|
Approach | Identification type | Features | Technique | Purpose |
|
Corbett et al. [3] | NIC vendor | Traffic rate | Spectral analysis | Protects networks |
Bratus et al. [4] | Chipsets and drivers | Response to crafted frames | Decision list learning | Identifies fake APs |
Gao et at. [5] | APs | Interarrival time | Wavelet analysis | Detects unsafe APs |
Loh et al. [8] | 802.11 devices | Probe request frames | Timing analysis | Protects networks |
Neumann et al. [10] | 802.11 devices | Transmission time and interarrival time | Histogram and cosine similarity | Device fingerprinting |
Dalai and Jena[11] | Wi-Fi devices | Correlation-based feature selection | Similarity measure | Device fingerprinting |
Shahid et al. [12] | Wi-Fi devices | Packet size and time | T-SNE technology, machine learning | Device behavior description |
Bezawada et al. [13] | Wi-Fi devices | Packet header and payload features | Machine learning | Device type identification |
Dong et al. [14] | Wi-Fi devices | Packet header and payload features | LSTM | Device type identification |
Mirsky et al. [19] | IP camera | Temporal statistics features | Unsupervised ANN | Anomaly detection |
Bovenzi et al. [20] | Wi-Fi devices | TCP/IP stack layer features | Deep autoencoders, machine learning, and double-censoring mechanism | Anomaly detection |
Acar et al. [22] | Wi-Fi, BLE, and ZigBee devices | Timing features, sensor state and controller state features, and controller location features | Machine learning | Device privacy leakage |
Singh et al. [23] | Wi-Fi-based wireless sensors | MAC address, cause-effect relationship | Granger causality, dead reckoning | Device privacy protection |
|