Table 3.
Comparison of accuracy of our proposed model with the state-of-the-art models.
| Ref. No. | Deep Learning Classifier | Type of Attacks | Dataset | Accuracy |
|---|---|---|---|---|
| [59] | Deep Q Network | Malware Detection | DREBIN | 98.79% |
| Multi-layer perceptron | 90.1% | |||
| Back Propagation Neural Network | 92.5% | |||
| [60] | Convolutional Neural Network | Spoofing,Sinkhole and Malicious Code | NSL-KDD | 98.3% |
| [63] | Support Vector Machines | IoT authentication | KDD-CUP99 | 97.36% |
| Multi-layer perceptron | 98.40% | |||
| [64] | Recurrent Neural Network-Long Short-Term Memory | Malware Detection | DREBIN | 98.18% |
| [51] | Deep Belief Network | Blackhole, DDoS, Wormhole | EMBER | 98.47% |
| [69] | Support Vector Machines | DoS, | AAGM | 96.7% |
| False Data Injection | 97.2% | |||
| Replay attack | 98.3% | |||
| Man-in-the-middle | 97.1% | |||
| Proposed Model | Convolutional Neural Network | Malware Detection | Malimg | 99% |