedness of the problem. The method is called a modified minimum norm solution (MMNS) method. Both the continuous MMNS and its discretization are studied. The error estimates hold for some classes of band-limited signals, when the maximum magnitude of the data error is known. These classes of band-limited signals are also characterized.