Professional Documents
Culture Documents
fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 1
Abstract—Automatic modulation classification (AMC), which The traditional feature-based AMC algorithms [3,4] can be
plays critical roles in both civilian and military applications, realized by three steps: data preprocessing, feature extraction
is investigated in this paper through a deep learning approach. and classificatory decision. Since good statistical features can
Conventional AMCs can be categorized into maximum likelihood
(ML)-based (ML-AMC) and feature-based AMC. However, the provide robust performance with low complexity [5], feature
practical deployment of ML-AMCs is difficult due to its high extraction-based AMC methods have been studied in many
computational complexity, and the manually extracted features existing papers [6,7], such as cyclic feature [8], higher order
require expert knowledge. Therefore, an end-to-end convolution moments and cumulants [9]–[11], wavelet transforms [12,13],
neural network (CNN)-based AMC (CNN-AMC) is proposed, et al. In classificatory decision, the conventional classifiers
which automatically extracts features from the long symbol-rate
observation sequence along with the estimated signal-to-noise includes random forest [14], multi-layer perception [15,16]
ratio (SNR). With CNN-AMC, a unit classifier is adopted to and support vector machine (SVM) [17,18]. These AMCs are
accommodate the varying input dimensions. The direct training combinations of different feature extraction algorithms and
of CNN-AMC is challenging with complicated model and complex classifications, and require both expert knowledge and artificial
tasks, so a novel two-step training is proposed, and the transfer design.
learning is also introduced to improve the efficiency of retraining.
Different digital modulation schemes have been considered in The AMC methods based on maximum likelihood (ML)
distinct scenarios, and the simulation results show that the CNN- are also studied in [19], where the ML-AMCs have been
AMC can outperform the feature-based method, and obtain a proved to achieve the optimal performance in the scenarios
closer approximation to the optimal ML-AMC. Besides, CNN- with mathematical channel models. Different from the feature-
AMCs have the certain robustness to estimation error on carrier based AMCs, ML-AMCs calculate likelihood functions of all
phase offset and SNR. With parallel computation, the deep
learning-based approach is about 40 to 1700 times faster than candidate modulations, and choose the modulation scheme
the ML-AMC regarding inference speed. with maximal likelihood value, working as an end-to-end
module. However, the high computational complexity makes it
Index Terms—automatic modulation classification, convolution
neural network, deep learning, two-step training. extremely difficult for the real-time implementation with low
cost [20]. Besides, the exact likelihoods in ML-AMC are hard
to achieve in the complex environments.
I. I NTRODUCTION In feature-based AMCs, the machine learning methods
In achieving the maximum possible transmission rates, simply work as a mapping function between features and
adaptive modulation schemes are used in the civilian wireless multi-hypothesis. To obviate feature engineering, the deep
communication system for full utilization of time-varying learning method [21] is developing rapidly in recent years
channels. Under these circumstances, information concerning both in algorithm design and hardware implementation [22],
a specific modulation scheme used in a communication process and it can learn significantly more complex functions than a
has to be shared by the transmitter with the receiver through a shallow one. The rapid development of deep learning has led
network protocol, at the cost of protocol overhead. Such over- to some successful applications in communications [23]–[25],
head could be eliminated when the receiver has the capability including modulation classification. In [26], the eye diagram of
of modulation scheme recognition. On the other hand, many the raw signal is used as input for Lenet-5-based classifier [27]
military applications also require automatic discovery of the and subtly relates the problem of AMC with the well-studied
modulation schemes used by signals from adversaries. Such area of image recognition. The convolutional network for
applications encompass signal interception and jamming. As a radio modulation recognition is proposed in [28,29], and the
result, automatic modulation classification (AMC) [1,2] is of simulations with a dataset called RML2016.10b show that the
great importance for both civilian and military applications classification accuracy is higher than those with expert feature
in achieving automatic receiver configuration, interference engineering. Then in [30], a LSTM-based AMC is proved
mitigation, and spectrum management. to outperform the CNN model with oversampled received
signals at small or medium scales. The long symbol-rate signal
Fan Meng and Lenan Wu are with the School of Information Science and is studied in [31], and the stacked auto-encoders achieves
Engineering, Southeast University, Nanjing 210096, China (email: mengxiao-
maomao@outlook.com, wuln@seu.edu.cn). excellent results with increased simulation runs.
Peng Chen is with the State Key Laboratory of Millimeter Waves, South- In our paper, we propose a deep neural network (DNN)
east University, Nanjing 210096, China (email: chenpengseu@seu.edu.cn), enabled AMC which can automatically learn to extract features
Peng Chen is the corresponding author of the paper.
Xianbin Wang is with the Department of Electrical and Computer Engi- from long symbol-rate signals at low SNR, and adopt the
neering, Western University, Canada (e-mail: xianbin.wang@uwo.ca). convolution neural network (CNN) to realize this AMC. The
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 2
CNN-AMC can approximate the ML-AMC with minimal 2) θc , the fixed phase offset introduced by the propagation
performance loss but significant speed improvement. The ML- delay as well as the initial carrier phase.
AMC is used as a benchmark for our proposed CNN-AMC, 3) N , the number of symbols in received sequence, i.e., the
and can also generate the training data in different scenarios. observation sequence space.
The contributions of this work are summarized as follows: 4) xk,i , the i-th constellation point under k-th modulation
• Directly using the long symbol-rate signals at low SNR, scheme where the symbols are equally distributed, i ∈
we propose an end-to-end trainable AMC based on {1, · · · , Mk }, k ∈ {1, · · · , C}. The mark Mk denoted
deep CNN. Different from the featured-based AMCs, the the symbol number using this modulation, and C is the
CNN-AMCs automatically learn to acquire features from number of candidate equiprobable modulation schemes,
the raw signals and simplify the AMC design. Besides, respectively. The modulated symbols are usually nor-
CNN-AMC makes an approximation to the ML-AMC, malized with unit power
and the related processing can be accomplished parallelly Mk
to accelerate the inference speed. 1 X
|xk,i |2 = 1. (4)
• We introduce a two-step training method including pre- Mk i=1
training and fine-tuning for the proposed CNN-AMC to
solve the challenging direct training problem. For varying In addition, symbols in a data block are assumed to be
combinations of different modulation schemes and com- independent and identical distributed (I.I.D).
munication channel scenarios, the transfer learning is also 5) ∆f , the residual carrier frequency after carrier removal.
proposed to improve the retraining efficiency further. 6) Ts , symbol interval.
• By dividing long preprocessed observation sequence into
7) g(t), the composite effect of the residual channel h(t)
signal segments with unit size, the unit classifier is and the pulse-shaping function p(t). It is expressed
introduced with the unit input dimension to accommodate by the equation g(t) = h(t) ∗ p(t), where ∗ is the
the varying input dimensions, and parallelly realize the convolution operator. Usually, p(t) is a root raised cosine
block calculation. pulse function with the duration Ts .
8) ε, the normalized epoch for timing offset between the
The remainder of this paper is organized as follows.
transmitter and the receiver, 0 ≤ ε < 1.
Section II outlines the modulation classification problem in
the framework of classical decision theory. In Section III In this paper, the received signal is expressed as the sampled
the structure of a CNN-AMC featured with multiple inputs complex baseband representation, and SNR of Es /N0 is con-
is provided, along with its corresponding two-step training, sidered. The discrete observation sequence sampled at symbol
T
transfer learning and the unit classifier. Then, in Section IV, interval Ts can be written as y = [y0 , · · · , yN −1 ] , where the
the proposed CNN-AMC, the ML-AMC and feature-based superscript T is the transpose operator. Each sampling instant
method are tested versus SNR in distinct environments, and the of the received signal is the output of a pulse-shaping matched
simulation results are analyzed. Conclusions and discussion filter at the receiver, which can be expressed as
are given in Section V. Z (n+1)Ts
1
yn = y(t)p(t − nTs ) dt. (5)
Ts nTs
II. S YSTEM M ODEL
A. Problem Formulation The pulse shape p(t) of the transmitting signal is unknown for
Without loss of generality, the complex envelope of received the receiver in the blind classification case. Similarly, instant
baseband signal can be expressed as yn is usually normalized to unit power
yn
y(t) = x(t; uk ) + v(t), (1) yn = q
1 H (6)
Ny y
where v(t) is the complex additive white Gaussian noise
(AWGN) with zero mean and double-sided power spectral where the superscript H is the conjugate transpose operator.
density being N0 /2. The noiseless signal x(t; uk ) is modeled In our classifiers, the estimation of symbol SNR is required,
via a comprehensive form as which is equivalent to having knowledge of both a and N0 .
N
X −1 The conditional probability density function (PDF) of yn
x(t; uk ) = aej(2π∆f t+θc ) xk,i
n g(t − nTs − εTs ), (2) under hypothesis Hk can be written as
n=0
Mk
|yn − xk,i 2
where uk is defined as a multi-dimensional parameter set, in
X 1 n (uk )|
p(yn |Hk , uk ) = √ exp
which a bunch of unknown signal and channel variables are i=1
Mk πN0 N0
stochastic or deterministic under the k-th modulation scheme, Mk ∗ k,i
|xk,i 2
and it is given as
1 X n | − 2Re{yn xn (uk )}
∝ exp .
n o Mk i=1 N0
uk = a, θc , ∆f, {xk,i }M
i=1 , h(t), ε .
k
(3) (7)
The symbols used in (2) and (3) are listed as follows: The elements in parameter set uk are deterministic or
1) a, the signal amplitude. random variables with known PDFs. Hence, the marginalized
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 3
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 4
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 5
TABLE I
S IMULATION PARAMETERS OF T WO - STEP T RAINING
k̂
m
A
a
g
x
r
Training Step Step 1 Step 2
Arg max Epoch Number 20 240
SNR Range/dB [0, 10] [−6, 12]
ξN
Ar
ax
m
g
Amount of Training Data 79, 200 228, 060
Softmax Amount of Validation Data 8, 800 25, 340
m
A
Using Output of ML Classifier No Yes
a
g
x
r
Dense 128 ReLU
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 6
1 1.4 1 1.2
Cross-entropy Loss
Cross-entropy Loss
0.6 0.84 0.6 0.72
Accuracy
Accuracy
0.5 0.7 0.5 0.6
Fig. 3. Training/validation curves of accuracy and loss. The x-axis denotes Fig. 4. Loss and accuracy curves of validation data versus epoch times. The
the training epoch. The validation and training curves are plotted in dark blue one using transfer learning takes less time to reach the saturation region than
and orange lines, and the accuracy and loss curves are depicted with solid the two-step training one, and the pre-training step is reduced.
and dotted lines.
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 7
impractical. On the other hand, it is also challenging to design Firstly, the conditional correct probabilities Pcc (y|Hk )
a classifier with changeable input space. curves of 7 modulations are illustrated in Fig. 5 (N = 1000).
According to the previous analysis in Subsection II-C, the The ML-AMC are presented with solid lines, and the CNN-
input y of a large N can be divided into segments with unit 1000 ones are plotted with dashed lines. Clearly 2PSK is
length. The CNN-based classifier with the input signal y of the most easily recognized modulation in our set, and its
unit input dimension is defined as a unit classifier. Theoreti- conditional correct probability is 100% even at the region of
cally, the segment has the same classification performance as low SNR. The 4PSK and 8PSK are also found to be more
the complete observation one in coherent environment. More easily identified compared to the other modulation schemes,
importantly, the scales of the unit classifier regarding training, and their Pcc (y|Hk ) is up to 1.00 when SNR = 2 dB and SNR
cache, and calculation are adjustable. Hence, the design of the = 4 dB, respectively. In high SNR field (SNR = 10 dB), all
AMC can be flexible to suit the hardware, and it is irrelevant the modulation schemes can be recognized almost without any
to the sequence space N . The inference can also be made error for ML-AMC.
parallelly for all unit classifiers. In general, there is a little Pcc (y|Hk ) performance loss when
comparing the CNN-1000 to the ML-AMC. At some SNR
IV. S IMULATION R ESULTS regions, for some modulations the Pcc (y|Hk ) are higher than
the ML-AMC when using CNN-1000. For example, in the case
The common sequence space N for AMC varies from where SNR ranging from −6 to −2 dB and the modulation
hundreds to thousands through our investigation, and consider- scheme being 16QAM, the CNN-1000 seems better than ML-
ing the hardware condition, the CNN-AMC with observation AMC. This is because the corresponding conditional probabil-
space N = 500 (named CNN-500) and N = 1000 (named ity is overestimated, at the cost of a lower correct classification
CNN-1000) are tested in this section. Besides, the ML-AMC probability of some other modulation scheme such as 64QAM.
and a feature-based AMC using 9 high-order cumulants [39] The Pc of CNN-1000 can only infinitely approximate the ML-
(named Cu-AMC) are served as the comparisons. The second AMC, and it cannot exceed the optimal one.
order, fourth order and sixth order cumulants are adopted as The curves of the overall correct probabilities for ML-
features, and a two-hidden-layer NN is used as a classifier. AMC, CNN-500, CNN-1000 and Cu-AMC are presented in
Couples of modulation schemes are considered. The overall Fig. 6. There is about 1% to 2% Pc performance degradation
and conditional correct probabilities versus SNR are the two between the CNN-AMC and the ML-AMC when N = 500
main issues needed to be studied. Besides, many other aspects and N = 1000. Moreover, the CNN-500 and CNN-1000
must be considered such as robustness against estimation error, are used as unit classifier proposed in III-E when sequence
inference time cost, and system storage requirement. space N is integer multiples of the unit input length, and
the approximation is almost very near the curves of ML-
A. Coherent AMC when N = 2000 and N = 8000. The simulation
results show that under the precise knowledge of the parameter
The simulations begin with the coherent environments, set, the CNN-based classifiers can closely approximate the
and seven modulation modes namely 2PSK, 4PSK, 8PSK, ML-AMC. On the other hand, its calculating speed is about
16QAM, 16APSK, 32APSK and 64QAM with equal prior 150× to 170× faster than ML enabled approach with given 7
probabilities are considered. The receiver is assumed to have modulation schemes in coherent scenario. Besides, the CNN-
precise knowledge of parameter set uk including the constant based classifiers outperform the Cu-AMC, and the feature-
phase offset θc and SNR, but the modulation scheme needs to based method works worse than CNN-AMCs when Pc > 90%.
be inferred. Apart from the power normalization by (6), the More importantly, the result indicates that the deep learning
observation sequence is compensated with the conjugate phase enabled approach has better performance than the manual
offset in the preprocessing period feature extraction-based method in this case.
y = ye−jθc . (24) 2) Estimation Error: The previous simulations are given
under the assumption of perfect knowledge of two parameters:
Combine (7), (8) and (9), the likelihood function under hy- fixed phase offset and SNR. However, the estimation error
pothesis Hk is explicitly given by is usually inevitable. In this part, simulations are conducted
N −1 Mk with varying degrees of estimation error, and the variable-
|xn | − 2Re{yn∗ xk,i
k,i 2
1 YX n } controlling approach is adopted.
Γ(y|Hk ) ∝ exp .
Mk n=0 i=1 N0 In the first simulation the deviation in carrier phase is inves-
(25) tigated. Because of the priori knowledge that common constel-
Hence, the normalized likelihood vector ξ N can be obtained, lation diagrams are symmetric, the influence of positive and
and tied with observation sequence y, plenty of training and negative deviation on carrier phase is equivalent. We use |∆θc |
validation data can be generated for the CNN-AMC. denote phase error and the |∆θc | ∈ {3◦ , 6◦ , 9◦ , 12◦ , 15◦ }.
1) Correct Probability: The deep learning enabled AMC is The result is shown in Fig. 7. Generally speaking, the Pc
greatly faster than the ML-AMC. While as an approximation to decreases as the |∆θc | rises. When |∆θc | < 6◦ the classifica-
the optimal AMC, it can suffer some performance degradation tion performance degradation is very small. The recognition
in terms of Pcc (y|Hk ) and Pc , which are studied in this has obviously deteriorated when |∆θc | ∈ [6◦ , 9◦ ], and the ML-
simulation. AMC gets even worse when SNR increases at the region of
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 8
1 1
0.9 0.9
0.8 0.8
0.7 0.7
2 PSK/ML
2 PSK/CNN c
=3°ML
0.6 0.6
Pcc (y|H k )
4 PSK/ML =3°CNN-1000
c
4 PSK/CNN
Pc
0.5 0.5 c
=6°ML
8 PSK/ML
8 PSK/CNN c
=6°CNN-1000
0.4 16 QAM/ML 0.4
c
=9°ML
16 QAM/CNN
0.3 0.3 c
=9°CNN-1000
16 APSK/ML
16 APSK/CNN c
=12°ML
0.2 32 APSK/ML 0.2 =12°CNN-1000
c
32 APSK/CNN
=15°ML
0.1 64 QAM/ML 0.1 c
64 QAM/CNN c
=15°CNN-1000
0 0
-6 -4 -2 0 2 4 6 8 10 -6 -4 -2 0 2 4 6 8 10
SNR (dB) SNR (dB)
Fig. 5. Pcc (y|Hk ) curves of 7 modulation modes with SNR range being Fig. 7. Pc curves of different estimation error on phase offset |∆θc | in set
[−6, 10] dB. The ML and CNN-based classifiers are depicted with solid and {3◦ , 6◦ , 9◦ , 12◦ , 15◦ }. The ML and the CNN-1000 classifiers are depicted
dotted lines, respectively. with solid and dashed lines, respectively.
1
be seen that the classifiers are sensitive to the SNR error,
0.9 even 0.5 dB estimation bias can bring about the inference
degradation. Besides, an overestimated SNR is more harmful
0.8
N=500 ML than an underestimated SNR for the classifier. We also notice
N=500 CNN-500
0.7 N=500 Cumulant+NN
that curves of CNN-AMC is smoother than the ML-AMC,
N=1000 ML and in many cases, the difference in Pc between the ML-
N=1000 CNN-500
Pc
0.6
N=1000 CNN-1000 AMC and CNN-based classifier can be larger than 10%. At
N=1000 Cumulant+NN the region of SNR = 12±1.5 dB, the Pc of ML-AMC remains
0.5 N=2000 ML
N=2000 CNN-500 1.0, which indicates that the performance saturation can be
0.4
N=2000 CNN-1000 against SNR estimation error to some extent. Meanwhile, the
N=2000 Cumulant+NN
N=8000 ML corresponding CNN-based classifier has the worst performance
N=8000 CNN-500
0.3
N=8000 CNN-1000
in all 5 comparisons. We suppose that this degradation can be
N=8000 Cumulant+NN attributed to inadequate training data. In Tab. I the training
0.2
-6 -4 -2 0 2 4 6 8 10 data is sampled in the range [−6, 12] dB, which means the
SNR (dB) CNN-AMC is not trained including this range.
Fig. 6. Pc curves of different sequence space N versus SNR. The ML-AMC, One feasible approach to make the classifier robust to the
CNN-500 and CNN-1000 are depicted with solid, dotted and dashed lines, SNR error is to train the CNN-AMC with adding noise on
respectively. SNR. This method can make the classifier less sensitive to the
deviation on SNR, but at the expense of overall classification
performance.
[8, 10] dB (|∆θc | = 9◦ ). The Pc remains about 0.86 when SNR
> 6 dB with |∆θc | = 12◦ . The Pc cannot even exceed 0.7
when |∆θc | = 15◦ . As comparison the CNN-based classifier
is about the same as the ML-AMC, and in some cases it can
perform better. In the extreme case with |∆θc | being up to
15◦ , the Pc of CNN-based classifier is higher than ML-AMC, B. Incoherent
which means the NN enabled approach is more robust.
In our second simulation, the influence of SNR estimation The coherent environment is usually not accessible in non-
error is studied. The mistake can lead to an erroneous assump- cooperative case, and the blind phase estimation algorithm is
tion of concentrations of data around the constellation points, required at the receiver. In the previous simulation classifi-
and the centers of these alphabets are also normalized with a cation performance of the CNN-AMC is greatly influenced
coefficient with deviation. when phase estimation error |∆θc | is larger than 6◦ , so the
As shown in Fig. 8, the x-axis is the SNR error ranging final classification accuracy highly depends on the employed
from −5 to 5 dB, and it is defined as ∆SNR = SNR d − SNR, estimation algorithms. In this subsection, the θc is assumed to
and the y-axis is the Pc . Curves of real SNR ∈ {0, 3, 6, 9, 12} be a random variable which is uniformly distributed in range
dB are labeled with different colors and markers. It can [0, 2π). Similarly, the likelihood function can be obtained as
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 9
1 1
0.9
0.9
0.8
0.7
0.8
0.6
Pc
Pc
0.5 0.7
SNR=0dB ML ML coherent
SNR=0dB CNN-1000 CNN coherent
0.4
SNR=3dB ML Cumulant+NN coherent
SNR=3dB CNN-1000 0.6
0.3 ML incoherent
SNR=6dB ML CNN incoherent
SNR=6dB CNN-1000 Cumulant+NN incoherent
0.2
SNR=9dB ML 0.5
SNR=9dB CNN-1000
0.1 SNR=12dB ML
SNR=12dB CNN-1000
0 0.4
-5 -4 -3 -2 -1 0 1 2 3 4 5 -6 -4 -2 0 2 4 6
SNR (dB) SNR (dB)
Fig. 8. Pc curves of different SNR in set {0, 3, 6, 9, 12} with estimation Fig. 9. Pc curves of ML-AMC and CNN-500 versus SNR in range [−6, 6]
error ∆SNR range being [−5, 5] dB. The ML-AMC and the CNN-1000 are dB. Classifiers in coherent case are plotted in light blue lines, and depicted
depicted with solid and dashed lines, respectively. with navy blue lines in incoherent scenario.
follows than that of the Cu-AMC in coherent case, but its advantage is
−1 X
Mk
( N
|xk,i 2
enlarged in the incoherent scenario. To achieve the Pc = 90%,
1 n |
Y
Γ(y|Hk ) ∝ Eθc exp − the SNR required for the incoherent classifier is up to 1.6 dB,
Mk n=0 i=1 N0
) which is about 1.0 dB more than the coherent counterpart.
2Re{yn∗ xk,i jθc
n e } Generally, the gap between the performance of the coherent
× exp and incoherent classifiers is gradually narrowed as the SNR
N0
N X
Z 2π Y Mk
" (26) increases. The correct decision probabilities of both classifiers
|xk,i |2
1 are all at the point of 100% when SNR = 4 dB.
= exp −
Mk 0 n=1 i=1 N0 The incoherent CNN-500 is also tested with observation
# space N ∈ {500, 1000, 2000}, and the ML-AMC is treated
2Re{yn∗ xk,i ejθc }
× exp dθc . as the optimal contrast in Fig. 10 with dotted lines. The text
N0 ‘slicing’ indicates that the unit classifier is used to calculate
In fact, the first Bessel function of zero order can be utilized the likelihood function, and the ML-AMC using segment is
to simplify (26), but it is not suggested in numerical calculation depicted with solid lines. Theoretically, the block calculation
due to its complex nature. In replace, the expectation over θc of observation signal is not optimal in the incoherent environ-
can be adequately approximated with the help of composite ment. From the illustration it can be seen there is about 0.2
Simpson’s rule [20] by splitting the integration interval [0, 2π) dB and 0.4 dB Pc loss with the segment number Ns being 2
into subintervals of same lengths. The range can be shorted to and 4. On the other hand, the approximation of unit CNN-500
[0, π), taking the use of constellation points symmetry in the is also very near the ones of ML-AMC using segment with
complex plane. The additional integration over θc makes the different input space. When N ∈ {1000, 2000}, the overall
total amount of calculations is several times of that in coherent correct probabilities of classifiers are up to 100% at the region
environment, and the computation time cost is proportional to of SNR = 2 dB. Although the unit classifier suffers some
the number of subintervals. performance loss, it is feasible as a sub-optimal approach
To reduce the total amount of calculations, only the 2PSK, in incoherent scenario. In addition, new CNN-AMCs with
4PSK, 8PSK and 16QAM are included in the modulation set. specific input space can be trained as an approximation to
Besides, only the CNN-500 is retrained with incoherent data the optimal method.
of observation space N = 500. In the preprocessing period, 2) Conditional Correct Probability: Through initial anal-
the phase compensation step in (24) is not needed. ysis on results of four conditional correct probabilities using
1) Coherent VS Incoherent: Both the overall probabilities ML-AMC and CNN-AMC, we find that the approximations of
of correct classification with ML-AMC, CNN-AMC and Cu- the CNN-based classifiers are quite close to ML-AMC in high
AMC are plotted in Fig. 9. It can be seen that under the SNR region, but they do not match well when SNR < 2dB.
incoherent scenario, the approximation of CNN-AMC is very To further study the conditional correct probability perfor-
close to the optimal one, and it also outperforms the Cu- mance, the confusion matrices of two AMCs are visualized in
AMC. As a comparison, the overall correct probabilities of Fig. 11 at the region of SNR = −6 dB. Each row represents
three classifiers in coherent scenario are illustrated with light the instances in a predicted class of AMC, and each column
blue lines. The Pc performance of CNN-AMC is a little better denotes the instances in a target modulation class. Comparing
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 10
2PSK
4PSK
36.3% 34.4%
Predict Class
8PSK
Precision 16QAM
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 11
1 1
0.9
0.9
0.8
0.7
0.8
0.6
ML 1 antenna
Pc
Pc
0.5 0.7
CNN 1 antenna
Cumulant+NN 1 antenna
0.4
ML 2 antennas
CNN 2 antennas 0.6
0.3 ML
Cumulant+NN 2 antennas
ML 4 antennas Cumulant+SVM-RBF
0.2 CNN 4 antennas Cumulant+SVM-POLY
0.5
Cumulant+NN 4 antennas Cumulant+DTC
0.1 Cumulant+NN
Proposed CNN
0 0.4
-18 -14 -10 -6 -2 2 6 10 14 18 -6 -4 -2 0 2 4 6
SNR (dB) SNR (dB)
Fig. 12. Plots of Pc versus SNR in range being [−18, 18] dB. The observation Fig. 14. Baseline test for various AMCs in the incoherent environment (N =
space N = 500 for each antenna, and the curves of 1, 2, 4 antennas are 500). Plots of Pc versus SNR.
illustrated with light blue, navy blue and green lines, respectively.
TABLE III
AVERAGE I NFERENCE T IME C OST OF C LASSIFIERS IN C OHERENT
E NVIRONMENT.
1 N 500 1000 2000 8000
ML-AMC/ms 10.8 21.3 42.6 164.5
0.9 CNN-500/ms 0.0727 0.146 0.288 1.15
CNN-1000/ms - 0.127 0.246 1.02
0.8
0.6
analysis and implement of ML-AMC and CNN-AMC. The
0.5
ML ML-AMC requires no memory, and the scales of the proposed
Cumulant+SVM-RBF
Cumulant+SVM-POLY
CNN-500 and CNN-1000 are small, and they occupy only 3.4
0.4 Cumulant+DTC MB and 6.3 MB capacity with N = 500 and N = 1000,
Cumulant+NN
Proposed CNN
respectively. The simulation platform is presented as: cpu
0.3 intel i7-6700k and gpu nvidia GTX-960. The ML-AMC is
-6 -4 -2 0 2 4 6 8 10
implemented with Matlab and mainly the cpu is utilized. While
SNR (dB) the CNN-AMC runs on the Python platform with the gpu.
1) Coherent: Tab. III lists the time cost per inference with
Fig. 13. Baseline test for various AMCs in the coherent environment (N = different input space of three classifiers in IV-A. The unit
500). Plots of Pc versus SNR.
time cost is defined as time cost per inference divided by
N , and they are 20.8µs, 0.144µs and 0.127µs for ML-AMC,
CNN-500 and CNN-1000, respectively. The calculating speed
is accelerated by more than two orders of magnitude with
shown in Fig. 13, the overall Pc of CNN-AMC and SVM-
NN-based approach in this specific case. More concretely, the
RBF is very similar, and the NN-based AMC suffers a little
CNN-500 and CNN-1000 are about 150 and 170 times faster
Pc performance loss compared to these two. the DTC classifier
than ML-AMC, respectively.
has the lowest classification correct rate.
2) Incoherent: In incoherent cases, the additional integra-
Then these AMCs are tested with the incoherent baseline tion on carrier phase is required for ML-AMC. Theoretically,
dataset (N = 500) in IV-B. Combining with Fig. 13, the the unit time cost is about Nθc times of that in coherent
correct classification performance of DTC classifier is still the scenario, where Nθc is the number of subintervals. After
worst, and the SVM-based methods are not stable in Fig. 14. simulation tests in IV-B1, the unit time costs listed in Tab. IV
Among the feature-based methods, only the NN-based AMC are 6.01µs and 268µs in coherent and incoherent scenarios,
can achieve the closest Pc performance to the CNN-AMC in respectively. The latter one is about 44.7 times of the coherent
both coherent and incoherent environments, and that is why one, which is approximately equal to the adopted variable
it is served as the Cu-AMC in previous simulations. Besides, value Nθc = 45.
the Pc performance of CNN-AMC is a little better than that On the other hand, the unit time cost of CNN enabled
of NN-based AMC. approach is only related to the architecture of the model.
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 12
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 13
[12] C. S. Park, J. H. Choi, S. P. Nah, W. Jang, and D. Y. Kim, “Auto- [37] D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,”
matic modulation recognition of digital signals using wavelet features Computer Science, 2014.
and svm,” in International Conference on Advanced Communication [38] S. J. Pan and Q. Yang, “A survey on transfer learning,” IEEE Transac-
Technology, 2008, pp. 387–390. tions on Knowledge & Data Engineering, vol. 22, no. 10, pp. 1345–1359,
[13] D. Boutte and B. Santhanam, “A hybrid ica-svm approach to continuous 2010.
phase modulation recognition,” IEEE Signal Processing Letters, vol. 16, [39] M. W. Aslam, Z. Zhu, and A. K. Nandi, “Automatic modulation
no. 5, pp. 402–405, 2009. classification using combination of genetic programming and knn,” IEEE
[14] Z. Zhang, Y. Li, X. Zhu, and Y. Lin, “A method for modulation Transactions on Wireless Communications, vol. 11, no. 8, pp. 2742–
recognition based on entropy features and random forest,” in IEEE 2750, 2012.
International Conference on Software Quality, Reliability and Security
Companion, 2017.
[15] M. L. D. Wong and A. K. Nandi, “Automatic digital modulation
recognition using artificial neural network and genetic algorithm,” Signal
Processing, vol. 84, no. 2, pp. 351–365, 2004.
[16] A. K. Nandi and E. E. Azzouz, “Modulation recognition using artificial
neural networks,” Signal Processing, vol. 56, no. 2, pp. 165–175, 1996.
[17] D. Wang, M. Zhang, Z. Cai, Y. Cui, Z. Li, H. Han, M. Fu, and B. Luo,
“Combatting nonlinear phase noise in coherent optical systems with Fan Meng was born in Jiangsu, China in 1992. He
an optimized decision processor based on machine learning,” Optics received the B.S. degree in 2015, from the school of
Communications, vol. 369, pp. 199–208, 2016. electronic engineering in the University of Electronic
[18] X. Zhang, J. Chen, and Z. Sun, “Modulation recognition of communi- Science and Technology of China, UESTC, and
cation signals based on schks-ssvm,” Journal of Systems Engineering & is currently working toward the Ph.D. degree in
Electronics, vol. 28, no. 4, pp. 627–633, 2017. the school of Information Science and Engineer-
[19] J. L. Xu, “Likelihood-ratio approaches to automatic modulation clas- ing, Southeast University, China. His main research
sification,” IEEE Transactions on Systems Man & Cybernetics Part C, topic is applying machine learning techniques in the
vol. 41, no. 4, pp. 455–469, 2011. wireless communication systems. Further research
[20] J. L. Xu, W. Su, and M. C. Zhou, “Software-defined radio equipped interests include machine learning in general, joint
with rapid modulation recognition,” IEEE Transactions on Vehicular demodulation and equalization, channel coding, re-
Technology, vol. 59, no. 4, pp. 1659–1667, 2010. source allocation, end-to-end communication system design with machine
learning tools.
[21] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521,
no. 7553, pp. 436–444, 2015.
[22] Y. H. Chen, T. Krishna, J. S. Emer, and V. Sze, “Eyeriss: An energy-
efficient reconfigurable accelerator for deep convolutional neural net-
works,” IEEE Journal of Solid-State Circuits, vol. 52, no. 1, pp. 127–
138, 2017.
[23] Y. He, F. R. Yu, N. Zhao, V. C. M. Leung, and H. Yin, “Software-defined
networks with mobile edge computing and caching for smart cities: A
big data deep reinforcement learning approach,” IEEE Communications
Magazine, vol. 55, no. 12, pp. 31–37, DECEMBER 2017. Peng Chen (S’15-M’17) was born in Jiangsu, China
[24] G. Gui, H. Huang, Y. Song, and H. Sari, “Deep learning for an in 1989. He received the B.E. degree in 2011 and
effective non-orthogonal multiple access scheme,” IEEE Transactions the Ph.D. degree in 2017, both from the School
on Vehicular Technology, pp. 1–1, 2018. of Information Science and Engineering, Southeast
[25] M. Chen, U. Challita, W. Saad, C. Yin, and M. Debbah, “Machine University, China. From Mar. 2015 to Apr. 2016, he
learning for wireless networks with artificial intelligence: A tutorial was a Visiting Scholar in the Electrical Engineering
on neural networks,” CoRR, vol. abs/1710.02913, 2017. [Online]. Department, Columbia University, New York, NY,
Available: http://arxiv.org/abs/1710.02913 USA.
[26] D. Wang, M. Zhang, Z. Li, J. Li, M. Fu, Y. Cui, and X. Chen, He is now an associate professor at the State
“Modulation format recognition and osnr estimation using cnn-based Key Laboratory of Millimeter Waves, Southeast Uni-
deep learning,” IEEE Photonics Technology Letters, vol. PP, no. 99, pp. versity. His research interests include radar signal
1–1, 2017. processing and millimeter wave communication.
[27] Y. Lécun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning
applied to document recognition,” Proceedings of the IEEE, vol. 86,
no. 11, pp. 2278–2324, 1998.
[28] T. J. O’Shea, J. Corgan, and T. C. Clancy, “Convolutional radio
modulation recognition networks,” pp. 213–226, 2016.
[29] T. J. O’Shea and J. Hoydis, “An introduction to deep learning for the
physical layer,” pp. 1–1, 2017.
[30] S. Rajendran, W. Meert, D. Giustiniano, V. Lenders, and S. Pollin, “Deep
learning models for wireless signal classification with distributed low-
Lenan Wu received his M.S. degree in the Electron-
cost spectrum sensors,” IEEE Transactions on Cognitive Communica-
ics Communication System from Nanjing University
tions and Networking, pp. 1–1, 2018.
of Aeronautics and Astronautics, China, in 1987,
[31] A. Ali, Y. Fan, and L. Shu, “Automatic modulation classification of
and the Ph.D. degree in Signal and Information
digital modulation signals with stacked autoencoders,” Digital Signal
Processing from Southeast University, China, in
Processing, vol. 71, 2017.
1997.
[32] P. E. Keller, Artificial Neural Networks: An Introduction, 2005.
Since 1997, he has been with Southeast Univer-
[33] V. Koltchinskii, “Neural network learning: Theoretical foundations,” sity, where he is a Professor, and the Director of
Annales De L Institut Henri Poincaré Probabilités Et Statistiques, Multimedia Technical Research Institute. He is the
vol. 39, no. 6, pp. 943–978, 2003. author or coauthor of over 400 technical papers
[34] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. MIT Press, and 11 textbooks, and is the holder of 20 Chinese
2016, http://www.deeplearningbook.org. patents and 1 International patent. His research interests include multimedia
[35] X. Glorot, A. Bordes, and Y. Bengio, “Deep sparse rectifier neural information systems and communication signal Processing.
networks,” in International Conference on Artificial Intelligence and
Statistics, 2011, pp. 315–323.
[36] P. Koniusz, F. Yan, and K. Mikolajczyk, Comparison of mid-level feature
coding approaches and pooling strategies in visual concept detection.
Elsevier Science Inc., 2013.
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2018.2868698, IEEE
Transactions on Vehicular Technology
IEEE TRANSACTION ON VEHICULAR TECHNOLOGY 14
0018-9545 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.