| | Method and reference | Number of modulation | Classification accuracy in % at SNR = 10 dB | No. of features |
| | Pdf of the received signal, zero crossing [32] | PSK, FSK (7) | 98 (15 dB of SNR) | 4 | | Spectral features [33] | PSK, FSK (6) | 90 | 5 | | Spectral features and moments [34] | PSK, FSK (9) | 96 (15 dB of SNR | 9 | | Constellation shape [35] | PSK, QAM (3) | 90 (5 dB of SNR | 4 | | HOC [36] | PSK, QAM (8) | 96 | 2 | | Neural networks [37] | PSK, FSK, QAM (7) | 93 (8 dB of SNR) | 5 | | Back propagation (BP) algo., resilient BP algo [26] | PSK, FSK, QAM (10) | 89.96, 99.95 | 17 | | Time frequency features [29] | FSK, PSK (6) | 97.64 | 4 | | Normalized cummulants [28] | PSK, QAM (5) | 97.5 | 1 | | Cummulants features, multiclass SVM [25] | PSK, QAM (9) | 97.56 | 12 | | Spectral features and moments [30] | FSK, QAM (12) | 99.95 | 8 | | HOC and spectral features, multiclass SVM [27] | PSK, QAM, FSK (11) | 97.45 | 11 | | Gabor filter based features [31] | PSK, FSK, QAM (18) | 100 | 3 | | HOC and SVM [38] | QAM (6) | 98.7 (5 dB of SNR) | 11 | | HOC and polynomial classifier [39] | PSK (5) | 98 | 10 | | Proposed optimized classifier | PSK, FSK, QAM (13) | 100 (8 dB of SNR) | 3 |
|
|
