With the popularization of computer and network in recent years, the information degree of daily life is increasing, the demand for information transmission and processing is increasing too. Data flow in the network transmission is very large, it is necessary to propose effective data stream processing method. Compressed sensing can reconstruct the entire signal with cost a small amount of observed data, this significant savings hardware resources and the cost of processing data. Compressed sensing ideas brought great improvements in data stream processing problems. In this paper, we use the latest ideas of compressed sensing to solve the optimization problem of the reconstruction of data streams, and provide adaptive weighted regularization method. The simulation examples show that the proposed method can reconstruct data stream well, and have some superiority on the reconstruction compare with other reconstruction algorithms.
[1]
Wang Liang-jun,et al.
Advances in Theory and Application of Compressed Sensing
,
2009
.
[2]
E.J. Candes,et al.
An Introduction To Compressive Sampling
,
2008,
IEEE Signal Processing Magazine.
[3]
Zongben Xu,et al.
$L_{1/2}$ Regularization: A Thresholding Representation Theory and a Fast Solver
,
2012,
IEEE Transactions on Neural Networks and Learning Systems.
[4]
Hou Biao,et al.
Development and Prospect of Compressive Sensing
,
2011
.
[5]
Xing Fu,et al.
Investigation on Solutions of Cubic Equations with One Unknown
,
2003
.