Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning
Radio access network optimization is a critical task in current cellular systems. For this purpose, Minimization of Drive Test (MDT) functionality provides mobile operators with georeferenced network performance statistics to tune radio propagation models in re-planning tools. However, some samples...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Open Journal of Vehicular Technology |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10818611/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832584006801031168 |
|---|---|
| author | J. M. Sanchez-Martin C. Gijon M. Toril S. Luna-Ramirez V. Wille |
| author_facet | J. M. Sanchez-Martin C. Gijon M. Toril S. Luna-Ramirez V. Wille |
| author_sort | J. M. Sanchez-Martin |
| collection | DOAJ |
| description | Radio access network optimization is a critical task in current cellular systems. For this purpose, Minimization of Drive Test (MDT) functionality provides mobile operators with georeferenced network performance statistics to tune radio propagation models in re-planning tools. However, some samples in MDT traces contain critical location errors due to the user equipment's energy-saving, thus making MDT data filtering vital to guarantee an adequate performance of MDT-driven algorithms. Supervised Learning (SL) allows to train automatic systems for detecting abnormal MDT measurements by using a labeled dataset. Unfortunately, labeling MDT data is a labor-intensive task, that can be alleviated by using Self-Supervised Learning (SSL). This work presents a novel SSL method to detect MDT measurements with abnormal position information in road scenarios. For this purpose, a dataset is first labeled by combining unlabeled MDT traces from high-mobility users and freely available land use maps, and then an SL classifier is trained. Model assessment is carried out using MDT data collected in a live Long-Term Evolution (LTE) network. Performance analysis includes the comparison of six well-known SL algorithms and 3 different sets of input features aiming to improve model accuracy, generalizability, and explainability, respectively. Results show that considering predictors regarding positioning error increases model accuracy, whereas omitting this information allows to cover a wider range of terminals. Likewise, Shapley Additive exPlanations (SHAP) analysis proves that the use of high-level predictors significantly improves model explainability. |
| format | Article |
| id | doaj-art-f0fcce3dce6f4b32aef1cc06e3a11155 |
| institution | Kabale University |
| issn | 2644-1330 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Open Journal of Vehicular Technology |
| spelling | doaj-art-f0fcce3dce6f4b32aef1cc06e3a111552025-01-28T00:02:14ZengIEEEIEEE Open Journal of Vehicular Technology2644-13302025-01-01639641110.1109/OJVT.2024.352405710818611Anomaly Detection in High Mobility MDT Traces Through Self-Supervised LearningJ. M. Sanchez-Martin0https://orcid.org/0000-0002-4707-6438C. Gijon1https://orcid.org/0000-0001-6204-0604M. Toril2https://orcid.org/0000-0003-3859-2622S. Luna-Ramirez3https://orcid.org/0000-0003-0171-5721V. Wille4https://orcid.org/0000-0003-2150-6564Instituto de Telecomunicación (TELMA), CEI Andalucía TECH, E.T.S. Ingeniería de Telecomunicación, Universidad de Málaga, Malaga, SpainInstituto de Telecomunicación (TELMA), CEI Andalucía TECH, E.T.S. Ingeniería de Telecomunicación, Universidad de Málaga, Malaga, SpainInstituto de Telecomunicación (TELMA), CEI Andalucía TECH, E.T.S. Ingeniería de Telecomunicación, Universidad de Málaga, Malaga, SpainInstituto de Telecomunicación (TELMA), CEI Andalucía TECH, E.T.S. Ingeniería de Telecomunicación, Universidad de Málaga, Malaga, SpainNokia, Cambridge, U.K.Radio access network optimization is a critical task in current cellular systems. For this purpose, Minimization of Drive Test (MDT) functionality provides mobile operators with georeferenced network performance statistics to tune radio propagation models in re-planning tools. However, some samples in MDT traces contain critical location errors due to the user equipment's energy-saving, thus making MDT data filtering vital to guarantee an adequate performance of MDT-driven algorithms. Supervised Learning (SL) allows to train automatic systems for detecting abnormal MDT measurements by using a labeled dataset. Unfortunately, labeling MDT data is a labor-intensive task, that can be alleviated by using Self-Supervised Learning (SSL). This work presents a novel SSL method to detect MDT measurements with abnormal position information in road scenarios. For this purpose, a dataset is first labeled by combining unlabeled MDT traces from high-mobility users and freely available land use maps, and then an SL classifier is trained. Model assessment is carried out using MDT data collected in a live Long-Term Evolution (LTE) network. Performance analysis includes the comparison of six well-known SL algorithms and 3 different sets of input features aiming to improve model accuracy, generalizability, and explainability, respectively. Results show that considering predictors regarding positioning error increases model accuracy, whereas omitting this information allows to cover a wider range of terminals. Likewise, Shapley Additive exPlanations (SHAP) analysis proves that the use of high-level predictors significantly improves model explainability.https://ieeexplore.ieee.org/document/10818611/Data filteringminimization of drive testoutlier detectionself-supervised learningshapleyuser positioning |
| spellingShingle | J. M. Sanchez-Martin C. Gijon M. Toril S. Luna-Ramirez V. Wille Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning IEEE Open Journal of Vehicular Technology Data filtering minimization of drive test outlier detection self-supervised learning shapley user positioning |
| title | Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning |
| title_full | Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning |
| title_fullStr | Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning |
| title_full_unstemmed | Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning |
| title_short | Anomaly Detection in High Mobility MDT Traces Through Self-Supervised Learning |
| title_sort | anomaly detection in high mobility mdt traces through self supervised learning |
| topic | Data filtering minimization of drive test outlier detection self-supervised learning shapley user positioning |
| url | https://ieeexplore.ieee.org/document/10818611/ |
| work_keys_str_mv | AT jmsanchezmartin anomalydetectioninhighmobilitymdttracesthroughselfsupervisedlearning AT cgijon anomalydetectioninhighmobilitymdttracesthroughselfsupervisedlearning AT mtoril anomalydetectioninhighmobilitymdttracesthroughselfsupervisedlearning AT slunaramirez anomalydetectioninhighmobilitymdttracesthroughselfsupervisedlearning AT vwille anomalydetectioninhighmobilitymdttracesthroughselfsupervisedlearning |