Identification of Communication Cables Based on Scattering Parameters and a Support Vector Machine Algorithm

The identification of cables is becoming increasingly important due to the higher complexity of technical equipment and installations. In this letter, we propose a method for identifying coaxial communication cables based on the scattering parameters at 101 frequencies for 16 coaxial communication c...

Full description

Saved in:
Bibliographic Details
Published in:IEEE sensors letters 2021-07, Vol.5 (7), p.1-4
Main Authors: Bader, Oumaima, Haddad, Dhia, Kallel, Ahmed Yahia, Hassine, Tarek, Amara, Najoua Essoukri Ben, Kanoun, Olfa
Format: Article
Language:English
Subjects:
Algorithms
artificial intelligence
cable identification
cable impedance
Cables
Classification
Classification algorithms
Coaxial cables
Communication
Communication cables
Connectors
Embedded systems
Parameter identification
Principal components analysis
Reflection
Scattering
Scattering parameters
Sensor applications
supervised learning
support vector machine (SVM)
Support vector machines
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The identification of cables is becoming increasingly important due to the higher complexity of technical equipment and installations. In this letter, we propose a method for identifying coaxial communication cables based on the scattering parameters at 101 frequencies for 16 coaxial communication cables with a characteristic impedance of 50 \Omega and various lengths, dimensions, and connector types. For classification, we consider the ready cable, with its specific length and type as a whole, as a unique class. The support vector machine (SVM) algorithm is used because of its ability to process high-dimensional spaces with linearly inseparable data. The achieved classification results when applying the principal component analysis on a database, including the input port reflection's magnitude values, show a testing accuracy of 100% with only five principal components. The method is suitable for realization as an embedded system.
ISSN:2475-1472
2475-1472
DOI:10.1109/LSENS.2021.3087539