sensors

Article
Dual Residual Denoising Autoencoder with Channel Attention
Mechanism for Modulation of Signals
Ruifeng Duan 1,2 , Ziyu Chen 1,2 , Haiyan Zhang 1,2, *, Xu Wang 1,2 , Wei Meng 1,2 and Guodong Sun 1,2

1 School of Information Science and Technology, Beijing Forestry University, Beijing 100083, China
2 Engineering Research Center for Forestry-Oriented Intelligent Information Processing of National Forestry
and Grassland Administration, Beijing 100083, China
* Correspondence: zhyzml@[Link]

Abstract: Aiming to address the problems of the high bit error rate (BER) of demodulation or low
classification accuracy of modulation signals with a low signal-to-noise ratio (SNR), we propose
a double-residual denoising autoencoder method with a channel attention mechanism, referred
to as DRdA-CA, to improve the SNR of modulation signals. The proposed DRdA-CA consists
of an encoding module and a decoding module. A squeeze-and-excitation (SE) ResNet module
containing one residual connection is modified and then introduced into the autoencoder as the
channel attention mechanism, to better extract the characteristics of the modulation signals and
reduce the computational complexity of the model. Moreover, the other residual connection is
further added inside the encoding and decoding modules to optimize the network degradation
problem, which is beneficial for fully exploiting the multi-level features of modulation signals and
improving the reconstruction quality of the signal. The ablation experiments prove that both the
improved SE module and dual residual connections in the proposed method play an important
role in improving the denoising performance. The subsequent experimental results show that the
proposed DRdA-CA significantly improves the SNR values of eight modulation types in the range of
−12 dB to 8 dB. Especially for 16QAM and 64QAM, the SNR is improved by 8.38 dB and 8.27 dB on
average, respectively. Compared to the DnCNN denoising method, the proposed DRdA-CA makes
the average classification accuracy increase by 67.59∼74.94% over the entire SNR range. When it
comes to the demodulation, compared with the RLS and the DnCNN denoising algorithms, the
Citation: Duan, R.; Chen, Z.; Zhang,
proposed denoising method reduces the BER of 16QAM by an average of 63.5% and 40.5%, and
H.; Wang, X.; Meng, W.; Sun, G. Dual
reduces the BER of 64QAM by an average of 46.7% and 18.6%. The above results show that the
Residual Denoising Autoencoder
proposed DRdA-CA achieves the optimal noise reduction effect.
with Channel Attention Mechanism
for Modulation of Signals. Sensors
Keywords: signal denoising; convolutional autoencoder; channel attention; residual connection; AWGN
2023, 23, 1023. [Link]
10.3390/s23021023

Academic Editor: Gwanggil Jeon

Received: 15 November 2022

1. Introduction
Revised: 3 January 2023 The improvement of the SNR plays an important role in both the cooperative and
Accepted: 13 January 2023 non-cooperative fields of wireless communication systems. In cooperative communication,
Published: 16 January 2023 improving the modulation signal’s SNR can enhance the signal transmission quality and
reduce the demodulation error rate. When it comes to non-cooperative communication,
the improvement of the SNR is conducive to improving the modulation classification
accuracy, and then signal interception or interference can be performed more effectively.
Copyright: © 2023 by the authors.
Therefore, the denoising method of modulation signals has always been the research focus
Licensee MDPI, Basel, Switzerland.
of wireless communication.
This article is an open access article
In the traditional signal denoising methods, Parolai et al. in Ref. [1] use the wavelet
distributed under the terms and
transform (WT) denoising method to deal with seismic wave signals, eliminate noise
conditions of the Creative Commons
Attribution (CC BY) license (https://
coefficients below the threshold value, and then reconstruct signals through inverse WT.
[Link]/licenses/by/
However, it is difficult for the WT denoising method to select the proper parameters, includ-
4.0/). ing wavelet base, threshold, and decomposition layers. Milani et al. [2] use the least mean

Sensors 2023, 23, 1023. [Link] [Link]

Sensors 2023, 23, 1023 2 of 18

square (LMS) algorithm to eliminate the noise of acoustic signals, and Albu et al. [3] use
the recursive least squares (RLS) algorithm for signal noise reduction. The RLS algorithm is
an improvement on the LMS algorithm and it adopts recursive calculation and has better
performance than the LMS adaptive transverse filter. But Haykin et al. [4] point out that
the performance of the RLS algorithm is not stable due to its internal positive feedback
mechanism. Li et al., in Refs. [5–7], use the principal component analysis (PCA) method
to reduce the signal’s noise. The PCA algorithm recombines the original variables into
a group of new irrelevant variables by reducing the dimension of the data, and extracts
important characteristic variables according to the actual needs. However, it cannot accu-
rately estimate the number of potential hidden variables in signals [8]. Moreover, the PCA
method cannot analyze the principal components accurately when the noise’s energy is
greater than the signal’s energy. Bekara et al. [9,10] adopt the singular value decomposition
(SVD) method for noise reduction and achieve satisfactory results. The SVD algorithm is a
generalization of spectral analysis theory on arbitrary matrices, but the selections of its ef-
fective rank and the reconstruction matrix have a great influence on the performance of the
algorithm, and the matrix decomposition of the SVD algorithm is not highly interpretable.
Zhang et al. [11] propose a noise reduction multi-carrier CDSK chaotic communication
system based on Schmidt orthogonalization, and obtain good BER performance only under
certain parameter conditions by using a sliding filter to reduce noise. It can be seen that
there are still some critical issues in the traditional denoising methods that have not been
well-addressed. Most of the methods require researchers to know the channel parameters
and obtain the channel transmission characteristics by sending a training sequence, which
leads to low transmission efficiency and poor channel utilization.
Recently, deep learning has not only achieved great success in the fields of computer
vision and natural language processing, but has also provided a new solution for the
noise reduction of modulation signals. Chang et al. [12] propose a convolutional neural
network (CNN)-based hybrid cascade structure to replace the traditional equalizer in the
communication system. Wada et al. [13] use two fully connected layers (FC) as denoising
autoencoders to reduce noise of modulation signals. Zhao et al. [14] propose a deep
neural network co-evolving simultaneously at two different scales to enhance the denoising
ability of the model. Johanna et al. [15] use a CNN-based denoising model to suppress
interference in real-world radar measurements. Hwanjin Kim et al. [16] develop and
compare a machine learning (ML)-based channel predictor and a vector Kalman filter
(VKF)-based channel predictor using the spatial channel model (SCM), which is widely
used in the 3GPP standard.
Due to the limitations of the simple network model, the above five methods cannot
extract deep signal features, resulting in limited SNR improvement. In deep convolutional
networks, shallow convolution can extract detailed edge data, while deep convolution can
selectively acquire useful semantic information. Therefore, deep convolutional networks
are also widely used in signal denoising. In 2017, Zhang et al. [17] propose the residual
learning of deep CNN for image denoising (DnCNN) and achieve excellent performance.
In 2019, Khan et al. [18] extend the DnCNN method to the field of noise reduction for
modulation signals, and experiments show that it has achieved a favorable denoising effect
on high-order QAM modulation signals that are susceptible to noise interference. In the
same year, Yin et al. [19] propose a full convolutional denoising autoencoder to reduce the
noise of underwater acoustic signals, and obtain better results in both the time domain and
frequency domain, compared with traditional methods. Xue et al. [20] design a wireless
signal enhancement network based on the specialized Generative Adversarial Networks,
which can adaptively learn the characteristics of signals and realize signal enhancement in
time-varying systems. However, such methods ignore the correlation between the deep
feature maps and the shallow feature maps, and they often have high training complexity
because of deep layers. Refs. [21–26] consider the correlation between feature maps of
different scales and depths in the model’s design, and the feature graphs of different levels
are added by using residual connections [27], which effectively alleviates the difficulty of
Sensors 2023, 23, 1023 3 of 18

training deep network models and resolves the problems of gradient disappearance and
explosion. The simple connections between feature maps of different scales and depths
still make it difficult to selectively strengthen some important features, degrading the
performance of the network model.
The channel attention mechanism is a very powerful method for optimizing network
performance. It can selectively strengthen feature values of important channels by changing
the weight of the channel, thus improving the expression ability of the network and
achieving higher performance. Yancheng et al. [28] employ the residual encoder-decoder
structure and multi-attention mechanism fusion module to perform feature map reuse and
self-recalibrating features, which can significantly improve the performance of ultrasound
image denoising. Li et al. [29] propose a new end-to-end semantic segmentation network,
which integrates lightweight spatial and channel attention modules that can refine features
adaptively. Zhou et al. [30] propose to incorporate an attention mechanism including
a spatial attention module and a channel attention module into a U-Net architecture to
re-weight the feature representation spatially and channel-wise to capture rich contextual
relationships for better feature representation.
Based on the above analysis, in this paper, we adopt a deep convolutional network to
construct a denoising autoencoder, in which more effective residual connections between
different scales and depths are explored to extract richer features and optimize network
degradation problem. We also try to integrate the squeeze-and-excitation (SE) module [31]
into the network as a typical channel attention mechanism. The SE module can first
obtain the global features of each channel in the denoising network through the global
average pooling operation [32], and the global features reflect the correlations between and
effectiveness of the channel feature maps. Then, the global features are weighted to make
the channel with a strong correlation and high effectiveness have more weight, and vice
versa. By selectively strengthening the channel weights and emphasizing the features of
important channels, the SE module makes the network have a strong expression ability,
leading to superior performance. The contributions of this paper can be summarized
as follows.
(1) We choose the convolutional denoising autoencoder as the noise reduction model for
modulation signals. Residual connections are added not only inside the encoding
(decoding) block but also between the encoding (decoding) blocks of different depths
to form a double residual connection.
(2) The SE module is improved and introduced into the denoising autoencoder to opti-
mize the feature extraction of modulation signals. The improved SE module retains
the advantages of selective enhancement for channel features, and the number of
parameters is fewer, which makes the network model easier to train and achieve a
better noise reduction effect.
(3) In the decoding module of the denoising autoencoder, transpose convolution (TConv)
is selected to complete upsampling. The parameters inside TConv can be initialized
randomly and updated iteratively. At the same time, the SmoothL1Loss(SLL) function
is used as the loss function of the denoising autoencoder, which can better measure
the error between the estimated value and the real value, reconstruct a more accurate
modulation signal, and improve the accuracy of the model training.
The remainder of this paper is organized as follows. Section 2 describes the commu-
nication system model. Section 3 analyzes the applicability of the neural network and
introduces the principles and details of the proposed DRdA-CA. Section 4 describes the
dataset generation and preprocessing. The simulation results and performance analysis are
summarized in Section 5. Finally, Section 6 concludes the paper.

2. Communication System Model

In this paper, we focus on the communication scenarios under the additive white
Gaussian noise (AWGN) channel [33] with a low SNR. The AWGN channel is a widely
used channel model between the transmitter and the receiver in wireless communication
Sensors 2023, 23, 1023 4 of 18

systems, in which the noise random variable obeys the zero-mean Gaussian distribution.
We consider cooperative communication systems and non-cooperative communications
systems and the modulation types of the systems are diversified. The system model is
shown in Figure 1. At the transmitter, the source bit is modulated and sent to the channel,
in which the modulation signal s(t) will be affected by AWGN.

s(t) Denoising
Soure bit Modulation
s '(t ) network s ''(t )
n(t)

CNN modulation
Bit recovery Demodulation
recognition network

Figure 1. Communication system model.

The received modulation signal s0 (t) can be expressed as Equation (1).

s0 (t) = s(t) + n(t) (1)

where s(t) is the modulation signal with IQ components and generated at transmitter and
n(t) represents zero-mean complex AWGN vector with variance σ2 in each signal dimen-
sion. When the noise interference is serious, the received signal quality is very poor, which
will badly affect the demodulation performance. In this case, the noise reduction needs to
be performed before subsequent processing. In the cooperative communication scenario,
the modulation type is known in advance for the receiver, thus the denoised signal will be
demodulated directly. For the application of non-cooperative communication, the receiver
does not know the current modulation type, thus it needs to firstly perform modulation
recognition, as shown in the dashed box in the Figure 1. We chose the automatic modulation
classification (AMC) module proposed in [34] to finish the modulation recognition.
This paper focuses on signal denoising based on deep learning, which corresponds to
the denoising network module in Figure 1. We will develop a novel network architecture
for noise reduction of multiple modulated signals. Based on the above analysis, eight
modulated signals, which are in good agreement with the communication system model
described in our paper, are generated using MATLAB software simulation.

3. DRdA-CA Architecture Details

3.1. The Proposed Denoising Autoencoder Structure
This paper proposes a dual residual denoising autoencoder with a channel atten-
tion mechanism, and the network model can be seen in Figure 2, which is composed of
an encoding module and decoding module in series. The encoding module consists of
a 3 × 3 convolutional layer and three encoding blocks, and each encoding block is the
improved SE-ResNet module. Similarly, the decoding module is also composed of a
3 × 3 convolutional layer and three decoding blocks. It is worth noting that, in the encod-
ing and decoding modules, we further introduce the residual connection between encoding
(decoding) blocks at different levels to optimize the network degradation problem.
Sensors 2023, 23, 1023 5 of 18

Encoding module Decoding module

Encoding block1

Decoding block3
Encoding block2

Decoding block2
s

Decoding block1
Encoding block3
s’’
s’

Convolution layer1 Residual connection Residual connection Convolution layer2

(3×3) (3×3)

Figure 2. Dual residual denoising autoencoder network framework with channel attention.

This paper focuses on the AWGN channel, and the received modulation signal s0 (t)
can be expressed as Equation (1). The task of the noise reduction autoencoder is to obtain
the optimal estimation s00 (t) of the pure modulation signal s(t) from the s0 (t).
In the decoding modules of the proposed DRdA-CA, the Tanh function defined
by Equation (2) is used as the activation function of the convolution layer, which is the
output layer.
e xi − e − xi
f Tanh ( xi ) = x (2)
e i + e − xi
where xi is the input variable. The detailed network parameters of the proposed DRdA-CA
are shown in Table 1.

Table 1. Denoising Autoencoder’s Parameters.

Layer Name Output Feature Map Size

convolutional layer 1 32 × 2 × 1024
encoding block 1 32 × 2 × 1024
encoding block 2 64 × 2 × 512
encoding block 3 128 × 1 × 256
decoding block 1 128 × 2 × 521
decoding block 2 64 × 2 × 1024
decoding block 3 32 × 2 × 1024
convolutional layer 2 1 × 2 × 1024

3.2. The Encoding and Decoding Blocks

The Squeeze-and-Excitation Deep Residual Network (SE-ResNet), as the champion
model of Imagenet 2017, is very similar to the Deep Residual Network (ResNet). However,
it adds an SE module so that the information between the channels can be exchanged,
thereby improving the performance of the network. Figure 3a depicts the original SE-
ResNet module. We modify the SE unit in SE-ResNet and use the improved SE-ResNet
to establish the encoding and decoding blocks. That is, the fully connected layer (FC) in
the original SE unit is replaced by a convolutional layer. This can be attributed as follows.
(1) The SE-ResNet was originally designed and implemented under the Caffe framework,
in which the FC supports the input of data dimension 1 × 1 × C. However, the deep
learning framework Pytorch used in this paper does not support the input of data with
the same dimension. When the FC is used in the model, both the input and output data
require dimension transformation, i.e., three-dimensional data 1 × 1 × C should be reduced
and flattened into one-dimensional data C as the input of the fully connected layer. It will
increase the computational complexity of the model and destroy the spatial structure of the
Sensors 2023, 23, 1023 6 of 18

data, leading to the loss of feature information. This problem can be addressed by replacing
the FC with a 1 × 1 convolutional layer. (2) Both the convolutional layer and FC perform a
dot product operation on the feature maps, thus they have the same functional form. That
is, the 1 × 1 convolutional layer can be used to replace the FC without changing the function
operation of the SE module itself. In addition, since the value of the modulation signals
is bipolar with positive and negative numbers, we choose LeakyReLU as the activation
function. The improved SE module, as shown in Figure 3b, mainly consists of the following
two parts.
(1) The squeeze part: Assume that the dimension of the original feature map is
C × H × W, where C represents the number of channels in the feature map, H represents
the height of the feature map, and W is the width of the feature map. The task of the
squeeze part is to compress the feature map dimension C × H × W to C × 1 × 1 , which
is equivalent to squeezing the matrix in the dimension of of H × W into the matrix in the
dimension of 1 × 1 . The squeezed matrix can still obtain the previous global field of view
and has a wider perception area.
The squeeze operation is implemented using the Global Average Pooling (GAP)
method, and its calculation is described as follows:
H W
1
gc = Fsq (vc ) =
H×W ∑ ∑ vc (i, j) (3)
i =1 i =1

where vc is the feature map output by the third convolutional layer in the improved SE
module. After GAP processing for vc , we get the feature map gc , and the same below.

x
x
Conv2d+BN+LReLU
Conv2d+BN+LReLU
Conv2d+BN+LReLU
Conv2d+BN+LReLU
Conv2d+BN+LReLU
Conv2d+BN+LReLU
vc C ×H ×W
vc C ×H ×W Squeeze
Squeeze GAP
GAP Fsq(·)
Fsq(·)
gc C ×1 ×1
gc C ×1 ×1

FC W1 Conv2d
C/r × 1 × 1 C/r × 1 × 1
vc ReLU vc δ LeakyReLU
C/r × 1 × 1 Excitation C/r × 1 × 1 Excitation
FC Fex(·) W2 Conv2d Fex(·)
C ×1 ×1 C ×1 ×1
Sigmoid Sigmoid
sc sc

Scale
C ×H ×W Scale
x C ×H ×W
x
C ×H ×W
C ×H ×W

(a) (b)
Figure 3. Improved SE-Resnet and Original SE-Resnet modules. “⊕” represents addition operation.
(a) Original SE-ResNet module. (b) Improved SE-ResNet module (encoding block).

(2) The excitation part: By adding a convolutional layer above and below the nonlinear
activation layers, we parameterize the gating mechanism to predict the importance of
each channel in the feature map gc . Then the weight values of different channels are
obtained and applied to the original feature map to complete the excitation. The excitation
operation process is performed as follows. First, the compressed feature map gc is fed
into a convolutional layer with C/r output channels, namely a dimensionality-reduction
layer with reduction ratio r. Then the convolution result is activated using the LeakyReLU
function, and finally the activated feature graph is dimensionally augmented through a
convolution layer with C output channels to guarantee channel dimensional consistency
between the output sc and feature map vc . The excitation function Fex is given by

sc = Fex ( gc , W ) = σ(W2 δ(W1 gc )) (4)

Sensors 2023, 23, 1023 7 of 18

C C
where W1 ∈ R r ×C and W2 ∈ RC× r are the weight values of the two convolutional layers
around the LeakyReLU layer, and δ is the LeakyReLU function with slope parameter
α = 0.001. The LeakyReLU function solves the problem of neuron necrosis in the ReLU
function, and σ represents the sigmoid activation function. Finally, the feature map sc
output by the sigmoid function is used to adjust the channel weights of feature map vc ,
and the calculation process is given by

x̃ = Fscale (vc , sc ) = sc vc (5)

Based on above analysis, the excitation operation can map the feature map gc to a set
of channel weights vc .
The decoding block of the noise reduction autoencoder is depicted in the Figure 4. It is
clear that the decoding block is similar to the encoding block, but it adopts TConv instead
of one of the convolution layers in the original encoding block to realize the upsampling
operation. Like the convolution operation, TConv is learnable. The advantage of transpose
convolution is that it can theoretically obtain the upsampled value that is most suitable for
the current dataset through continuously updating the parameters.

Conv2d+BN+LReLU
TConv2d+BN+LReLU

Conv2d+BN+LReLU

Squeeze
GAP Fsq(·)

Conv2d
LeakyReLU
Excitation
Conv2d Fex(·)

Sigmoid

Scale

Figure 4. Decoding block.

3.3. Loss Function

In the existing work, most denoising autoencoders use the two loss functions of Mean
Absolute Error (MAE) and Mean Square Error (MSE), and carry out parameter estimation
by minimizing the error. MAE is the average absolute value of the true value minus
the predicted value. By contrast, MSE is the average square of the absolute value of the
predicted value minus the true value. Both MSE and MAE range from 0 to infinity, and
they can be calculated, respectively, using

N
1
MAE =
N ∑ |(si − si00 )| (6)
i =1

N
1
MSE =
N ∑ |(si − si00 )|2 (7)
i =1

where N is the length of the input signal, si is the noise-free signal value, and si00 is the
signal-reconstructed value (predicted value) output from our autoencoder, the same as
below. The problem existing in the MAE training process is that the update gradient is
always the same. Even for a small loss value, the gradient is relatively large, which is not
Sensors 2023, 23, 1023 8 of 18

conducive to a training model. Unlike MAE, MSE will assign more weight to outliers, and,
at the expense of the error of other samples, the model will be updated towards reducing
the error of the outliers, which will result in the performance degradation of the model. In
order to overcome the above defects and make the output of the denoising autoencoder
more accurate, the SLL is adopted in the proposed DRdA-CA in this paper. The SLL is
calculated as follows:
1 N
N i∑
SLL(s, s00 ) = zi (8)
=1
where the variable zi is defined as follows:

0.5 × |si − si00 |2 , if|si − si00 | < 1


zi = (9)
|si − si00 | − 0.5, otherwise

It is clear that SLL is an integration of MSE and MAE. When taking the derivative of
SLL, we have: 
dSSL si , if|si | < 1
= (10)
dsi ±1, otherwise
In Equation (10), it can be seen that, when si is small, the derivative of SSL with respect
to si also becomes small. Otherwise, the absolute value of the derivative reaches the upper
bound of 1, and the network stability will not be impacted by the large derivative. As we
know, MSE has derivatives at the origin and is easy to converge, and, in the boundary area,
MAE enables our denoising autoecoder to correct back when the error tends to be large.
Regarding the two error metrics, therefore, the SLL is more robust with respect to outliers.

4. Dataset
4.1. Experimental Data and Environment
The dataset of the denoising autoencoder includes eight modulation signals generated
by MATLAB; they are BPSK, QPSK, 8PSK, CPFSK, GMSK, OQPSK, 16QAM, and 64QAM,
and the detailed parameters of the dataset are shown in Table 2. The SNR of the modulation
signals ranges from −12 dB to 8 dB, and the SNR interval is 2 dB. In the transmission
process of the signal, the transmitter completes the modulation and up-conversion of the
signal, and then transmits radio frequency signals into the channel. The receiver first
performs down-conversion, and subsequently estimates the carrier frequency and phase,
restores the received signal to the baseband signal, and performs modulation recognition,
demodulation, and other operations. Without the loss of generality, this paper takes the
baseband signal as an example to generate the dataset of the modulation signals. The
rate of the baseband signal is set to 256 kHz, and the sampling frequency of the signal
is 1024 kHz—that is, the oversampling rate of the signal is 4 times higher to increase the
fault tolerance rate of the signal. The modulated signals are shaped and filtered using
a root raised cosine filter with a roll-off factor of 0.3, and the transmission channel is
modelled as an AWGN channel. Pytorch is used as the deep learning framework, and
an NVIDIA GeForce RTX2080 Ti graphics card is used to implement the GPU parallel
acceleration calculation.

Table 2. Parameter settings for data sample sets

Parameter Value/Description
SNR −[Link] (dB)
Symbol rate 256k Baud
Sample rate 1024 kHz
Oversampling rate 4
Forming filters Root raised cosine filter
Roll-off factor 0.3
Channel AWGN channel
Sensors 2023, 23, 1023 9 of 18

4.2. Dataset Preprocessing

In order to facilitate the addition of noise and the measurement of the SNR, as well
as to accelerate the training of the model and reduce the computational complexity of the
data, we normalize the power of the modulation signals generated by MATLAB simulation
using the following formulae:
∑ N | s i |2
Ps = i=1 (11)
N

S
Ns = √ (12)
Ps
where si is the initial noise-free modulation signal, N is the length of the signal, Ps is the
signal power, and Ns is the noise-free modulation signal with normalized power. In the
training stage of the denoising autoencoder, the signal dataset with 11 different SNR values
is made into an IQ data sample with the shape of 2 × 1024 for each piece. We split the
training set and the values set by the ratio of approximately 8:2, and there are 70,400 pieces
of training data and 17,600 pieces of validation data. Then we generate 8800 pieces of test
data under the same parameters. The Adaptive Moment Estimation (Adam) optimization
algorithm is used to complete model training with 40 training epochs, and the initial
learning rate is set to 0.001 while decreasing by 0.1 times every 20 epochs.

5. Experiment and Analysis

5.1. Layer Number Optimization
We conduct simulation experiments to evaluate our denoising autoencoder, and we
also compare it with a set of typical models including RLS, LMS, PCA, and the deep learning
method DnCNN proposed by Khan et al. [18]. The evaluation metrics are SNR, Error Vector
Magnitude (EVM), and demodulation BER before and after noise reduction. In addition,
the modulation recognition accuracy is considered and the modulation recognition model
is the CNN [34].
The SNR is defined as the ratio of noiseless signal power to noise power, which can
intuitively measure the quality of signal. The calculation formula is given by the following:

||s||22
SNRdB = 10lg( ) (13)
||s0 − s||22
where s represents the noiseless modulation signal, and s0 represents the noisy modulation
signal. The parameter EVM can comprehensively measure the amplitude error and phase
error of the modulation signal, which is defined as the ratio of the root mean square value
of the error vector signal and the root mean square value of the ideal signal, and expressed
in the form of a percentage. The calculation formula is given by the following:
v
u1 N
u
u N ∑ |si − si00 |2
u i =1
EV M = u × 100% (14)
t 1 N
u
N ∑ i | s | 2
i =1

where s is the same as before, and si00 represents the signal-reconstructed value (predicted
value). In the first experiment, we determine the optimal depth of encoder-decoder modules.
We design three encoder module structures, whose encoder-decoder depths equal 2, 3, and
4, respectively, and the corresponding neural networks are called DRdA-CA2, DRdA-CA3,
and DRdA-CA4, respectively. The training loss and validation loss of the three network
models are shown in Figure 5. The detailed value of the training loss with different epochs
and the numbers of parameters are listed in Table 3. In terms of the training loss, DRdA-
CA3 is 1.4–1.9% lower than DRdA-CA2 on average when the epoch is larger than 10, and
the validation loss presents a similar trend. It should be noted that the performance of
Sensors 2023, 23, 1023 10 of 18

DRdA-CA3 and DRdA-CA4 is almost the same. When it comes to the complexity, the size
of the model parameters of DRdA-CA2, DRdA-CA3, and DRdA-CA4 are 181 k, 263 k, and
345 k, respectively. Therefore, considering both the loss and model complexity, we choose
DRdA-CA3 as the neural network model structure for the denoising tasks. Although the
loss values have been small at the epoch equaling 10, they continue to decline and basically
stabilize at 40 rounds. In this paper, the model parameters adopt the training results of
40 rounds.

0.2
tra-loss2
0.18 val-loss2
tra-loss3
0.16 val-loss3
tra-loss4
0.14 val-loss4

0.12
loss

0.1

0.08

0.06

0.04

0.02

0
0 5 10 15 20 25 30 35 40
epoch

Figure 5. Classification accuracy at different depths.

Table 3. The Parameters of Different Depths.

Model Parameter
DRdA-CA2 181,408
DRdA-CA3 263,416
DRdA-CA4 345,424

5.2. Ablation Experiments

In this paper, the SE module is first improved and then introduced into the encoder as
the channel attention module, so that the modulation signal features can be better extracted
and the computational complexity of the model can be reduced. In order to validate
the efficiency of the improved SE module as the channel attention mechanism, we carry
out the first ablation experiment based on the control variable method. We compare the
improved SE module (impr-SE) using the convolutional layer with the original SE module
(orig-SE) using FC and the denoising autoencoder without a channel attention mechanism
(withoutCA). In this case, the SE module is removed in the coding block in Figure 3b.
Figure 6a displays the training loss (tra-loss) and validation loss (val-loss) of the three
methods. As is shown, the module with the improved SE module gives better performance.
When the epoch equals 40, it lower the validation loss by 35.7% and 17.9% compared to
the module with the original SE and the model without a channel attention mechanism.
In addition, it should be noted that the denoising model without a channel attention
mechanism is even superior to the model with the original SE. This can be attributed to
the loss of feature information caused by data dimension transformation. Therefore, a
mismatched channel attention module may cause network performance degradation, and
the improved SE is quite effective.
Sensors 2023, 23, 1023 11 of 18

0.35
tra-loss-withoutCA tra-loss-single
val-loss-withoutCA val-loss-single
0.3 tra-loss-impr-SE 0.3
tra-loss-dual
val-loss-impr-SE val-loss-dual
tra-loss-orig-SE
0.25 0.25
val-loss-orig-SE

0.2 0.2
0.018

loss

loss
0.016
0.15 0.014 0.15
0.012
0.1 0.01 0.1
0.008
0.05 32 34 36 38 40 0.05

0 0
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
epoch epoch
(a) (b)

Figure 6. Ablation experiments. (a) Original SE and improved SE. (b) Single and dual residual connections.

In the proposed DRdA-CA, “dual” represents the residual connections added not
only between different encoding (decoding) blocks but also inside the encoding (decoding)
blocks of different depths, so it is called “dual”. In contrast, “single” refers to the residual
connection only existing inside the encoding (decoding) block, and each encoding (decod-
ing) block is the improved SE-ResNet module, which contains one residual connection.
In order to demonstrate the advantages of “dual residual connections”, one of the main
innovations, we have also conducted an ablation experiment based on the control variable
method, comparing it with the “single residual connection” method, where the residual
connection only exists inside the encoding (decoding) block. The loss curves are given in
Figure 6b. As shown in the figure, the dual residual model achieves a lower training loss
and validation loss than the single residual model. When the epoch increases to 40, the
validation loss of the dual residual connections is only 50% of that of the single residual
connection. Comparing Figure 6a,b, apparently, both the improved SE module and dual
residual connections play an important role in improving denoising performance, and dual
residual connections are more effective.

5.3. Comparison of SNR Improvement

In the second experiment, we compare the SNR and EVM variations of different
denoising methods. Table 4 displays the average values of SNR improvement in overall
SNR, ranging from −12 dB to 8 dB for all eight modulation modes with five different
denoising methods. In the SNR range of −12 dB to 8 dB, taking the GMSK signal as an
example, the SNRs of the modulation signal are increased by 6.76 dB, 6.5 dB, and 7.18 dB,
respectively, on average, after denoising with RLS, LMS, and PCA algorithms. When the
DnCNN model is used for noise reduction, the SNRs of the signal are improved by 11.59 dB
on average. Remarkably, the proposed DRdA-CA brings an increase of 16.8 dB on average
in terms of the SNRs. When DRdA-CA is applied to 8PSK, 16QAM, and 64QAM signals,
the average values of SNR enhancement are 9.47 dB, 8.38 dB, and 8.27 dB, respectively. For
two deep learning-based denoising methods, DnCNN and DRdA-CA, the average values
of SNR improvement gradually go up with the decrease in modulation order, except in
a few cases. However, traditional noise reduction methods do not show similar laws. It
is clear that DRdA-CA outperforms the DnCNN method and RLS, LMS, and PCA. This
can be attributed to the well-designed structure in the proposed method, in which both
double-residual connections and the introduction and improvement of the SE module
contribute to a desirable denoising effect.
It would be complicated to use curves to display dynamic ranges of SNR and EVM
after denoising with the increase of SNR for all eight modulation modes. As a result, we only
present four typical modulation modes in Figures 7 and 8. Figure 7a–d respectively show
the SNR variations of GMSK, 8PSK, 16QAM, and 64QAM, and four signals are processed
using different denoising methods. It can be seen from Figure 7 that the deep learning
Sensors 2023, 23, 1023 12 of 18

denoising methods, denoted as DRdA-CA and DnCNN, are superior to the three traditional
denoising methods, denoted as LMS, RLS, and PCA. The proposed denoising method
achieves different denoising effects for different modulation modes. Although denoising
is more challenging for high-order modulation due to the complex signal patterns, the
proposed denoising achieves a desirable denoising performance for high-order modulation.
Moreover, for the four modulation types, the proposed DRdA-CA has always maintained
the maximum SNR improvement over the entire SNR range.

35 18
RLS RLS
16 LMS
30 LMS
PCA 14 PCA
25 DRdA-CA DRdA-CA

SNR after denoising/dB

SNR after denoising/dB

DnCNN 12 DnCNN
20
10

15 8

10 6

4
5
2
0
0
−5 −2
−12 −10 −8 −6 −4 −2 0 2 4 6 8 −12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR before denoising/dB SNR before denoising/dB
(a) (b)

16 14

RLS RLS
14 12
LMS LMS
12 PCA 10 PCA
DRdA-CA DRdA-CA
SNR after denoising/dB
SNR after denoising/dB

DnCNN 8
10 DnCNN
8 6

6 4

4 2

2 0

0 −2

−2 −4
−12 −10 −8 −6 −4 −2 0 2 4 6 8 −12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR before denoising/dB SNR before denoising/dB
(c) (d)

Figure 7. SNR of modulation signal after denoising [2,3,5–7,18]. (a) GMSK. (b) 8PSK. (c) 16QAM.
(d) 64QAM.

Figure 8a–d also show the EVM curves of the above four modulation signals processed
using different noise reduction methods. The lower the EVM, the better the noise reduction
performance. As seen from Figure 8, the EVMs obtained using the two deep learning-
based denoising methods are significantly lower than those obtained using the other three
traditional denoising methods. After PCA denoising, the EVMs of the signals do not
steadily decline with the increase in the SNR, which indicates that the robustness of PCA
is poor. The signal EVM curves obtained after RLS and LMS denoising are close to each
other when the SNR is greater than 4 dB. However, when the SNR falls in −12 dB∼4 dB,
RLS is better than LMS. The deep learning algorithm DnCNN performs better than the
three traditional algorithms do, and the proposed DRdA-CA makes a better achievement
in denoising than DnCNN does.
Sensors 2023, 23, 1023 13 of 18

Table 4. Average value of SNR improvement under different denoising methods.

RLS LMS PCA DnCNN DRdA-CA

GMSK 6.76 dB 6.50 dB 7.18 dB 11.59 dB 16.84 dB
CPFSK 6.14 dB 5.75 dB 8.45 dB 10.11 dB 15.39 dB
BPSK 6.15 dB 5.74 dB 8.46 dB 11.12 dB 11.86 dB
QPSK 6.14 dB 5.74 dB 6.21 dB 9.17 dB 13.44 dB
OQPSK 6.14 dB 5.76 dB 6.17 dB 8.99 dB 13.43 dB
8PSK 6.13 dB 5.74 dB 5.44 dB 8.29 dB 9.47 dB
16QAM 6.13 dB 5.75 dB 5.43 dB 7.9 dB 8.38 dB
64QAM 6.14 dB 5.78 dB 6.19 dB 7.79 dB 8.27 dB

140 140
RLS RLS
120 LMS 120
LMS
PCA PCA
DRdA-CA
DRdA-CA
Error vector amplitude (%)

Error vector amplitude(%)

100 100 DnCNN
DnCNN

80 80

60 60

40 40

20 20

0 0
−12 −10 −8 −6 −4 −2 0 2 4 6 8 −12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR/dB SNR/dB

(a) (b)

140 140

RLS RLS
120 LMS 120 LMS
PCA PCA
DRdA-CA
Erro vector amplitude(%)
Error vector amplitude(%)

100 100 DRdA-CA

DnCNN
DnCNN
80 80

60 60

40 40

20 20

0 0
−12 −10 −8 −6 −4 −2 0 2 4 6 8 −12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR/dB SNR/dB
(c) (d)

Figure 8. EVM after denoising of modulation signal [2,3,5–7,18]. (a) GMSK. (b) 8PSK. (c) 16QAM.
(d) 64QAM.

5.4. Comparison of AMC Performance

For non-cooperative communication applications, the receiver needs to firstly per-
form modulation classification, followed by demodulation and decoding. Therefore, the
modulation classification experiment is also done to further verify the performance of the
denoising autoencoder. The LMS and RLS algorithms need to know the modulation mode
and send the training sequence when they are used to reduce noise in communication sys-
tems, leading to low transmission efficiency. Thus, they are unsuitable for non-cooperative
communication applications. PCA also has the lowest denoising performance. As a result,
Sensors 2023, 23, 1023 14 of 18

in the experiment of the noise reduction cascade AMC, we compare the DnCNN to the
proposed method. Moreover, the CNN [34] is adopted to implement AMC, and the dataset
is given in Section 4.1, including eight modulation signals under eleven SNRs.
The experiment schemes are as follows. (1) The CNN model is used for AMC directly,
denoted as CNN. (2) The DnCNN method is used to reduce the noise of modulation signals,
and then CNN is used for AMC, denoted as DnCNN+CNN. (3) The proposed DRdA-CA
is used to decrease the noise, followed by CNN for AMC, denoted as DRdA-CA+CNN.
Figure 9 displays the average classification accuracy of all the modulation modes for three
different schemes. The modulation recognition accuracy is greatly increased when using
denoising methods based on deep learning. The DRdA-CA method clearly outperforms
the DnCNN method. When the SNR varies from −12 dB to 8 dB, compared with the non-
denoising scheme, the scheme with DnCNN denoising improves the average classification
accuracy by 0.01%∼12.5%. After using the proposed denoising approach, the average
classification accuracy is increased by 1.88%∼29.25%. Compared to the DnCNN denoising
method, the proposed DRdA-CA makes the average classification accuracy increase from
67.59%∼74.94% over the entire SNR range, which further shows the superiority of the
proposed method.

100

90

80

70
Accuracy (%)

60

50

40
CNN
30 DRdA-CA+CNN
DnCNN+CNN
20
−12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR/dB

Figure 9. Modulation classification accuracy under different deep learning networks.

Figure 10a,b give the recognition accuracy of each modulation type with the latter
two schemes. As seen in Figure 10a, after DnCNN denoising, the recognition accuracy of
the three low-order modulations, BPSK, GMSK, and CPFSK, is relatively high, and almost
reaches 100% under the SNR of −4 dB. This is because these three types are low-order
modulations and the patterns of the signals are simple and significantly different from
those of the high-order PSK and QAM. Therefore, their classification accuracy is also high,
even through general DnCNN denoising.
However, the recognition performance of the high-order modulations is poor. In
particular, the recognition accuracy of the 16QAM and 64QAM signals fluctuates widely.
This is attributed to the highly similar characteristics of higher-order modulation signals
and the insufficient noise reduction capability of DnCNN. It should be noted that the
classification accuracy of QPSK and OQPSK is lower than that of 8PSK in most of the SNR
intervals, although their modulation order is smaller than that of 8PSK. This is because the
QPSK and OQPSK signal patterns are too similar, and they are very easy to confuse when
the noise reduction effect is limited with the DnCNN method.
According to the Figure 10b, DRdA-CA denoising makes the recognition accuracy
of each modulation signal increase significantly. Although 16QAM and 64QAM still
cannot achieve accurate classification when the SNR is larger than 2 dB, their recognition
accuracy exhibits a continuous rising trend with an increase in the SNR. It demonstrates
Sensors 2023, 23, 1023 15 of 18

that the strong denoising ability of the proposed DRdA-CA is beneficial for AMC. The
classification accuracy of 16QAM and 64QAM can be further improved by adopting better
modulation classification model after denoising, and this paper focuses on the design of
the denoising network.

100 100

90 90

80 80

70 70
Classification accuracy (%)

Classification accuracy (%)

60 60

50 50

40 40

8PSK 8PSK
30 BPSK 30 BPSK
CPFSK CPFSK
GMSK GMSK
20 20
OQPSK OQPSK
16QAM 16QAM
10 64QAM 10
64QAM
QPSK QPSK
0 0
−12 −10 −8 −6 −4 −2 0 2 4 6 8 −12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR/dB SNR/dB
(a) (b)

Figure 10. Classification accuracy of different networks for each modulation type. (a) DnCNN+CNN.
(b) DRdA-CA+CNN.

5.5. Comparison of Demodulation Performance

In order to further verify the performance of the proposed DRdA-CA, the demod-
ulation experiments after denoising have also been done. We present two higher-order
modulation modes, 16QAM and 64QAM, since their demodulations are more challenging
in the low SNR range for practical communication systems. The RLS algorithm with better
performance in the traditional methods and the DnCNN are selected as the comparison
methods. Figure 11 shows the BER curves of demodulation for the 16QAM and 64QAM
signals after denoising with different methods.
From Figure 11, it can be seen that the three noise reduction methods can improve
the BER performance of signal demodulation. The DnCNN improves BER more than the
traditional RLS, and the proposed DRdA-CA considerably outperforms both of these two
methods. When the SNR ranges from −12 dB to 8 dB, the proposed method decreases
the BER by 22.3%∼99.5% for 16QAM and 12.6%∼93.7% for 64QAM compared to the non-
denoising method. In addition, compared with RLS and DnCNN, DRdA-CA reduces the
BER of 16QAM by 63.5% and 40.5% on average, and makes the BER of 64QAM decline
by 46.7% and 18.6% on average, across the entire SNR range. For 16QAM signals at an
SNR of 8 dB, when using the RLS and DRdA-CA, the BER of demodulation is reduced by
two orders of magnitude, from 0.05 to 0.0005. Besides, compared with DnCNN, the BER
of DRdA-CA is only 1/6 of the former. When it comes to the 64QAM signals at an SNR
of 8 dB, compared with the non-denoising method, DRdA-CA makes BER decrease from
0.195 to 0.0123, and compared to the RLS and the DnCNN, it reduces the BER of 64QAM
by 91.49% and 43.34%, respectively.
Sensors 2023, 23, 1023 16 of 18

10 0

10−1

Bit Error Rate

10−2

Non-denosing (16QAM)
DRdA-CA (16QAM)
DnCNN (16QAM)
10−3 RLS (16QAM)
Non-denoding (64QAM)
DRdA-CA (64QAM)
DnCNN (64QAM)
RLS (64QAM)
10−4
−12 −10 −8 −6 −4 −2 0 2 4 6 8
SNR/dB

Figure 11. Demodulation BER after denoising.

6. Conclusions
In this paper, a dual residual denoising autoencoder with channel attention (DRdA-
CA) is designed based on CNN to deal with the problems caused by a low SNR in wireless
communications. First, as a channel attention mechanism, the SE module with one residual
connection is adapted and applied to the proposed DRdA-CA to improve the feature
extraction ability of neural networks. Then, the other residual connection between the
different coding (decoding) blocks is introduced to solve the network degradation problem,
which promotes the fusion of model characteristic information at different depths, and
further enhances the network noise reduction performance. Moreover, a dataset including
eight modulation types is created under the AWGN channel model.
The ablation experiments are done to verify the efficiency of the improved SE module
and the dual residual connections, and the dual residual connections are more helpful
in improving the denoising performance. Then, the experiments on SNR improvement,
modulation classification, and demodulation are also done to verify the advantage of the
proposed method. The simulation results show that the proposed DRdA-CA surpasses
the traditional denoising algorithms and the DnCNN noise reduction method in both
improving the SNR and recovering the original signals for all eight modulation modes.
Taking GMSK and 8PSK as examples, the proposed DRdA-CA improves the SNR by
16.84 dB and 9.47 dB on average, respectively. Compared to the DnCNN denoising method,
the proposed DRdA-CA makes the recognition accuracy of each modulation signal increase
significantly. Even the recognition accuracy of 16QAM and 64QAM goes up steadily
with the increase in the SNR. Besides, demodulation performance is also dramatically
improved after using the proposed DRdA-CA to reduce noise. Taking the SNR of 8 dB as
an example, for 64QAM signals, the proposed DRdA-CA lowers BER by 43.34% compared
to the DnCNN method. These experiments further demonstrate our method’s strength.
In the future, we will extend our designs and make it suitable for the fading-channel
scenario, where modulation signals are prone to be affected by multipath propagation and
the Doppler frequency shift.

Author Contributions: Conceptualization, R.D. and Z.C.; methodology, R.D. and H.Z.; validation,
R.D., Z.C. and H.Z.; formal analysis, H.Z. and X.W.; investigation, G.S.; resources, W.M.; data curation,
Z.C.; writing—original draft preparation, R.D. and X.W.; writing—review and editing, Z.C. and
W.M.; visualization, H.Z. and G.S.; supervision, W.M.; project administration, R.D. and H.Z.; funding
acquisition, R.D. All authors have read and agreed to the published version of the manuscript.
Funding: The work was supported by the Fundamental Research Funds for the Central Universities
(BLX201623), Beijing Natural Science Foundation (L202003) and National Natural Science Foundation
of China (No.31700479).
Sensors 2023, 23, 1023 17 of 18

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: The dataset is available from the website [Link]
0/mod_dataset.git. (accessed on 14 November 2022).
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript:

BER Bit Error Rate

SNR Signal-to-Noise Ratio
DRdA-CA Double-Residual Denoising Autoencoder Method with Channel Attention Mechanism
WT Wavelet Transform
LMS Least Mean Square
RLS Recursive Least Squares
PCA Principal Component Analysis
SVD Singular Value Decomposition
CNN Convolutional Neural Network
FC Fully Connected Layers
ML Machine Learning
VKF Vector Kalman Filter
SCM Spatial Channel Model
DnCNN Residual Learning of Deep CNN for Image Denoising
SE Squeeze-and-Excitation
TConv Transpose Convolution
SLL SmoothL1Loss
AWGN Additive White Gaussin Noise
AMC Automatic Modulation Classification
SE-ResNet Squeeze-and-Excitation Deep Residual Networks
ResNet Deep Residual Network
GAP Global Average Pooling
MAE Mean Absolute Error
MSE Mean Square Error
Adam Adaptive Moment Estimation
EVM Error Vector Magnitude

References
1. Parolai, S. Denoising of Seismograms Using the S Transform. Bull. Seismol. Soc. Am. 2009, 99, 226–234. [CrossRef]
2. Milani, A.A.; Panahi, I.M.S.; Briggs, R.W. LMS-Based Active Noise Cancellation Methods for fMRI Using Sub-band Filtering. In
Proceedings of the 2006 International Conference of the IEEE Engineering in Medicine and Biology Society, New York, NY, USA,
30 August–3 September 2006; pp. 513–516.
3. Albu, F.; Paleologu, C. A recursive least square algorithm for active noise control based on the Gauss-Seidel method.
In Proceedings of the 2008 15th IEEE International Conference on Electronics, Circuits and Systems, St. Julian’s, Malta,
31 August–3 September 2008; pp. 830–833.
4. Haykin, S. Adaptive Filter Theory; Pearson Education: Upper Saddle River, NJ, USA, 1986.
5. Li, X.L.; Anderson, M.; Adalı, T. Principal component analysis for noncircular signals in the presence of circular white gaussian
noise. In Proceedings of the 2010 Conference Record of the Forty Fourth Asilomar Conference on Signals, Systems and Computers,
Pacific Grove, CA USA, 7–10 November 2010; pp. 1796–1801.
6. Pyatykh, S.; Hesser, J.W.; Zheng, L. Image Noise Level Estimation by Principal Component Analysis. IEEE Trans. Image Process.
2013, 22, 687–699. [CrossRef] [PubMed]
7. Zhao, Q.; Meng, D.; Xu, Z.; Zuo, W.; Zhang, L. Robust Principal Component Analysis with Complex Noise. In Proceedings of the
ICML, Beijing, China, 21–26 June 2014.
8. Peng, F.; Gao, Y. Noise reduction of BPSK signals based on convolutional self-coding networks. Inf. Commun., 2020, 8, 41–44.
9. Bekara, M.; van der Baan, M. Local singular value decomposition for signal enhancement of seismic data. Geophysics 2007,
72, V59–V65. [CrossRef]
10. Rajwade, A.; Rangarajan, A.; Banerjee, A. Image Denoising Using the Higher Order Singular Value Decomposition. IEEE Trans.
Pattern Anal. Mach. Intell. 2013, 35, 849–862. [CrossRef] [PubMed]
Sensors 2023, 23, 1023 18 of 18

11. Zhang, G.; He, H.; Zhang, P. NR-MC-CDSK Chaotic Communication System Based on Schmidt Orthogonalization. J. Electron. Inf.
Technol. 2021, 43, 1930–1938.
12. Chang, Z.; Wang, Y.; Li, H.; Wang, Z. Complex CNN-Based Equalization for Communication Signal. In Proceedings of the 2019
IEEE 4th International Conference on Signal and Image Processing (ICSIP), Wuxi, China, 19–21 July 2019; pp. 513–517.
13. Wada, T.; Toma, T.; Dawodi, M.; Baktash, J.A. A Denoising Autoencoder based wireless channel transfer function estimator for
OFDM communication system. In Proceedings of the 2019 International Conference on Artificial Intelligence in Information and
Communication (ICAIIC), Okinawa, Japan, 11–13 February 2019; pp. 530–533.
14. Zhao, T.; Zhong, Y.; Wang, Y. Parallel multi-scale CNN for image denoising. In Proceedings of the 5th International Conference
on Communication and Information Processing (ICCIP ’19), Chongqing, China, 15–17 November 2019.
15. Rock, J.; Tóth, M.; Meissner, P.; Pernkopf, F. Deep Interference Mitigation and Denoising of Real-World FMCW Radar Signals. In
Proceedings of the 2020 IEEE International Radar Conference (RADAR), Washington, DC, USA, 28–30 April 2020; pp. 624–629.
16. Kim, H.; Kim, S.; Lee, H.; Choi, J. Massive MIMO Channel Prediction: Machine Learning Versus Kalman Filtering. In Proceedings
of the 2020 IEEE Globecom Workshops (GC Wkshps), Taipei, Taiwan, 7–11 December 2020; pp. 1–6.
17. Zhang, K.; Zuo, W.; Chen, Y.; Meng, D.; Zhang, L. Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image
Denoising. IEEE Trans. Image Process. 2017, 26, 3142–3155. [CrossRef] [PubMed]
18. Khan, S.; Khan, K.S.; Shin, S.Y.K. Symbol Denoising in High Order M-QAM using Residual learning of Deep CNN. In Proceedings of the
2019 16th IEEE Annual Consumer Communications & Networking Conference (CCNC), Las Vegas, NV, USA, 11–14 January 2019; pp. 1–6.
19. Yin, J.; Luo, W.; Li, L.; Han, X.; Guo, L.; Wang, J. Enhancement of underwater acoustic signal based on denoising automatic-encoder.
J. Commun. 2019, 40, 119–126.
20. Zhou, X.; Sun, Z.; Wu, H. Wireless signal enhancement based on generative adversarial networks. Ad. Hoc. Netw. 2020, 103, 102151.
[CrossRef]
21. Jiang, Y.; Li, H.; Rangaswamy, M. Deep Learning Denoising Based Line Spectral Estimation. IEEE Signal Process. Lett. 2019,
26, 1573–1577. [CrossRef]
22. Jin, K.H.; McCann, M.T.; Froustey, E.; Unser, M.A. Deep Convolutional Neural Network for Inverse Problems in Imaging. IEEE
Trans. Image Process. 2017, 26, 4509–4522. [CrossRef] [PubMed]
23. Wang, Y.; Tu, L.; Guo, J.; Wang, Z. Residual learning based RF signal denoising. In Proceedings of the 2018 IEEE International
Conference on Applied System Invention (ICASI), Tokyo, Japan, 13–17 April 2018; pp. 15–18.
24. Wang, Y.; Huang, H.; Xu, Q.; Liu, J.; Liu, Y.; Wang, J. Practical Deep Raw Image Denoising on Mobile Devices. arXiv 2020,
arxiv:2010.06935.
25. Kang, E.; Chang, W.; Yoo, J.; Ye, J.C. Deep Convolutional Framelet Denoising for Low-Dose CT via Wavelet Residual Network.
IEEE Trans. Med. Imaging 2018, 37, 1358–1369. [CrossRef] [PubMed]
26. Casas, L.; Navab, N.; Belagiannis, V. Adversarial Signal Denoising with Encoder-Decoder Networks. In Proceedings of the 2020
28th European Signal Processing Conference (EUSIPCO), Amsterdam, The Netherlands, 18–21 January 2021; pp. 1467–1471.
27. He, K.; Zhang, X.; Ren, S.; Sun, J. Deep Residual Learning for Image Recognition. In Proceedings of the 2016 IEEE Conference on
Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 27–30 June 2016; pp. 770–778.
28. Yancheng, L.; Zeng, X.; Dong, Q.; Wang, X. RED-MAM: A residual encoder-decoder network based on multi-attention fusion for
ultrasound image denoising. Biomed. Signal Process. Control 2023, 79, 104062.
29. Li, H.; Qiu, K.; Chen, L.; Mei, X.; Hong, L.; Tao, C. SCAttNet: Semantic Segmentation Network With Spatial and Channel
Attention Mechanism for High-Resolution Remote Sensing Images. IEEE Geosci. Remote Sens. Lett. 2019, 18, 905–909. [CrossRef]
30. Zhou, T.; Canu, S.; Ruan, S. Automatic COVID-19 CT segmentation using U-Net integrated spatial and channel attention
mechanism. Int. J. Imaging Syst. Technol. 2020, 31, 16–27. [CrossRef]
31. Hu, J.; Shen, L.; Albanie, S.; Sun, G.; Wu, E. Squeeze-and-Excitation Networks. IEEE Trans. Pattern Anal. Mach. Intell. 2020,
42, 2011–2023. [CrossRef] [PubMed]
32. Zhang, Y.; Li, G.; Lei, J.; He, J. FF-CAM: Crowd counting based on frontend-backend fusion through channel-attention mechanism.
Comput. Sci. 2020, 44, 304–317.
33. Chirag, B.; Lohith, A.; Prasantha, H.S. Comparative performance analysis of various digital modulation schemes in AWGN
channel. In Proceedings of the 2017 Innovations in Power and Advanced Computing Technologies (i-PACT), Vellore, India, 21–22
April 2017; pp. 1–5.
34. O’Shea, T.; Corgan, J.; Clancy, T.C. Convolutional Radio Modulation Recognition Networks. arXiv 2016, arxiv:1602.04105.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.