A machine learning based approach for 5G network security monitoring
This paper constructs a 5G network security detection system based on the functional requirements of network security detection and the P2DR process model. The structure can be broken down into three layers from the bottom to the top: acquisition layer, analysis layer, and display layer. The design...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Sciendo
2024-01-01
|
| Series: | Applied Mathematics and Nonlinear Sciences |
| Subjects: | |
| Online Access: | https://doi.org/10.2478/amns-2024-0071 |
| _version_ | 1797303158207152128 |
|---|---|
| author | Chen Bin |
| author_facet | Chen Bin |
| author_sort | Chen Bin |
| collection | DOAJ |
| description | This paper constructs a 5G network security detection system based on the functional requirements of network security detection and the P2DR process model. The structure can be broken down into three layers from the bottom to the top: acquisition layer, analysis layer, and display layer. The design focuses on writing the SVM algorithm into the software of the communication network security vulnerability monitoring system, and in order to solve the defect of the long training time of the model of this machine learning algorithm, the incremental learning vector machine model is used, which is combined into the CSV-KKT-ISVM model. Test datasets that cover system performance, effectiveness, and leakage are used to test the system after it is completed. The test data was analyzed to prove that the system’s memory usage was maintained at 46M, CPU usage was 5% to 10%, and the response time was no later than 1 s. The monitoring accuracy was 98.5% at the highest but decreased with the increase of the percentage of vulnerability data, and the accuracy dropped to 93.9% at 50%, the minimum was not lower than 90%, and the error rate was no less than 0.8%. To achieve the best outcome, the system threshold should be set to 5, and there should be no false alarms or misreporting. |
| first_indexed | 2024-03-07T23:48:55Z |
| format | Article |
| id | doaj.art-2eb6a3b28257454db8e4330ed6a8b6d5 |
| institution | Directory Open Access Journal |
| issn | 2444-8656 |
| language | English |
| last_indexed | 2024-03-07T23:48:55Z |
| publishDate | 2024-01-01 |
| publisher | Sciendo |
| record_format | Article |
| series | Applied Mathematics and Nonlinear Sciences |
| spelling | doaj.art-2eb6a3b28257454db8e4330ed6a8b6d52024-02-19T09:03:34ZengSciendoApplied Mathematics and Nonlinear Sciences2444-86562024-01-019110.2478/amns-2024-0071A machine learning based approach for 5G network security monitoringChen Bin01Huaxin Consulting Co., Ltd., Hangzhou, Zhejiang, 310052, China.This paper constructs a 5G network security detection system based on the functional requirements of network security detection and the P2DR process model. The structure can be broken down into three layers from the bottom to the top: acquisition layer, analysis layer, and display layer. The design focuses on writing the SVM algorithm into the software of the communication network security vulnerability monitoring system, and in order to solve the defect of the long training time of the model of this machine learning algorithm, the incremental learning vector machine model is used, which is combined into the CSV-KKT-ISVM model. Test datasets that cover system performance, effectiveness, and leakage are used to test the system after it is completed. The test data was analyzed to prove that the system’s memory usage was maintained at 46M, CPU usage was 5% to 10%, and the response time was no later than 1 s. The monitoring accuracy was 98.5% at the highest but decreased with the increase of the percentage of vulnerability data, and the accuracy dropped to 93.9% at 50%, the minimum was not lower than 90%, and the error rate was no less than 0.8%. To achieve the best outcome, the system threshold should be set to 5, and there should be no false alarms or misreporting.https://doi.org/10.2478/amns-2024-0071machine learningincremental vector machine modelp2drsecurity monitoringcommunication network security05c82 |
| spellingShingle | Chen Bin A machine learning based approach for 5G network security monitoring Applied Mathematics and Nonlinear Sciences machine learning incremental vector machine model p2dr security monitoring communication network security 05c82 |
| title | A machine learning based approach for 5G network security monitoring |
| title_full | A machine learning based approach for 5G network security monitoring |
| title_fullStr | A machine learning based approach for 5G network security monitoring |
| title_full_unstemmed | A machine learning based approach for 5G network security monitoring |
| title_short | A machine learning based approach for 5G network security monitoring |
| title_sort | machine learning based approach for 5g network security monitoring |
| topic | machine learning incremental vector machine model p2dr security monitoring communication network security 05c82 |
| url | https://doi.org/10.2478/amns-2024-0071 |
| work_keys_str_mv | AT chenbin amachinelearningbasedapproachfor5gnetworksecuritymonitoring AT chenbin machinelearningbasedapproachfor5gnetworksecuritymonitoring |