Supervised Learning of Neural Networks for Active Queue Management in the Internet
The paper examines the AQM mechanism based on neural networks. The active queue management allows packets to be dropped from the router’s queue before the buffer is full. The aim of the work is to use machine learning to create a model that copies the behavior of the AQM <inline-formula><ma...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-07-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/15/4979 |
| _version_ | 1827686403407872000 |
|---|---|
| author | Jakub Szyguła Adam Domański Joanna Domańska Dariusz Marek Katarzyna Filus Szymon Mendla |
| author_facet | Jakub Szyguła Adam Domański Joanna Domańska Dariusz Marek Katarzyna Filus Szymon Mendla |
| author_sort | Jakub Szyguła |
| collection | DOAJ |
| description | The paper examines the AQM mechanism based on neural networks. The active queue management allows packets to be dropped from the router’s queue before the buffer is full. The aim of the work is to use machine learning to create a model that copies the behavior of the AQM <inline-formula><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>P</mi><msup><mi>I</mi><mi>α</mi></msup></mrow></semantics></math></inline-formula> mechanism. We create training samples taking into account the self-similarity of network traffic. The model uses fractional Gaussian noise as a source. The quantitative analysis is based on simulation. During the tests, we analyzed the length of the queue, the number of rejected packets and waiting times in the queues. The proposed mechanism shows the usefulness of the Active Queue Management mechanism based on Neural Networks. |
| first_indexed | 2024-03-10T09:09:42Z |
| format | Article |
| id | doaj.art-f9764d147a2b4d088739d9b6a8425e7d |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T09:09:42Z |
| publishDate | 2021-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-f9764d147a2b4d088739d9b6a8425e7d2023-11-22T06:08:38ZengMDPI AGSensors1424-82202021-07-012115497910.3390/s21154979Supervised Learning of Neural Networks for Active Queue Management in the InternetJakub Szyguła0Adam Domański1Joanna Domańska2Dariusz Marek3Katarzyna Filus4Szymon Mendla5Faculty of Automatic Control, Electronics and Computer Science, Department of Distributed Systems and Informatic Devices, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandFaculty of Automatic Control, Electronics and Computer Science, Department of Distributed Systems and Informatic Devices, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandInstitute of Theoretical and Applied Informatics Polish Academy of Sciences, Bałtycka 5, 44-100 Gliwice, PolandFaculty of Automatic Control, Electronics and Computer Science, Department of Distributed Systems and Informatic Devices, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandInstitute of Theoretical and Applied Informatics Polish Academy of Sciences, Bałtycka 5, 44-100 Gliwice, PolandFaculty of Automatic Control, Electronics and Computer Science, Department of Distributed Systems and Informatic Devices, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandThe paper examines the AQM mechanism based on neural networks. The active queue management allows packets to be dropped from the router’s queue before the buffer is full. The aim of the work is to use machine learning to create a model that copies the behavior of the AQM <inline-formula><math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>P</mi><msup><mi>I</mi><mi>α</mi></msup></mrow></semantics></math></inline-formula> mechanism. We create training samples taking into account the self-similarity of network traffic. The model uses fractional Gaussian noise as a source. The quantitative analysis is based on simulation. During the tests, we analyzed the length of the queue, the number of rejected packets and waiting times in the queues. The proposed mechanism shows the usefulness of the Active Queue Management mechanism based on Neural Networks.https://www.mdpi.com/1424-8220/21/15/4979neural networksHurst exponentself-similarityinternet trafficcongestion controldropping packets |
| spellingShingle | Jakub Szyguła Adam Domański Joanna Domańska Dariusz Marek Katarzyna Filus Szymon Mendla Supervised Learning of Neural Networks for Active Queue Management in the Internet Sensors neural networks Hurst exponent self-similarity internet traffic congestion control dropping packets |
| title | Supervised Learning of Neural Networks for Active Queue Management in the Internet |
| title_full | Supervised Learning of Neural Networks for Active Queue Management in the Internet |
| title_fullStr | Supervised Learning of Neural Networks for Active Queue Management in the Internet |
| title_full_unstemmed | Supervised Learning of Neural Networks for Active Queue Management in the Internet |
| title_short | Supervised Learning of Neural Networks for Active Queue Management in the Internet |
| title_sort | supervised learning of neural networks for active queue management in the internet |
| topic | neural networks Hurst exponent self-similarity internet traffic congestion control dropping packets |
| url | https://www.mdpi.com/1424-8220/21/15/4979 |
| work_keys_str_mv | AT jakubszyguła supervisedlearningofneuralnetworksforactivequeuemanagementintheinternet AT adamdomanski supervisedlearningofneuralnetworksforactivequeuemanagementintheinternet AT joannadomanska supervisedlearningofneuralnetworksforactivequeuemanagementintheinternet AT dariuszmarek supervisedlearningofneuralnetworksforactivequeuemanagementintheinternet AT katarzynafilus supervisedlearningofneuralnetworksforactivequeuemanagementintheinternet AT szymonmendla supervisedlearningofneuralnetworksforactivequeuemanagementintheinternet |