Software Implementation of Real-time Discrete Wavelet Transform Algorithm with Filter Banks for Network Traffic Parameters Estimation
The 5th International Conference on Information Systems and Technologies (ICIST 2015): Proceedings 2015
Nikolajs Bogdanovs, Elans Grabs, Ernests Pētersons

This article describes real-time discrete wavelet transform algorithm implementation for high-level programming language. The article describes multiscale transform algorithms both for direct discrete wavelet transform and inverse discrete wavelet transform. This algorithm has been implemented in C++ programming language and tested with Raspberry Pi microprocessor system. This article proposes the improved delay line algorithm without full shifting of register. New algorithm requires single reading operation, single writing operation and one division calculation for any length of delay line. The article includes experimental measurements of processing time on Raspberry Pi for various scale numbers. The algorithm described in this article can be used with any software tool capable of using high level programming language, for example Matlab, Octave, Opnet, as well as microprocessors in communications devices, such as routers, switches and access points. This is the main purpose – to create algorithm which is not tied strictly to hardware implementation but also, nonetheless, provides real-time discrete wavelet analysis capability.


Keywords
Discrete Wavelet Transform, Filter Banks, Multiscale analysis, Raspberry Pi microprocessor system

Bogdanovs, N., Grabs, E., Pētersons, E. Software Implementation of Real-time Discrete Wavelet Transform Algorithm with Filter Banks for Network Traffic Parameters Estimation. In: The 5th International Conference on Information Systems and Technologies (ICIST 2015): Proceedings, Turkey, Istanbul, 21-23 March, 2015. Istanbul: 2015, pp.1-7.

Publication language
English (en)
The Scientific Library of the Riga Technical University.
E-mail: uzzinas@rtu.lv; Phone: +371 28399196