Automatic modulation classification using modulation fingerprint extraction

doi:10.23919/JSEE.2021.000069

Abstract

Abstract:

An automatic method for classifying frequency shift keying (FSK), minimum shift keying (MSK), phase shift keying (PSK), quadrature amplitude modulation (QAM), and orthogonal frequency division multiplexing (OFDM) is proposed by simultaneously using normality test, spectral analysis, and geometrical characteristics of in-phase-quadrature (I-Q) constellation diagram. Since the extracted features are unique for each modulation, they can be considered as a fingerprint of each modulation. We show that the proposed algorithm outperforms the previously published methods in terms of signal-to-noise ratio (SNR) and success rate. For example, the success rate of the proposed method for 64-QAM modulation at SNR=11 dB is 99%. Another advantage of the proposed method is its wide SNR range; such that the probability of classification for 16-QAM at SNR=3 dB is almost 1. The proposed method also provides a database for geometrical features of I-Q constellation diagram. By comparing and correlating the data of the provided database with the estimated I-Q diagram of the received signal, the processing gain of 4 dB is obtained. Whatever can be mentioned about the preference of the proposed algorithm are low complexity, low SNR, wide range of modulation set, and enhanced recognition at higher-order modulations.

Key words: automatic modulation classification, in-phase-quadrature (I-Q) constellation diagram, spectral analysis, feature based modulation classification

Jafar NOROLAHI, Paeiz AZMI, Farzaneh AHMADI. Automatic modulation classification using modulation fingerprint extraction[J]. Journal of Systems Engineering and Electronics, 2021, 32(4): 799-810.

Figures/Tables 12

Table 1

Fig 1

Fig 2

Fig 3

Fig 4

Table 2

Fig 5

Fig 6

Fig 7

Fig 8

Table 3

Table 4

Performance comparison of the proposed algorithm with other automatic classifiers"

Algorithm classification	AMC	Modulation set	Complexity	Success rate/%	SNR for 16QAM/dB	Advantage	Disadvantage
Proposed algorithm: FPMC	FB	OFDM, 2FSK, 4FSK, MSK, 4QAM, 16QAM, 2QAM, 64QAM, 8PSK, 16PSK	Low	?99?	3	Low SNR Wide modulation set Recognition of higher-order QAM	?
ALRT [5], L=1, ηA=1 L=likelihood ratio; ηA= average likelihood ratio	LB	BPSK, QPSK, 16QAM, 32QAM, 64QAM	High	99	7	Maximum probability of classification	Multidimensional integration Impractical [48]
HLRT [5], μ, H not specified μH= hybrid likelihood ratio	LB	BPSK, QPSK, 8PSK, 16PSK, 16QAM, 64QAM	High	99	9	Treated as deterministic known Overcome to nested Constellation	Un-conditional PDF
Quasi-HLRT [5], threshold = 1	LB	16QAM, 32QAM, 64QAM	Low	99	19	Low-complexity Enhanced performance	Require high SNR Disabled in high-order QAM
Cumulant-based [11], Nm=2, μH=?0.68 Nm=Number of modulation	FB	2ASK, 2FSK, 4FSK, 8FSK, BPSK, QPSK, MSK	Low	99	9	Low SNR Robust in phase and frequency	Sub-optimal performance [37]
AMC with ELM [29]?	FB	BPSK, QPSK, 8-PSK, 16-QAM, 64-QAM, 4-ASK	Low	?99 ?	?7?	High accuracy Low SNR levels Robustness	Impractical Limited modulation set
GPOS [37], symbol length 4096	FB	BPSK, QPSK, 8-PSK, 16-QAM, 64-QAM	Low	?99	15	Accelerate the feature process Satisfactory performance Stronger robustness	Limited modulation set
[26]	FB	BPSK, QPSK, 16QAM, 64QAM	Low	?98	11	Good performance	Disable in high-order QAM
AMC with HMLN [36]	FB	? BPSK, QPSK, 16QAM, LFM, SF, 2FSK, 4FSK	Low	?99.26	?30?	Wide modulation set	Requires high SNR
[27]	FB	4ASK, 8ASK, 16ASK, 2PSK, 4PSk, 8PSK, 16PSK, 4FSK8FSK, 6FSK, 8QAM, 16QAM, MSK, OOK	Low	100	20	Reduced complexity	Large set of modulations Need high SNR Need many features
[34]	FB	4QAM, 16QAM, 32QAM, 64QAM, 128QAM, 256QAM	Low	100	15	Enhanced recognition Higher-order QAM	Require high SNR
[35]	FB	16QAM, 32QAM, 64QAM	Low	100	15	Enhanced performance Without prio information	Need high SNR Disable in high-order QAM

Table 4

References 48

1	MAREY M, DOBRE O A Blind modulation classification algorithm for single and multipleantenna systems over frequency-selective channels. IEEE Signal Processing Letters, 2014, 21 (9): 1098- 1102. doi: 10.1109/LSP.2014.2323241
2	GOKCAY D, EKEN A, BALTACI S Binary classification using neural and clinical features: an application in fibromyalgia with likelihood-based decision level fusion. IEEE Journal of Biomedical and Health Informatics, 2019, 23 (4): 1490- 1498. doi: 10.1109/JBHI.2018.2844300
3	GHASEMZADEH P, BANERJEE S, HEMPEL M, et al Accuracy analysis of feature-based automatic modulation classification with blind modulation detection. Proc. of the International Conference on Computing, Networking and Communications, 2019, 1000- 1004.
4	ALI A, FAN Y Y Automatic modulation classification using principle composition analysis based features selection. Proc. of the Computing Conference, 2017, 294- 296.
5	DOBRE O A, ABDI A, BAR-NESS Y, et al Survey of automatic modulation classification techniques: classical approaches ‎and new trends. IET Communications, 2007, 1 (2): 137- 156. doi: 10.1049/iet-com:20050176
6	KIM K, POLYDOROS A. Digital modulation classification: the BPSK versus QPSK case. Proc. of the Military Communications Conference, 1988: 431−436.
7	LAYN E, POLYDOROS A Per-survivor processing for channel acquisition, data detection and modulation classification. Proc. of the 28th Asilomar Conference on Signals, Systems and Computers, 1994, 1169- 1173.
8	LIANG H, HOK C Antenna array likelihood modulation classifier for BPSK and QPSK signals. Proc. of the IEEE Military Communications Coference, 2002, 647- 651.
9	SILLS J A Maximum-likelihood modulation classification for PSK/QAM. Proc. of the IEEE Military Communications Conference, 1999, 217- 220.
10	LEE J H, KIM J, KIM B, et al Robust automatic modulation classification technique for fading channels via deep neural network. Entropy, 2017, 19, 1- 11.
11	SWAMI A, SADLER B M Hierarchical digital modulation classification using cumulants. IEEE Trans. on Communications, 2000, 48 (3): 416- 429. doi: 10.1109/26.837045
12	UYS L Y, GOUWS M, STRYDOM J J, et al The performance of feature-based classification of digital modulations under varying SNR and fading channel conditions. Proc. of the IEEE AFRICON, 2017, 198- 203.
13	WANG F G, WANG X D Fast and robust modulation classification via kolmogorov-smirnov test. IEEE Trans. on Communications, 2010, 58 (8): 2324- 2332. doi: 10.1109/TCOMM.2010.08.090481
14	KURNIANSYAH H, WIJANTO H, SURATMAN F Y Automatic modulation detection using non-linear transformation data extraction and neural network classification. Proc. of the International Conference on Control, Electronics, Renewable Energy and Communications, 2018, 213- 216.
15	LEE J H, KIM K, SHIN Y. Feature image-based automatic modulation classification method using CNN algorithm. Proc. of the International Conference on Artificial Intelligence in Information and Communication, 2019: 1−4.
16	XU Y H, WANG J L, WU Q H, et al Opportunistic spectrum access in unknown dynamic environment: a game-theoretic stochastic learning solution. IEEE Trans. on Wireless Communications, 2012, 11 (4): 1380- 1391. doi: 10.1109/TWC.2012.020812.110025
17	GAO Z K, ZHU Z C. Distribution test based low complexity modulation classification in MIMO systems. Proc. of the 10th International Conference on Wireless Communications and Signal Processing, 2018: 1−5.
18	HAN L B, GAO F F, LI Z, et al Low complexity automatic modulation classification based on order-statistics. IEEE Trans. on Wireless Communications, 2017, 16 (1): 400- 411. doi: 10.1109/TWC.2016.2623716
19	AZARMANESH O, BILEN S G. New results on a two-stage novel modulation classification technique for cognitive radio applications. Proc. of the Military Communications Conference, 2011: 266−271.
20	AZARMANESH O, BILEN S G I-Q diagram utilization in a novel modulation classification technique for cognitive radio applications. EURASIP Journal on Wireless Communications and Networking, 2013, 2013 (1): 289. doi: 10.1186/1687-1499-2013-289
21	BAGGA J, TRIPATHI N Automatic modulation classification using statistical features in fading environment. International Journal of Advanced Research in Electried, Electronics and Instrumentation Engineering, 2013, 2 (8): 2320- 3765.
22	SHEN L, LI S J, SONG C S, et al. Automatic modulation classification of MPSK signals using high order cumulants. Proc. of the 8th International Conference on Signal Processing, 2006, DOI: 10.1109/ICOSP.2006.344507.
23	DAS D, BORA P K, BHATTACHARJEE R. Cumulant based automatic modulation classification of QPSK, OQPSK, 8-PSK and 16-PSK. Proc. of the 8th International Conference on Communication Systems and Networks, 2016: 1−5.
24	ALI A, FAN Y Y Higher-order statistics based modulation classification using hierarchical approach. Proc. of the IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, 2016, 370- 374.
25	ZHU Z, ASLAM M W, NANDI A K. Support vector machine assisted genetic programming for MQAM classification. Proc. of the International Symposium on Signals, Circuits and Systems, 2011: 1−6.
26	ASLAM M W, ZHU Z, NANDI A K Automatic modulation classification using combination of genetic programming and KNN. IEEE Trans. on Wireless Communications, 2012, 11 (8): 2742- 2750.
27	WALENCZYKOWSKA M, KAWALEC A Type of modulation identification using wavelet transform and neural network. Bulletin of the Polish Academy of Sciences: Technical Sciences, 2016, 64 (1): 257- 261. doi: 10.1515/bpasts-2016-0028
28	PRAKASAM P, MADHESWARAN M Digital modulation identification model using wavelet transform and statistical parameters. Journal of Computer Systems, Networks, and Communications, 2008, 1- 8.
29	GUNER A, ALÇIN O F, ŞENGUR A Automatic digital modulation classification using extreme learning machine with local binary pattern histogram features. Measurement, 2019, 145, 214- 225. doi: 10.1016/j.measurement.2019.05.061
30	MOSER E, MORAN M K, HILLEN E, et al Automatic modulation classification via instantaneous features. Proc. of the National Aerospace and Electronics Conference, 2015, 218- 223.
31	VALIPOUR M H, HOMAYOUNPOUR M M, MEHRALIAN M A Automatic digital modulation recognition in presence of noise using SVM and PSO. Proc. of the 6th International Symposium on Telecommunications, 2012, 378- 382.
32	ZHU X L, LIN Y, DOU Z Automatic recognition of communication signal modulation based on neural network. Proc. of the IEEE International Conference on Electronic Information and Communication Technology, 2016, 223- 226.
33	SATIJA U, MANIKANDAN M S, RAMKUMAR B. Performance study of cyclostationary based digital modulation classification schemes. Proc. of the 9th International Conference on Industrial and Information Systems, 2014: 1−5.
34	WANG L, LI Y B Constellation based signal modulation recognition for MQAM. Proc. of the 9th IEEE International Conference on Communication Software and Networks, 2017, 826- 829.
35	CHOU Z D, JIANG W N, XIANG C B, et al Modulation recognition based on constellation diagram for M-QAM signals. Proc. of the 11th IEEE International Conference on Electronic Measurement & Instruments, 2013, 70- 74.
36	WANG F, HUANG S S, WANG H, et al Automatic Modulation classification exploiting hybrid machine learning network. Mathematical Problems in Engineering, 2018, 2018, 1- 14.
37	HUANG S, JIANG Y Z, QIN X Q, et al Automatic modulation classification of overlapped sources using multi-gene genetic programming with structural risk minimization principle. IEEE Access, 2018, 6, 48827- 48839. doi: 10.1109/ACCESS.2018.2868224
38	HARING L, CHEN Y, CZYLWIK A. Efficient modulation classification for adaptive wireless OFDM systems in TDD mode. Proc. of the IEEE Wireless Communication and Networking Conference, 2010: 1−6.
39	LI H, BAR-NESS Y, ABDI A, et al. OFDM modulation classification and parameters extraction. Proc. of the 1st International Conference on Cognitive Radio Oriented Wireless Networks and Communications, 2006: 1−6.
40	JOSHI H, DARAK S J. Sub-Nyquist sampling and machine learning based online automatic modulation classifier for multi-carrier waveform. Proc. of the 32nd General Assembly and Scientific Symposium of the International Union of Radio Science, 2017: 1−4.
41	THAKUR P S, MADAN S, MADAN M A novel method for automatic classification of multicarrier signals on noisy HF channel. Proc. of the International Conference on Computer, Communication and Control, 2015, 1- 6.
42	ZHANG J, LI Y, YIN J P Modulation classification method for frequency modulation signals based on the time-frequency distribution and CNN. IET Radar, Sonar & Navigation, 2018, 12 (2): 244- 249.
43	ZHU X, FUJII T A novel modulation classification method in cognitive radios based on features clustering of ‎time-frequency. Proc. of the IEEE Radio and Wireless Symposium, 2016, 45- 47.
44	YAN X, FENG G Y, WU H C, et al Innovative robust modulation classification using graph-based cyclic-spectrum analysis. IEEE Communications Letters, 2017, 21 (1): 16- 19. doi: 10.1109/LCOMM.2016.2618868
45	TAN J L, SHA'AMERI A Z B, CHEE Y M Signal analysis and classification of digital communication signal using time-frequency analysis techniques in the multipath fading environment. Proc. of the 10th International Conference on Information Science, Signal Processing and their Applications, 2010, 678- 681.
46	YA T, LIN Y, WANG H Modulation recognition of digital signal based on deep auto-ancoder network. Proc. of the IEEE International Conference on Software Quality, Reliability and Security Companion, 2017, 256- 260.
47	AL-NUAIMI D H, HASHIM I A, ABIDIN I S Z L, et al Performance of feature-based techniques for automatic digital modulation recognition and classification-a review. Electronics, 2019, 8 (12): 1407. doi: 10.3390/electronics8121407
48	VAN TREES H L, BELL L. Detection estimation and modulation theory, part I: detection, estimation, and filtering theory. 2nd ed. New York: Wiley, 2013.

Algorithm	Feature type	Decision making	Extracted feature	Modulation set
[36]	Time-frequency	Naive Bayesian and SVM	Discrete Fourier transform	Binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), 16-QAM, LFM, SF, 2-FSK, 4-FSK
[36]	Time-frequency	Naive Bayesian and SVM	Instantaneous autocorrelation
[29]	Spectral	Extreme learning machine	Amplitude	BPSK, QPSK, 8-PSK, 16-QAM, 64-QAM, 4-ASK
[29]	Spectral	Extreme learning machine	Phase information	BPSK, QPSK, 8-PSK, 16-QAM, 64-QAM, 4-ASK
[37]	Statistical	Genetic programming	Original cumulants	BPSK, QPSK, 8-PSK, 16-QAM, 64-QAM
[11]	Statistical	PDF-based	Fourth-order cumulants	BPSK, 4-ASK, 16-QAM, 8-PSK, V32, V29, V29c
[32]	Cyclo-stationary	ANN	Cyclic frequency	2-ASK, 2-FSK, 4-FSK, 8-FSK, BPSK, QPSK, MSK
[32]	Cyclo-stationary	ANN	Spectral	2-ASK, 2-FSK, 4-FSK, 8-FSK, BPSK, QPSK, MSK
[31]	Spectral and statistical	SVM with PSO	Higher-order	2-ASK, 4-ASK, 2-FSK, 4-FSK, 2-PSK, 4-PSK
[31]	Spectral and statistical	SVM with PSO	Statistical and wavelet	2-ASK, 4-ASK, 2-FSK, 4-FSK, 2-PSK, 4-PSK
[25]	Statistical	GP and SVM	Higher-order cumulants	16-QAM, 64-QAM
[27]	Statistical	PCA and ANN	Mean value and variance	4-ASK, 8-ASK, 16ASK, 2-PSK, 4-PSK, 8-PSK, 16-PSK, 4-FSK, 8-FSK, 16-FSK, 8-QAM, 16-QAM, MSK, on off keying (OOK)
[27]	Statistical	PCA and ANN	Central moments
[30]	Spectral	DT	Amplitude	2-ASK, 4-ASK, 8-ASK, 2-FSK, 4-FSK, 8-FSK, 2-PSK, 4-PSK, 8-PSK
[30]	Spectral	DT	Phase and frequency
[22]	Statistical	DT	Fourth-order cumulant	BPSK, QPSK, FSK, MSK
[23]	Statistical	DT	Fourth-order	QPSK, offset quadrature phase shift keying (OQPSK), 8-PSK, 16-PSK
[23]	Statistical	DT	Zero-conjugate cumulant
[24]	Statistical	DT	Instantaneous amplitude	2-ASK, 4-ASK, 8-ASK, BPSK, QPSK, 8-PSK
[24]	Statistical	DT	Higher-order cumulants	2-ASK, 4-ASK, 8-ASK, BPSK, QPSK, 8-PSK
[33]	Cyclo-stationary	KNN	Cyclo-stationarity	BPSK, QPSK, FSK, MSK
[26]	Statistical	GP and KNN	Higher-order cumulants	BPSK, QPSK, 16-QAM, 64-QAM
[34]	Time-frequency	ML	Constellation	4-QAM, 16-QAM, 32-QAM, 64-QAM, 128-QAM, 256-QAM.
[35]	Time-frequency	ML	Constellation	16-QAM, 32-QAM, 64-QAM

Modulation type	Sequence length	Reference coordinate	Reference sequence
64-QAM	64	(1,1)	d={0.0 2.0 2.0 2.0 2.0 2.82 2.82 2.82 2.82 4.0 4.0 4.0 4.0 4.47 4.47 4.47 4.47 4.47 4.47 4.47 4.47 5.65 5.65 5.65 5.65 6.0 6.0 6.0 6.0 6.36 6.32 6.32 6.32 6.32 6.32 6.32 6.32 7.21 7.21 7.21 7.21 7.21 7.21 7.21 7.21 8.0 8.0 8.24 8.24 8.24 8.24 8.48 8.48 8.48 8.48 8.94 8.94 8.94 8.94 10.0 10.0 10.0 10.0 11.31};
32-QAM	32	(?3, ?5)	d={0.0 2.0 2.0 2.8 2.82 4.0 4.0 4.47 4.47 4.47 5.65 6.0 6.0 6.32 6.32 6.32 7.21 7.21 8.0 8.24 8.24 8.24 8.48 8.94 8.94 10.0 10.0 10.0 10.19 10.77 11.31 11.66};
16-QAM	16	(3,3)	d = {0.0 2.0 2.0 2.82 4.0 4.0 4.47 4.47 5.65 6.00 6.32 6.32 6.0 8.48 7.21 7.21}
8-QAM	8	(3,3)	d={2.0 2.82 4.0 4.47 4.47 5.65 6.32 ?? 7.21}?
4-QAM	4	(1,1)	d={0 2.0 2.0 ??2.82}?
4-FSK	Variable	(0,0)	Sequence with any elements ={1, 1, 1, 1, 1 $,\cdots $ }
2-FSK	Variable	(0,0)	Sequence with any elements ={1, 1, 1, 1, 1 $,\cdots $ }
MSK	Variable	(0,0)	Sequence with any elements ={1, 1, 1, 1, 1 $,\cdots $ }
?4-PSK	4	(0,0)	d ={0.0 1.41 1.41 2.0}
8-PSK	8	(0,0)	d={0.0 0.76 0.76 1.41 1.41 1.84 1.84 2.0}
16-PSK	16	(0,0)	d={0 0.39 0.39 0.76 0.76 1.11 1.11 1.41 1.41 1.66 1.66 1.84 1.84 1.96 1.96 2.0}

Modulation type	SNR/dB (99% success rate)	SNR/dB (80% success rate)	Symbol	Iteration
64-QAM	11	10	$ {10}^{3} $	$ {10}^{5} $
32-QAM	9	8	$ {10}^{3} $	$ {10}^{5} $
16-QAM	3	2	$ {10}^{3} $	$ {10}^{5} $
4-QAM	3	2	$ {10}^{3} $	$ {10}^{5} $
4-FSK	5	4	$ {10}^{3} $	$ {10}^{5} $
2-FSK	7	6	$ {10}^{3} $	$ {10}^{5} $
MSK	7	6	$ {10}^{3} $	$ {10}^{5} $
8-PSK	9	8	$ {10}^{3} $	$ {10}^{5} $
16-PSK	13	12	$ {10}^{3} $	$ {10}^{5} $

[1]	Sheng Zhang, Guangcai Sun, and Mengdao Xing. Full aperture imaging algorithm for highly squinted TOPS SAR [J]. Journal of Systems Engineering and Electronics, 2016, 27(6): 1168-1175.
[2]	Hamidreza Hosseinzadeh. Tracking performance of large margin classifier in automatic modulation classification with a software radio environment [J]. Journal of Systems Engineering and Electronics, 2014, 25(5): 735-741.
[3]	Xiaojian Xu and Jia Li. Ultrawide-band radar imagery from multiple incoherent frequency subband measurements [J]. Journal of Systems Engineering and Electronics, 2011, 22(3): 398-404.
[4]	Yunhe Cao, Tao Zhang, Shouhong Zhang, and Binfeng Luo. Forward scattering bistatic radar imaging method and#br# practice data processing [J]. Journal of Systems Engineering and Electronics, 2011, 22(2): 206-211.