Local Interpolated Compressive Sampling for Internet Traffic Reconstruction
Irawati I.D.a, Suksmono A.B.a, Edward I.Y.M.a
a School of Electrical and Informatics, Institut Teknologi Bandung, Bandung, Indonesia
Abstract
© 2017 IEEE.In this paper, we propose the integration of local interpolation on compressive sampling (CS) for application on internet traffic reconstruction. CS managed to solve the problem of missing Internet traffic but not in the case of extreme missing like Missing Elements at Random (MER) and Combine Missing Patterns (CMP). To overcome the problem is done by merging between local interpolations with CS. In this study conducted a comparison of local interpolation using correlation method and Euclidean norm and tested the effect of similarity parameters between rows and similarity between columns. The simulation results show that the incorporation of local interpolation can improve accuracy with insignificant time increments, mainly on the addition of Euclidean norm technique.
Author keywords
Compressive sampling,Euclidean norm,Internet traffic,Local interpolation,Similarity parameter,Time increments
Indexed keywords
compressive sampling,correlation,Euclidean norm,internet traffic matrix,local interpolation
Funding details
ACKNOWLEDGMENT The authors would like to thank Telkom Foundation, Indonesia Ministry of Higher Education, and LPPM ITB for financial support in this research.