Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Network slicing architecture.

More »

Fig 1 Expand

Table 1.

Comparison of 5G and other networks.

More »

Table 1 Expand

Table 2.

Overview of the existing related models.

More »

Table 2 Expand

Fig 2.

Our INBSI model for 5G network slicing, utilizing VAE for anomaly detection, CNN with NADAM for slice prediction, and SHAP/LIME for interpretability to ensure efficient resource allocation and QoS compliance.

More »

Fig 2 Expand

Fig 3.

Graphical representation of our proposed INBSI system.

More »

Fig 3 Expand

Fig 4.

The proposed Network Bandwidth Slicing Identification (INBSI) system’s top-level paradigm.

More »

Fig 4 Expand

Fig 5.

Working methodology of our proposed INBSI system.

More »

Fig 5 Expand

Fig 6.

Variational autoencoder architecture.

More »

Fig 6 Expand

Fig 7.

The proposed Model Architecture using NADAM Opitmizer.

More »

Fig 7 Expand

Table 3.

VAE performance baseline.

More »

Table 3 Expand

Fig 8.

Anomaly detection of the constructed data.

More »

Fig 8 Expand

Fig 9.

Training Loss of VAE.

More »

Fig 9 Expand

Fig 10.

Learning curves for the ML models.

Learning curves demonstrate model stability with increasing data, critical for dynamic slicing where traffic patterns evolve for machine learning models.

More »

Fig 10 Expand

Fig 11.

Confusion Matrices for the ML models.

More »

Fig 11 Expand

Fig 12.

Classification Report for the ML models.

More »

Fig 12 Expand

Fig 13.

Confusion Matrix and Classification report for DL models and Proposed CNN Model.

More »

Fig 13 Expand

Fig 14.

Training accuracy and loss curves for the DL models with our proposed model.

More »

Fig 14 Expand

Fig 15.

Comparison of the models.

More »

Fig 15 Expand

Fig 16.

Achieved scores of these models based on performance metrics.

More »

Fig 16 Expand

Fig 17.

2D PCA and t-SNE plots showing distinct clusters for network slices, indicating effective feature extraction and slice differentiation by the INBSI model.

More »

Fig 17 Expand

Fig 18.

3D PCA and t-SNE plots demonstrating well-separated clusters for network slices, further validating the INBSI model’s ability to learn meaningful, discriminative feature representations.

More »

Fig 18 Expand

Table 4.

Comparison of the models.

More »

Table 4 Expand

Table 5.

Inference Time Comparison of Models (Lower is Better).

More »

Table 5 Expand

Table 6.

Statistical Analysis of Model Performance.

More »

Table 6 Expand

Fig 19.

LIME interpretability with decision tree.

More »

Fig 19 Expand

Fig 20.

LIME interpretability with our proposed CNN.

More »

Fig 20 Expand

Fig 21.

LIME interpretibility with MLP.

More »

Fig 21 Expand

Fig 22.

SHAP feature importance and summary plot of proposed CNN.

More »

Fig 22 Expand

Fig 23.

SHAP feature importance and summary plot of proposed CNN model.

More »

Fig 23 Expand

Fig 24.

The tree-structured output of the XGBoost classifier on the SHAP model.

More »

Fig 24 Expand

Fig 25.

Slice distribution using our proposed hybrid CNN model.

More »

Fig 25 Expand

Table 7.

Model comparison used.

More »

Table 7 Expand