|
| Method year, and reference | Features used | Classification accuracy at 10 dB of SNR |
|
| Zero Crossing (1995) [21] | PDF of cross related variables | 98%
15 dB of SNR |
|
| Hierarchical Architecture (1990) [22] | Spectral features | 90% |
|
| Multilayer Perceptron, Hierarchical SVM + Bees Algorithm for Optimization (2012) [11] | Spectral features
HOM | 97.45%
(w/o optimization)
99.83%
(optimization) |
|
| SVM + PSO (2012) [12] | Spectral features
Statistical features
Wavelets features | 98.8% |
|
| Artificial Neural Network (2003) [23] | Spectral features | 93%
8 dB of SNR |
|
| Genetic Algorithm based Clustering (2011) [10] | Spectral features | 98.32% (GA)
98.12% (-mean) |
|
| Hierarchical Architecture (2000) [24] | HOC & HOM | 96% |
|
| Fuzzy based Classifier (2000) [25] | Kurtosis
phase histogram | 90%
5 dB of SNR |
|
| Multilayer Perceptron Neural Network Recognizer (2004) [13] | Spectral features
cumulants | 99.94% |
|
| Binary SVM, Multi SVM GA for Optimization (2010) [14] | HOM & HOC | 98.5%
(w/o optimization)
99.36%
(optimization) |
|
| PSO-SVM based Intelligent Classifier (2013) [17] | HOC | 96% |
|
| Gabor Filter (2014) [20] | Shift, scale, and modulation parameters | 100% |
|
| Proposed MGF Network based classifier | Shift, scale, modulation, and weights | 100%
(8 dB of SNR) |
|
