DocumentCode :
3695824
Title :
Statistical performance analyzer for compressive sensing gradient algorithm
Author :
Stefan Vujović;Irena Orović;Miloš Daković;Ljubiša Stanković
Author_Institution :
Faculty of Electrical Engineering, University of Montenegro, Podgorica, Montenegro
fYear :
2015
Firstpage :
101
Lastpage :
104
Abstract :
The paper presents a statistical performance analyzer for compressive sensing gradient algorithm. It is composed of two main parts. The first part is designed as a tool for signal reconstruction based on the gradient algorithm. The possibility to generate various signals is incorporated through the special panel which is used to set the length, sparsity, percent of available samples and the ranges of amplitude. Also, a noise with specified variance can be added through this panel as well. The second part of presented instrument is dedicated to the statistical analysis. Different analysis can be performed through this part such as the calculation of MSE as a function of noise variance or signal sparsity, as well as the analysis of MSE and computational time in terms of sparsity and the amount of missing samples. This instrument can be useful for both the research and educational purpose for the area of compressive sensing and signal reconstruction.
Keywords :
"Instruments","Algorithm design and analysis","Discrete Fourier transforms","Compressed sensing","Biomedical measurement","Matching pursuit algorithms","Signal processing algorithms"
Publisher :
ieee
Conference_Titel :
ELMAR (ELMAR), 2015 57th International Symposium
Print_ISBN :
978-953-184-209-9
Type :
conf
DOI :
10.1109/ELMAR.2015.7334506
Filename :
7334506
Link To Document :
بازگشت