Table 8 Computational complexity of the proposed model over existing model.
From: A hybrid deep learning model for detection and mitigation of DDoS attacks in VANETs
Metric | ADA | EGOA | Hybrid (ADA +  EGOA) | Description |
|---|---|---|---|---|
Execution time (sec) | 12.5 | 10.2 | 15.8 | Total time taken for feature selection |
Memory usage (MB) | 180 | 165 | 210 | RAM consumption during execution |
Computational complexity | O(n2) | O(nlogn) | O(n log n) | Theoretical complexity analysis |
Convergence iterations | 65 | 55 | 48 | Number of iterations to reach the optimal solution |
Feature selection rate (%) | 72.5 | 68.9 | 81.2 | Percentage of total features selected |
Feature reduction (%) | 27.5 | 31.1 | 18.8 | Reduction in feature dimensionality |
Detection accuracy (%) | 93.2 | 91.8 | 96.5 | Accuracy of the selected features in classification |
Precision (%) | 90.4 | 89.7 | 94.1 | Correct positive predictions |
Recall (%) | 92.1 | 90.2 | 95.6 | Correctly identified attacks |
F1-Score | 91.2 | 89.9 | 94.8 | Harmonic mean of precision & recall |
