Table 3.
Different state–of–the–art DFD systems used IoT–based architecture and machine learning algorithms.
| Cited. | Sensors/Parameters | Algorithms | Cloud Environment | Processing Cost | Overhead |
|---|---|---|---|---|---|
| [1] Hu X et al. (2015) | Drivers’ social context | NA | Cloud Server | Low | Low |
| [126] Muñoz et al. (2016) | PERCLOS, blink time and blink rate | NN | Body Sensor Networks (BSNs) with Vehicular ad hoc Networks (VANETs) | Middle | Middle |
| [136] Ming et al. (2017) | EEG | Fuzzy Entropy and SVM | Cloud Server | High | High |
| [159] Škrjanc et al. (2018) | Driver by speed, revolutions, steering-wheel and pedals etc., without using intelligence sensors | NA | Cloud Server | High | High |
| [160] Dasgupta et al. (2018) | mobile sensors | NA | multi-tier vehicular social network (M-VSN) |
SVM: support vector machine.