electronics

Review
Deep Learning for Channel Estimation in Physical Layer
Wireless Communications: Fundamental, Methods,
and Challenges
Chaoluo Lv 1 and Zhongqiang Luo 1,2, *

1 School of Automation and Information Engineering, Sichuan University of Science and Engineering,
Yibin 644000, China; 322085404127@stu.suse.edu.cn
2 Artificial Intelligence Key Laboratory of Sichuan Province, Sichuan University of Science and Engineering,
Yibin 644000, China
* Correspondence: luozhongqiang@suse.edu.cn

Abstract: With the rapid development of wireless communication technology, intelligent communica-
tion has become one of the mainstream research directions after the fifth generation (5G). In particular,
deep learning has emerged as a significant artificial intelligence technology widely applied in the
physical layer of wireless communication for achieving intelligent receiving processing. Channel
estimation, a crucial component of physical layer communication, is essential for further information
recovery. As a motivation, this paper aims to review the relevant research on applying deep learning
methods in channel estimation. Firstly, this paper briefly introduces the conventional channel esti-
mation methods and then analyzes their respective merits and drawbacks. Subsequently, this paper
introduces several common types of neural networks and describes the application of deep learning
in channel estimation according to data-driven and model-driven approaches, respectively. Then, this
paper extends to emerging communication scenarios and discusses the existing research on channel
estimation based on deep learning for reconfigurable intelligent surface (RIS)-aided communication
systems. Finally, to meet the demands of next-generation wireless communication, challenges and
future research trends in deep-learning-based channel estimation are discussed.

Citation: Lv, C.; Luo, Z. Deep

Keywords: channel estimation; deep learning; wireless communication; physical layer; intelligent
Learning for Channel Estimation in
communication; RIS
Physical Layer Wireless
Communications: Fundamental,
Methods, and Challenges. Electronics
2023, 12, 4965. https://doi.org/
1. Introduction
10.3390/electronics12244965
In recent decades, there has been a rapid advancement in wireless communication
Academic Editor: Dimitra I.
technology. As a mobile communication technology that enables the “Internet of Every-
Kaklamani
thing”, fifth generation (5G) technology has high-speed, large-connectivity, and low-latency
Received: 3 November 2023 properties. Its primary application scenarios include massive machine-type communica-
Revised: 8 December 2023 tions (mMTC), enhanced mobile broadband (eMBB), and ultra-reliable and low-latency
Accepted: 9 December 2023 communications (uRLLC) [1]. According to forecasts from the International Data Corpora-
Published: 11 December 2023 tion (IDC), by 2025, an astonishing 152,200 Internet of Things (IoT) devices will be deployed
every minute. Additionally, the data volume generated by global IoT devices is projected
to reach 73.1 ZB. The exponential growth of wireless device access presents significant
challenges to 5G, as its performance increasingly needs to improve to meet the explosive
Copyright: © 2023 by the authors.
increase in mobile data traffic and the diversified service demands. Given this, forward-
Licensee MDPI, Basel, Switzerland.
looking research on sixth generation (6G) technology is actively underway. Compared to
This article is an open access article
distributed under the terms and
5G, 6G is expected to achieve order-of-magnitude improvements in key metrics such as
conditions of the Creative Commons
throughput, spectral efficiency, energy efficiency, transmission rate, and security [2,3].
Attribution (CC BY) license (https://
In wireless communication systems, the channel is complex and dynamic. Signals may
creativecommons.org/licenses/by/ suffer from attenuation and distortion due to the multipath effect, Doppler effect, shadow
4.0/). fading, etc., thus affecting the performance of the communication systems. The accuracy of

Electronics 2023, 12, 4965. https://doi.org/10.3390/electronics12244965 https://www.mdpi.com/journal/electronics

Electronics 2023, 12, 4965 2 of 38

transmitted and received signals heavily relies on channel state information (CSI), and the
goal of channel estimation is to estimate CSI at the receiver. Accurate channel estimation
helps in compensating for the amplitude and phase of the received signal, facilitating the
precise recovery of the original data when demodulated at the receiver. For these reasons,
channel estimation is a key part of physical layer wireless communication. In light of
this, enhancing its performance can significantly improve the quality of communication
networks, help overcome the performance limitations of 5G, and pave the way for achieving
the “Intelligent Interconnection of Everything” goal of 6G [4,5].
Conventional channel estimation methods are typically classified as pilot-based, blind,
and semi-blind. Due to the simplicity of the principle behind the pilot-based channel esti-
mation method, it can be implemented relatively quickly in practice. However, this method
requires transmitting both data and pilot symbols, which leads to low spectrum utilization
and reduced transmission efficiency. While the blind channel estimation method does not
require pilots, it is computationally complex, sluggish to converge, and demands many
statistical samples. The semi-blind channel estimation method represents a compromise
approach that integrates the benefits of pilot-based and blind channel estimation meth-
ods. Nevertheless, it is worth noting that the implementation complexity of this method
remains substantial.
It is evident that, with their significant limitations, conventional channel estimation
methods are inadequate for addressing the growing complexity of future communication
systems. Hence, there is a pressing need to develop a novel approach that exhibits en-
hanced precision, efficacy, and intelligence in estimating CSI. The ongoing advancement of
artificial intelligence in recent years has sparked a fresh surge of technological innovation in
human society. As a significant research avenue in artificial intelligence, deep learning has
successfully addressed a range of previously challenging issues. It has exhibited excellent
performance across numerous areas, including speech recognition, natural language pro-
cessing, and computer vision [6]. In wireless communications, applying deep learning to
the physical layer has become a hot research topic, and papers [7,8] summarize the possible
research directions. In this paper, the focus is on exploring the application of deep learning
to channel estimation.
The primary purpose of using deep learning for channel estimation is to enhance the
receiver’s performance, thereby optimizing the recovery process for impaired transmit-
ted data. Compared to traditional methods, deep-learning-based methods have greater
advantages in terms of estimation accuracy or computation overhead in complex channel
environments. After thoroughly reviewing the relevant literature, we observe that deep-
learning-based channel estimation methods can fall into two distinct classes: data-driven
and model-driven.
The data-driven approach abandons the modular structure of wireless communication
systems and replaces it with neural networks trained on massive data. In 2017, the authors
of [9] started to utilize a data-driven deep learning approach for implicit channel estimation.
In this procedure, the orthogonal frequency division multiplexing (OFDM) receiver is
regarded as a “black box”, and the channel estimation and signal detection modules
are substituted by a deep neural network (DNN). The DNN is trained offline to obtain
the best parameters and then deployed online to directly recover the transmitted data.
For massive multiple-input–multiple-output (MIMO) systems, the data-driven approach
proposed in [10] employs a multilayer perceptron (MLP). This method does not leverage
channel statistics and can achieve considerable performance gains over traditional channel
estimators in low signal-to-noise ratio (SNR) regions.
The model-driven approach is based on known physical layer prior knowledge, de-
creasing the need for data [11]. The ComNet developed in [12] is a typical example of
model-driven schemes. Unlike the data-driven method in [9], ComNet retains the basic
OFDM system framework. In ComNet’s channel estimation module, a DNN learns the
error between the least square (LS) estimation and the actual channel response, resulting in
a more accurate channel estimation. In ComNet’s signal detection module, the network
Electronics 2023, 12, 4965 3 of 38

with nonlinear activation functions introduces nonlinearity, making the module a nonlinear
signal detector that can more effectively process and recognize nonlinear features and dis-
tortions within the signal. In the model-driven scheme proposed in [13] for millimeter-wave
massive MIMO systems, neural networks are fused into an iterative signal reconstruction
algorithm to estimate channels. This scheme improves the dilemma of difficult CSI ac-
quisition in scenarios where antenna arrays are dense, and receivers are equipped with
restricted radio frequency (RF) chains.
Numerous studies have shown that deep-learning-based methods can outperform
traditional methods, and they hold the potential to perform well in future wireless commu-
nications. In recent years, some scholars [14–18] have reviewed the research on channel es-
timation based on deep learning. The main contributions of these reviews are summarized
in Table 1. While these previous reviews offer valuable assistance on deep-learning-based
channel estimation, their discussion is limited to traditional communication scenarios.
However, it is crucial to recognize that the emergence of techniques such as the recon-
figurable intelligent surface (RIS) [19] has spawned novel communication scenarios that
present novel challenges and opportunities. Therefore, to better assist future research,
it is also necessary to review state-of-the-art deep-learning-based methods for emerging
scenarios, such as scenarios incorporating the RIS technique.

Table 1. Summary of recent review papers on deep-learning-based channel estimation.

Article Year Main Contribution

It groups the challenges (such as feedback overhead) in massive MIMO channel modeling and
[14] 2019 estimation according to four technical details, followed by a discussion of each group’s relevant
deep-learning-based solutions.
It outlines how to use deep learning to enhance the performance of massive MIMO channel
[15] 2020
estimation while reducing training overhead, followed by introducing some data-driven methods.
It reviews various deep learning models used for channel estimation. Subsequently, it introduces
[16] 2022
the channel estimation methods that use deep learning in different systems.
It overviews recent channel estimation methods using deep learning in doubly-dispersive channels
[17] 2022 and conducts experimental comparisons and simulation analyses under various frame sizes,
modulation orders, and mobility scenarios.
It provides a comprehensive summary of artificial intelligence-assisted channel estimation methods
[18] 2022
in multicarrier systems, which include machine learning and neural networks.
This paper discusses data-driven channel estimation methods based on different fundamental
neural networks. It also introduces model-driven methods, which are categorized according to
This review 2023
various benchmark algorithms. Additionally, this paper overviews recent studies on
deep-learning-based channel estimation in RIS-aided communication systems.

RIS, also referred to as intelligent reflection surface (IRS), is a promising 6G com-

munication technique. It can achieve multiple purposes, including capacity expansion,
coverage expansion, interference suppression, high beamforming gain, and improving
the wireless propagation environment [20]. Nevertheless, RIS is an approximately passive
device that cannot actively transmit and receive signals or further process them. This
limitation, coupled with the introduction of cascaded channels with high dimensionality,
presents a challenge in channel estimation. As a result, conventional channel estimation
methods are computationally expensive and inaccurate for practical RIS-aided communi-
cation systems. In this context, deep learning is particularly apt for addressing channel
estimation challenges in RIS-aided communication systems, owing to its ability to effec-
tively solve nonlinear mapping problems. Specifically, this approach achieves its goal by
learning an approximate mapping function from the training data to the CSI of a single or
cascaded channel.
Electronics 2023, 12, 4965 4 of 38

Given the ongoing advancements in wireless communication technologies and the

vital role played by deep learning in solving channel estimation problems, we provide a
review of deep-learning-based channel estimation. The main contributions of this paper
are as follows:
• Before discussing deep-learning-based channel estimation methods, conventional
methods are introduced and classified as pilot-based, blind, and semi-blind, followed
by an analysis of the advantages and disadvantages of each category.
• Dividing deep-learning-based channel estimation methods into data-driven and
model-driven approaches, then reviewing their recent studies based on neural network
types and benchmark algorithm types, respectively.
• We extend the discussion on applying deep learning to channel estimation to RIS-
aided communication systems, which are emerging scenarios in next-generation
wireless communications.
The remainder of this paper is structured as follows: Section 2 introduces the three
main classes of conventional channel estimation methods and analyzes their advantages
and disadvantages; Section 3 presents several commonly used neural networks and re-
views recent research on deep-learning-based channel estimation; Section 4 discusses the
application of deep learning in channel estimation for RIS-aided communication systems;
Section 5 describes the challenges and future research trends of deep-learning-based chan-
nel estimation; and finally, Section 6 summarizes the entire paper. Figure 1 illustrates the
organization of this paper.

Ⅱ. Conventional channel Introducing the conventional channel estimation

estimation methods methods and analyzing their respective
characteristics

Ⅲ. Deep learning-based
Summarizing this paper Ⅵ. Conclusion The shortcomings of conventional
channel estimation
at the end methods algorithms make them unsuitable for
modern communication, so attempting to
introduce deep learning to solve the
This review channel estimation issue

Ⅳ. Channel estimation
Proposing challenges and future Ⅴ. Challenges and future based on deep learning Extending the application of deep
research trends based on existing research trends for RIS-aided learning in channel estimation to
research communication systems RIS-aided communication systems

Figure 1. Organization of this paper.

2. Conventional Channel Estimation Methods

Irrespective of the specific channel model employed in any wireless communication
system, a shared challenge arises: how to provide accurate and rapid channel estimation
to facilitate precise signal demodulation, equalization, decoding, and other baseband
processing tasks [21]. As delineated in Section 1, the conventional channel estimation
methods already developed for wireless communications mainly include pilot-based, blind,
and semi-blind methods. Figure 2 shows the classification of different conventional channel
estimation methods.
Electronics 2023, 12, 4965 5 of 38

Least Square

Pilot-Based Channel Minimum Mean

Estimation Square Error

Linear
Minimum Mean
Square Error

Subspace

Conventional Channel
Blind Channel Higher Order
Estimation
Estimation Statistics
Methods

Maximum
Likehood

Subspace
Semi-Blind Channel
Estimation
Expectation
Maximization

Figure 2. Classification of conventional channel estimation methods.

2.1. Pilot-Based Channel Estimation Method

In wireless communications, the pilot-based method is the most used channel esti-
mation method, with its process illustrated in Figure 3. The central idea of this method is
inserting known pilots into the transmitted signal according to specific rules. After that,
the corresponding received signal is extracted at pilot positions at the receiver. Following
this, the CSI at pilot positions is computed by combining them with the known transmit-
ted pilots. Finally, various interpolation algorithms are employed to estimate the CSI at
non-pilot positions (data positions) [22–25]. Among them, the LS algorithm [26,27], the
minimum mean square error (MMSE) algorithm [28,29], and the linear MMSE (LMMSE)
algorithm [30,31] are the most classical pilot-based channel estimation methods. While
we will introduce these pilot-based estimation methods using the OFDM system as an
example, it is essential to note that these methods are not limited to OFDM systems.

Pilot Pilot CSI at Pilot

Interpolation
Insertion Extraction Positions

Transmitted Result of Channel

Signal Estimation

Figure 3. Flowchart of pilot-based channel estimation.

2.1.1. Pilot Arrangements

According to the different ways of pilot insertion, the pilot arrangements can be
roughly categorized into block-type, comb-type, and lattice-type.
• Block-Type Pilot Arrangement
Figure 4 shows the block-type pilot arrangement. This pilot arrangement periodically
transmits pilot symbols (i.e., OFDM symbols with pilots inserted) along the frequency axis.
For each pilot symbol, pilots are inserted into all subcarriers (as pilot subcarriers). Then,
Electronics 2023, 12, 4965 6 of 38

these pilots are used to interpolate in the time domain and estimate the channel along the
time axis. To track the characteristics of the time-selective channel, the period of the pilot
symbols must be equal to the channel coherence time. Let St denote the period of the pilot
symbols, and since the channel coherence time is in inverse proportion to the Doppler
frequency deviation f Doppler , St must be satisfied:

1
St ≤ (1)
f Doppler

Since the pilots are periodically inserted into all subcarriers of the pilot symbols, the
block-type pilot arrangement is more suitable for frequency-selective channels compared
to time-selective channels.

an OFDM symbol St

pilot subcarrier

data subcarrier
Frequency

Time

Figure 4. Block-type pilot arrangement.

• Comb-Type Pilot Arrangement

Figure 5 shows the comb-type pilot arrangement. The pilots are regularly inserted into
the subcarriers within each OFDM symbol, utilized to interpolate in the frequency domain
and estimate the channel along the frequency axis. To accurately track the frequency-
selective channel’s properties, the pilots’ period on the frequency axis must be consistent
with the coherent broadband. The period of the pilots on the frequency axis is denoted
by S f . Since the coherence bandwidth depends on the reciprocal of the maximum delay
extension σmax , S f must be satisfied:

1
Sf ≤ (2)
σmax

Unlike the block-type pilot arrangement, the comb-type pilot arrangement is more
suitable for time-selective channels compared to frequency-selective channels.

an OFDM symbol

pilot subcarrier

Sf
data subcarrier
Frequency

Time

Figure 5. Comb-type pilot arrangement.

Electronics 2023, 12, 4965 7 of 38

• Lattice-Type Pilot Arrangement

Figure 6 shows the lattice-type pilot arrangement. The pilots are inserted at a specified
period and concurrently into both the time and frequency domains. With the regular
distribution along the time and frequency axes, they facilitate the interpolation of channel
estimation in the two domains. The insertion of the pilots must satisfy Equations (1) and (2)
to trace the channel’s time-selective and frequency-selective characteristics. This pilot
arrangement is a compromise between the block-type and comb-type pilot arrangements.
It has a superior performance against time-selective fading compared to the block-type
pilot arrangement and performance against frequency-selective fading compared to the
comb-type pilot arrangement.

St an OFDM symbol

pilot subcarrier
Sf
data subcarrier
Frequency

Time

Figure 6. Lattice-type pilot arrangement.

2.1.2. LS Channel Estimation

The criterion of the LS algorithm is to minimize the weighted error between the
estimated value and the true one. Considering a simple OFDM system, let Np denote the

T
number of pilots inserted, y = y(0), y(1), . . . , y Np − 1 denote
 the frequency domain

received signal at the pilot position, and X = diag{x} = diag x (0), x (1), . . . , x Np − 1
denote the corresponding transmitted signal at the pilot position. Furthermore, assume

T
w = w (0), w (1), . . . , w Np − 1 represents the noise, and the cost function of the LS
estimation result ĥLS can be mathematically represented as:
  2
J ĥLS = y − XĥLS (3)

The LS algorithm aims to estimate the channel by minimizing the cost function, and
this is achieved by setting the partial derivative to zero, as follows:
 H  
  
∂ J ĥLS ∂ y − X ĥ LS y − X ĥ LS
= =0 (4)
∂ĥLS ∂ĥLS

thus, XH XĥLS = XH y can be obtained to derive the LS estimation of pilot subcarriers:

"
#T
  −1 y (0) y (1) y Np − 1
ĥLS = XH X X H y = X −1 y = , ,...,
(5)
x (0) x (1) x Np − 1

The LS algorithm is widely adopted because of its low computational requirement

and simple implementation. Nevertheless, due to the omission of channel and noise
factors in the computation process, the estimation performance is vulnerable to inter-carrier
interference and noise, thereby limiting the algorithm’s precision.
Electronics 2023, 12, 4965 8 of 38

2.1.3. MMSE Channel Estimation

The purpose of the MMSE algorithm is to minimize the mean square error between
the actual value and the estimated value, and the cost function of the MMSE estimation
result ĥMMSE can be represented as:
    H 
J ĥMMSE = E h − ĥMMSE h − ĥMMSE (6)

 
in a similar vein, setting the partial derivative of J ĥMMSE to zero, it can be concluded:

ĥMMSE = Rhy R− 1
yy y (7)

where Rhy represents the cross-correlation matrix between the actual channel response h
and the received signal y, and Ryy is the autocorrelation matrix of y. Rhy can be further
expressed as: n o n o
Rhy = E hyH = E h(Xh + w)H = Rhh XH (8)

also, Ryy can be further expressed as:

n o n o
Ryy = E yyH = E (Xh + w)(Xh + w)H = XRhh XH + σ2 I (9)

where Rhh is the autocorrelation matrix of the actual channel response and σ2 is the noise
variance. Therefore, it can be concluded:
   −1  −1
2 H
ĥMMSE = Rhh Rhh + σ XX ĥLS (10)

Since the MMSE algorithm considers the noise impact and introduces the priori infor-
mation Rhh and σ2 , its channel estimation accuracy is high. Under the same MSE accuracy
condition, at least 10 dB of SNR gain can be obtained compared to the LS algorithm.
However, the practical implementation of the MMSE algorithm is constrained due to chal-
lenges in acquiring prior knowledge and second-order statistical features in communication
systems. Moreover, the need for multiple matrix inversions significantly amplifies the
algorithm’s complexity.

2.1.4. LMMSE Channel Estimation

As expressed in Equation (10), the inverse of the matrix needs to be recalculated when
the pilot signal X changes, making the computation more difficult. The LMMSE algorithm

−1
isna simplified version of the MMSE algorithm by replacing nXXHo with its expectation

−1 o
E XXH . Furthermore, the SNR is defined as SNR = E | x |2 /σ2 , so that there is:

  −1
β
ĥLMMSE = Rhh Rhh + I ĥLS (11)
SNR
n o n o
where β = E | x |2 E 1/| x |2 is the modulation type parameter. For instance, β = 1
corresponds to QPSK modulation, and β = 17/9 corresponds to 16QAM modulation. If
both Rhh and SNR are known, only one matrix inverse operation is performed, effectively
reducing computational complexity.

2.2. Blind Channel Estimation Method

Unlike the pilot-based channel estimation, the blind channel estimation does not
require any pilots. The core concept of this method is to interpret the CSI. This is achieved
by leveraging the inherent statistical characteristics of both received and transmitted signals
Electronics 2023, 12, 4965 9 of 38

while adaptively tracking the dynamic changes of channels [32]. In general, blind channel
estimation may be performed using the subspace algorithm, the higher-order statistics
(HOS) algorithm, and the maximum likelihood (ML) algorithm.

2.2.1. Subspace-Based Blind Channel Estimation

The subspace-based blind channel estimation does not require any known training
sequence, making it useful for systems where the training sequence cannot be sent or takes
up a significant amount of bandwidth. The core concept of this algorithm is to use the
statistical characteristics of the signal to separate the signal and noise subspaces and then
estimate the channel parameters [33]. In [34], Kang et al. proposed a generalized subspace
algorithm via a minimized cost function, which focuses on integrating the kernel matrix
set from signals that share the target channel through a weighted sum of projection errors.
Experiments demonstrate that the proposed algorithm has exceptional performance, which
can be further enhanced by employing a particular weight set for the cost function and
adding more kernel matrices. Most subspace-based blind algorithms primarily estimate
channel coefficients by utilizing the orthogonality of the correlation matrix of the signal
and the noise subspaces of the received signal. However, in practice, the correlation matrix
is unknown and necessitates time-averaging estimation using multiple received samples.
For this reason, Tu et al. [35] proposed a scheme that leverages the frequency correlation
between neighboring subcarriers in MIMO-OFDM systems to minimize the time averaging.

2.2.2. HOS-Based Blind Channel Estimation

The utilization of HOS information on the signal can yield additional insights into
both the signal and noise details, which has the potential to enhance the channel estimation
accuracy. The HOS-based blind channel estimation mainly estimates the CSI through
the HOS characteristics of received signals. Acar et al. [36], faced with channels with
constant scaling phases and identical alignments, proposed to compute channel estimation
values in a recursive scheme using the parallel factor (PARAFAC) decomposition and
the HOS tensor of the system output. The proposed method eliminates pre-whitening
and has a lower estimation error than conventional methods. In [37], Choqueuse et al.
investigated a blind estimation algorithm for space-time block coding (STBC) systems based
on HOS. Leveraging the statistical independence of sources before space–time encoding,
this algorithm estimates the channel matrix by minimizing a kurtosis-based cost function
following zero-forcing (ZF) equalization. In simulations, the proposed algorithm exhibits
superior generalization ability over other methods, and it applies to the whole linear
STBC class.

2.2.3. ML-Based Blind Channel Estimation

Some researchers have employed the ML algorithm to perform blind channel estima-
tion as well. The ML-based blind algorithm constructs a likelihood function to measure
the similarity between the predicted and actual received signals under the given channel
parameters. By optimizing this likelihood function, one can obtain the best channel estima-
tion result. In [38], Cirpan et al. investigated the deterministic ML (DML) and stochastic
ML (SML) iterative methods for jointly estimating Doppler frequency shift and channel
parameters. These methods offer maximum likelihood solutions and thus perform better
than most HOS-based and blind adaptive methods, particularly if only a short data record
is available. In [39], Necker et al. developed the first fast converging, fully blind channel
estimation algorithm based on the ML principle. This algorithm recovers the absolute phase
of the channel transfer function without reference symbols. Compared to the pilot-based
method, there is a higher spectrum utilization and a lower mean square error (MSE) and
bit error rate (BER).
The blind channel estimation method can effectively utilize spectrum resources and
adaptively track the changes in CSI, without requiring extra pilots. Nevertheless, the
blind method is not without its limitations. These include the method’s high computing
Electronics 2023, 12, 4965 10 of 38

cost, relatively slow convergence speed, extra observation data, and the presence of phase
ambiguity [33,35]. Thus, these challenges pose difficulties for the practical use of such
a method.

2.3. Semi-Blind Channel Estimation Method

The semi-blind channel estimation method combines the computational simplicity
of the pilot-based method with the increased spectrum utilization of the blind method. It
utilizes a few pilots and integrates the received signal statistics from the blind channel
estimation method, intending to reduce computational complexity and ensure accurate
channel estimation. Among them, researchers commonly use the subspace and expectation
maximization (EM) algorithms for semi-blind channel estimation.

2.3.1. Subspace-Based Semi-Blind Channel Estimation

Similar to the subspace-based blind algorithm, the semi-blind subspace-based algo-
rithm also leverages the statistical characteristics of the signal. It aims to find the channel
parameters that best match the signal or noise subspace to achieve channel estimation. In
the scheme presented by Wu et al. [40], the received signals are initially projected into the
subspace with minimal interference. Following this, a low-complexity modified power
method is employed to recursively establish the subspace base. Finally, an initial estimation
of the projected channel is performed using a small number of pilots. Through experiments,
the scheme demonstrates its ability to effectively reduce the impact of pilot contamina-
tion. In [41], Kawasaki et al. utilized out-of-band emission to suppress redundancy in
their subspace-based scheme. Subsequently, they employed the LS estimation on pilot
subcarriers to solve the scalar ambiguity. Through experiments, their scheme can perform
channel estimation for the remaining subcarriers with a high accuracy equal to that of the
pilot subcarriers.

2.3.2. EM-Based Semi-Blind Channel Estimation

As an iterative optimization algorithm, the EM algorithm [42] works to estimate pa-
rameters in probabilistic models with hidden variables. In semi-blind channel estimation,
this algorithm is executed in the expectation step (E-step) and the maximization step (M-
step). In the E-step, the algorithm fixes the current channel parameters and calculates the
expectation of the correlation between the training sequence and the received signal. In
the subsequent M-step, the algorithm updates the parameters to maximize the probability
of consistency between predicted and actual data. Then, iterate these two steps until the
parameters converge and achieve the channel estimation. In [43], Obradovic et al. pro-
posed an EM-based temporal semi-blind method for MIMO-OFDM time-varying channels.
Their method solves the problem of inadequate pilot sequence arrangement in standard
pilot-aided channel estimation, which hinders the direct application of interpolation ap-
proaches. Furthermore, the method can handle systems that do not satisfy the sampling
theorem. Specifically, in cases where the sampling theorem is not met, the method can aptly
initialize the EM algorithm by performing local Wiener interpolation on the frequency axis.
In [44], Nayebi et al. proposed two EM-based semi-blind algorithms (one introduces prior
information while the other does not) for time division duplexing (TDD) massive MIMO
systems. Experimentally, it is demonstrated that the EM algorithm, with the introduction
of prior information, can perform better under low SNR conditions.
The semi-blind channel estimation adds a few pilots, building on the foundation of
the blind channel estimation. Compared to blind channel estimation, it can achieve faster
convergence, lower computational complexity, and address phase ambiguity. In addition,
it boasts a higher bandwidth utilization than the pilot-based method. However, the semi-
blind channel estimation still requires substantial computation, making it relatively difficult
to realize.
Electronics 2023, 12, 4965 11 of 38

Based on the introduction to conventional channel estimation methods, Table 2 sum-

marizes their advantages and disadvantages.

Table 2. Summary of advantages and disadvantages of conventional channel estimation methods.

Methods Advantages Disadvantages

Low computational complexity Waste of spectrum resources
Pilot-based channel estimation Easy to implement Low transmission efficiency
Pilot can improve estimation performance
Efficient use of spectrum resources High computational complexity
Able to adaptively track the dynamic changes of the channel Slow convergence speed
Blind channel estimation
Presence of phase ambiguity
Long observation time
Relatively high bandwidth efficiency Relatively large amount of calculation
Semi-blind channel estimation Able to solve phase ambiguity Relatively difficult to implement
Relatively fast convergence speed

From Table 2, we can conclude that conventional channel estimation methods have
limitations that will restrict their application in increasingly complex future communication
systems. Therefore, researchers in related fields have set out to figure out how to improve
the performance of channel estimation while simultaneously reducing its computational
complexity. Deep learning has made a big splash in physical layer wireless communications
in recent years, opening up a new horizon for scholars. Numerous studies have introduced
deep learning as a novel approach to channel estimation and have obtained outstanding
results. Section 3 highlights the application of deep learning in channel estimation.

3. Deep-Learning-Based Channel Estimation Methods

Deep learning is transforming the paradigms and methods of physical layer commu-
nications, with application areas including channel estimation, resource allocation, and
signal detection [45,46]. Conventional channel estimation methods work well in systems
with a clear and linear representation of the input–output relationship. However, when the
wireless environment becomes more complicated, the input–output relationship exhibits
nonlinear characteristics, potentially reducing the effectiveness of conventional techniques.
Deep learning aims to enable machines to learn to deal with complex and highly nonlinear
relationships between input datasets and desired outputs without human intervention [47].
As stated in Section 1, the existing research on deep-learning-based channel estimation can
be broadly classified into two main categories: data-driven and model-driven approaches.
Therefore, this section will describe deep-learning-based channel estimation according to
these two approaches and introduce their representative studies.

3.1. Overview of Neural Networks

As a branch of artificial intelligence technology, deep learning uses neural networks
as models, so it is necessary to briefly introduce neural networks before introducing deep-
learning-based channel estimation methods. Neural networks, which mimic the neuronal
network of the human brain, consist of input, hidden, and output layers with numerous
interconnected neurons through the weights in each layer. By adjusting these weights
through learning, neural networks can process input information and produce output.
According to the application of deep learning in various fields, the structure of the basic
network is different, primarily including the deep neural network (DNN), convolutional
neural network (CNN), recurrent neural network (RNN), generative adversarial network
(GAN), etc. Most deep learning models can be combined or optimized based on these four
networks. Next, we will introduce these common neural networks.

3.1.1. DNN
A basic structure of DNN is shown in Figure 7, which consists of one input layer, L
hidden layers, and one output layer, with several neurons in each layer and full connectivity
Electronics 2023, 12, 4965 12 of 38

between layers. A DNN can be considered a neural network that contains numerous hidden
layers, and the purpose of adding multiple hidden layers is to enhance the learning and
mapping capabilities of the network. Additionally, to introduce the nonlinearity, activation
functions like the sigmoid, rectified linear unit (ReLU), or hyperbolic tangent (Tanh) are
applied after the outputs of each layer.

Hidden Hidden Hidden Hidden

Layer 1 Layer 2 Layer L-1 Layer L

···
Input Layer
··· Output Layer

···
···
···
···
···
···
Figure 7. Basic structure of DNN.

3.1.2. CNN
CNN is among the most representative model structures for deep learning, with the
earliest proposition dating back to the publication of the seminal literature [48]. Figure 8
shows a basic CNN structure comprising an input layer, multiple convolutional and pooling
layers, a fully connected layer, and an output layer. The input layer’s primary function is to
preprocess the data. After this, the convolutional layer extracts features from the processed
data, and its output is then mapped nonlinearly via the activation function. Following
this, the pooling layer is introduced to decrease computation and prevent overfitting. The
convolutional and pooling layers are alternately stacked, and after a series of operations,
the output is obtained through the fully connected and output layers. In the convolutional
layer, the convolutional kernel is locally connected to its input feature map. For each
position in the output feature map, the value is obtained by the weighted sum of the
local inputs and connection weights, plus the bias. As this process is equivalent to the
convolutional operation, the network is termed a convolutional neural network.

Convolutional
Layer Pooling Fully Connected
Activation Layer Output
Input Layer Function Layer Layer

···

Figure 8. Basic structure of CNN.

3.1.3. RNN
Figure 9 shows a basic RNN structure. In contrast to the previously discussed DNN
and CNN, RNN is uniquely designed to handle sequential data. It can remember infor-
mation from past moments and utilize it for the current output calculation. In this case,
the nodes among the hidden layers are interconnected. Moreover, the hidden layer’s
input depends on both the input layer’s value and the hidden layer’s output from the
previous moment.
Electronics 2023, 12, 4965 13 of 38

Output
O Ot-1 Ot Ot+1
Layer

V V V V

Unfold W W W W
Hidden W
S St-1 St St+1
Layer

U U U U

Input
X Xt-1 Xt Xt+1
Layer

Figure 9. Basic structure of RNN.

During the training process, RNNs may suffer from long-term dependency, leading to
gradient vanishing or gradient explosion. Therefore, the literature [49] conducted further
research and proposed a special RNN called the long short-term memory (LSTM) network
to effectively resolve this issue.

3.1.4. GAN
Figure 10 represents the general GAN structure, which consists of a generator and a
discriminator. The generator receives a latent vector (usually random noise) as input and
generates samples analogous to the training data. On the other hand, the discriminator
receives samples (which can be real or generated by the generator) as input and predicts
their veracity. Notably, the generator and the discriminator train against each other in a
competitive and collaborative dynamic. Specifically, the generator’s objective is to trick the
discriminator by bringing the generated samples closer and closer to the real samples to
the point where they cannot be distinguished accurately. Concurrently, the discriminator
is tasked with accurately classifying the samples as possible, thereby heightening the
distinction between the real and generated samples. During the iterative adversarial
training process, the generator and the discriminator keep adjusting their parameters until
they reach an equilibrium point. That is to say, the generator can generate realistic samples,
while the discriminator cannot distinguish between real and generated samples.

Random
Generator
Noises

Generated
Ture Samples Discriminator
Samples

Fake/Real
0/1

Figure 10. Basic structure of GAN.

Sometimes, GAN is “too free”, and thus, we cannot control the generator to generate
the required data. For this reason, researchers have proposed a variant of GAN, conditional
GAN (CGAN) [50], where conditions are introduced into the modeling of both the generator
and the discriminator to guide the data generation process.
Electronics 2023, 12, 4965 14 of 38

3.2. Toy Example of Deep Learning Application in Channel Estimation

Before reviewing the research on deep-learning-based channel estimation, we pro-
vide a toy example to illustrate the application of deep learning in channel estima-
tion. The example aims to demonstrate the ability of deep learning to enhance channel
estimation performance.
This simulation is set in a vehicular communication scenario within an OFDM system
that conforms to the IEEE 802.11p standard. In this environment, the channel exhibits
time-varying characteristics, and there is a strong correlation between adjacent OFDM
symbols. The classic LS algorithm, performed based on two preambles in an OFDM frame,
struggles to track rapid changes in the channel, resulting in poor performance. In contrast,
the constructed data pilot (CDP) algorithm constructs virtual pilots from data subcarriers
and uses the correlation between adjacent OFDM symbols. It treats the data subcarriers of
the previous symbol as a preamble to conduct the current symbol’s channel estimation. This
algorithm somewhat improves the channel estimation performance but reduces reliability
due to error propagation from one symbol to the next.
The flowchart of our example is shown in Figure 11. In our example, we aim to
employ a DNN to capture more channel time–frequency features, thereby reducing the
error between the CDP estimation results and the actual channel response and improving
the channel estimation accuracy. This DNN comprises three hidden layers, each with
20 neurons, and an output layer, and it takes the coarse estimation result from the CDP
algorithm as its input.

Equalizaiton
Constructing LS CDP
Equalization and
Data Pilot Estimation Decision
Demapping

Received OFDM Coarse Estimation More Accurate

Symbols Result Estimation Result
CDP Algorithm DNN
Hidden Layer 1

Hidden Layer 2

Hidden Layer 3

Output Layer

Figure 11. Flowchart of the toy example.

In the experiments, we used the VTV Urban Canyon (VTV-UC) channel model with
a velocity setting of 48 km/h and a Doppler shift of 500 Hz. To obtain a robust DNN
model, we trained it at a high SNR of 40 dB. Then, we tested it in a SNR range of 0–40 dB
and compared its performance with the LS algorithm and the original CDP algorithm.
Figure 12a,b display the performance comparison in terms of BER and NMSE, respectively.
These results indicate that the DNN-enhanced CDP approach significantly improves the
channel estimation accuracy in most scenarios. Even though the performance is not the
best in a few low-SNR conditions, it still outperforms the other two traditional algorithms
in general, showing the potential of deep learning in channel estimation applications.
Electronics 2023, 12, 4965 15 of 38

(a) (b)

Figure 12. Performance comparison with the LS algorithm and the original CDP algorithm; (a) BER
performance comparison; and (b) NMSE performance comparison.

3.3. Data-Driven Deep-Learning-Based Channel Estimation Methods

Figure 13 illustrates the structure of a simple channel estimator based on data-driven
deep learning. The core idea of the data-driven approach is to consider neural networks as
“black boxes”, replacing the original traditional communication system structure with them
and perpetually updating their parameters through training with massive amounts of data.
Next, we will summarize the data-driven deep learning channel estimation methods by
categorizing them according to different kinds of neural networks.

Channel
Estimator

Channel
DL Model Channel Data Data-Driven
Estimation
Selection Acquisition Learning
Result

Data-Driven
Learning

Traning Data Neural Network

Figure 13. Structure of a data-driven deep-learning-based channel estimator.

3.3.1. Application of DNN in Data-Driven Channel Estimation

DNN is a fundamental network for deep learning with a simple model, and as such, it
was first applied in channel estimation. It allows end-to-end learning, i.e., from the original
input to the target output. Figure 14 shows a DNN-based channel estimator proposed
in [9], serving as one of the typical examples. In this model, the authors estimated the CSI
end-to-end by recovering the transmitted data from the received data, including the pilot
and data blocks. Moreover, DNN is highly adaptable and flexible, and its structure and
hyperparameters can be modified to accommodate various application scenarios.
Electronics 2023, 12, 4965 16 of 38

Transmitter Channel
X(k) Transmit Insert x(n) w(n)
S/P IDFT P/S h(n)
symbols CP
y(n)

Receiver
Xˆ (k ) Recover Remove
P/S DFT S/P
Data CP
Online deployment
Offline training
Transmit
Symbols

Simulation
DNN
Data

Channel
Models

Figure 14. DNN-based end-to-end channel estimator [9].

In 2020, Ma et al. [51] investigated a DNN-based method that combines channel

estimation and pilot design. The proposed DNN model is composed of a dimensionality
reduction subnetwork and a reconstruction subnetwork. In the first subnetwork, the
fully connected layer is designed to compress the high-dimensional channel vector into
low-dimensional received measurements, and this compression process treats the weights
in the layer as pilot signals. In the second subnetwork, several cascaded convolutional
layers and a fully connected layer are designed to recover the high-dimensional channel.
Through experimental comparison, this method exhibits a superior NMSE performance
to the simultaneous orthogonal matching pursuit (SOMP) algorithm. Furthermore, when
the DNN trained with multi-carrier samples is tested on single-carrier samples, and it still
performs well, demonstrating its excellent generalization ability.
The interpolation approach is closely connected with the performance of the pilot-
based method. In [52], Ge et al. proposed to utilize DNN to replace the interpolation
process of the conventional pilot-based method. Firstly, the position index and the channel
estimation result of pilot subcarriers are used as training data to train the DNN model.
Then, the position index of data subcarriers is used as the input to the DNN, and the
channel estimation of data subcarriers is finally obtained. This method does not require
prior statistics or matrix inverse operations like the MMSE algorithm and thus has a lower
complexity. Moreover, the proposed method exhibited superior estimation accuracy in
experiments compared to the LS algorithm, which requires interpolation. Finally, the author
tested the model trained with eight multipath paths in conditions with 4, 8, and 12 mul-
tipath paths. The BER performance remains almost unchanged, showing the excellent
generalization ability of the proposed model.
In 2021, Zheng et al. [53] developed an online DNN-based channel estimation method
under limited pilot conditions. This method can dynamically learn and compute in real
time, inferring and updating the DNN weights from the pilot symbols received online
without knowing the real channel matrix to adapt to the actual communication environ-
ment. The proposed DNN-based method exhibits the strongest robustness in experiments.
Furthermore, its estimation accuracy approaches the group orthogonal matching pursuit
(GOMP) algorithm and the burst LASSO algorithm while surpassing the MMSE algo-
rithm. Significantly, it surpasses all these methods by orders of magnitude in terms of
computational speed. Additionally, the proposed DNN model, trained at an SNR of 30 dB,
consistently surpasses the MMSE algorithm in terms of NMSE performance when tested
at lower SNR levels. Specifically, the experiments conducted on the 3GPP SCM TR 25.996
Electronics 2023, 12, 4965 17 of 38

and the DeepMIMO channel models can both yield this result, demonstrating the model’s
strong generalization on lower SNR.
In the realm of ocean exploration, underwater acoustic (UWA) communication is piv-
otal. In 2022, Zhang et al. [54] investigated a DNN-based scheme for underwater acoustic
channel estimation in OFDM systems. The DNN utilizes transmitted pilots and received
symbols in this scheme to reconstruct the UWA channel. Encouragingly, the scheme shows
outstanding signal detection performance on the estimated channel. Simulation results
demonstrate its efficacy, reducing the BER by over 40% in comparison to the LS algorithm.
Furthermore, as the pilot number grows, the BER performance approaches that of the
MMSE algorithm.
To improve the estimation performance in impulse noise environments, Li et al. [55]
developed a method based on the denoising autoencoder-DNN (DAE-DNN). The proposed
method consists of three steps: first, utilizing DAE for preprocessing, namely learning the
impaired data and restoring clean received signals under conditions where impulse noise
is present; second, employing the preprocessed data from DAE to train the DNN offline;
and finally, the DNN estimates the CSI online. Experimental results demonstrate that this
method has strong robustness under impulse noise conditions, outperforming the MMSE
algorithm, the orthogonal matching pursuit (OMP) algorithm, and the LS algorithm in
terms of MSE and BER.

3.3.2. Application of CNN in Data-Driven Channel Estimation

CNN is advantageous for channel estimation since it can reduce the computation
of feature parameters and automatically select the appropriate training weights. In ad-
dition, CNN’s excellent feature extraction capability contributes to its excellent image
processing performance. Hence, the channel estimation problem can be transformed into
an image processing problem in the application, enabling it to learn from the training
data and obtain accurate CSI. Figure 15 shows examples of transforming the channel time–
frequency response into a 2D image under the Vehicular-A (VehA) channel model [56].
Specifically, Figure 15a,b represent the channel time–frequency response images under
the perfect channel model without noise and the noisy channel model with an SNR of
22 dB, respectively.

(a) (b)

Figure 15. Two-dimensional image examples of the channel time–frequency response: (a) under
the VehA perfect channel model (without noise); and (b) under the VehA noisy channel model
(SNR = 22 dB).
Electronics 2023, 12, 4965 18 of 38

In 2019, the ChannelNet proposed by Soltani et al. [57] with regard to the channel time–
frequency response as a low-resolution 2D image and the values are only known at pilot
positions. The ChannelNet consists of a super-resolution CNN (SRCNN) and a denoising
CNN (DnCNN), which are cascaded into a two-step channel estimator. In the first step,
the low-resolution image is enhanced into a high-resolution image by the SRCNN. In the
second step, the DnCNN removes the noise effect to obtain a higher-quality image (i.e., an
accurately estimated channel). Through experimental verification, the ChannelNet’s MSE
performance is superior to the ALMMSE algorithm (an approximation to linear MMSE)
but inferior to the MMSE algorithm in low SNR. However, the performance of ChannelNet
trained at the SNR of 22 dB shows a decreasing trend when the SNR exceeds 23 dB, which
indicates that its generalization ability on real noise is not good enough. Notably, this
literature is a pioneer in using image processing methods for channel estimation, providing
a novel idea for many subsequent studies. Inspired by this, Li et al. [58] investigated a
deep residual channel estimation network (ReEsNet) based on residual learning, which can
work in a wide range of scenarios and significantly reduce the complexity while improving
estimation accuracy.
The numerous convolutional layers in the denoising structure (DnCNN) of Channel-
Net [57] lead to a large amount of computation and a long execution time, which greatly
consumes memory. For this reason, Pradhan et al. [59] made improvements and proposed
the channel estimation network (CENet). Compared with ChannelNet, CENet also uses
SRCNN for image resolution enhancement, with the difference that a convolutional blind
denoising network (CBDNet) is used as the denoising structure instead of DnCNN. The
CBDNet reduces the overall number of convolutional layers, thereby reducing the complex-
ity of the proposed CNN model. Through experimental comparison, the CENet is superior
to the ChannelNet but inferior to the ideal MMSE algorithm in terms of MSE. Furthermore,
the author tested the CENet trained with 48 pilots under conditions of different numbers of
pilots. The results showed that CENet maintains high performance even with fewer pilots,
unlike other methods, which exhibited a significant decrease in performance under the
same conditions.
Most research utilizing deep learning for channel estimation focuses on constructing
complex neural networks, which leads to increased storage and computational require-
ments. Li et al. [60] addressed this phenomenon by combining the CNN and Transformer
and proposing a lightweight channel estimation Transformer (LCET). This scheme treats
the channel response matrix as a 2D image, with the channel features extracted using a
lightweight feature extraction CNN (LFEC) and then transmitted to a lightweight-adjusted
Transformer (LAT) for channel estimation. Through experimental comparison, the estima-
tion accuracy of the LCET surpasses the LS algorithm and some neural networks, closely
approximating the performance of LMMSE algorithms in multi-pilot scenarios.
In general, the transceiver of millimeter wave (mmWave) massive MIMO systems
employs a hybrid precoding structure, and thus, the acquisition of CSI in the low SNR
regime poses difficulties. To overcome this challenge, Zhao et al. [61] proposed ResNet-
UNet, a network that combines the residual network (ResNet) and the U-shaped network
(U-Net) for channel estimation. In particular, the proposed ResU-net consists of a denoiser
and an estimator. Firstly, the received noisy pilot signal is converted into an image and
processed by the denoiser. Then, the obtained clean pilot signal image is fed to the estimator
to obtain the estimation result. Through experimental verification, the Res-UNet surpasses
the conventional algorithms and the deep CNN in estimation accuracy and is robust to
noisy environments.
In 2023, Rahman et al. [62] proposed a deep residual convolutional blind denoising
network (ResCBDNet) for massive MIMO visible light communication systems to estimate
more realistic and accurate CSI. The proposed ResCBDNet comprises two subnetworks:
the noise estimation network and the non-blind denoising network, and it transforms the
sparse channel matrix into a 2D image. Initially, the scheme employs a noise estimation
network to enhance the generalization ability of the true noise. It then interactively reduces
Electronics 2023, 12, 4965 19 of 38

the effect of noise in the channel matrix through the adjustment of the noise level mapping.
Subsequently, the noiseless estimated channel is recovered using the non-blind denoising
network. Experiments demonstrate that the ResCBDnet surpasses some existing deep-
learning-based methods in terms of normalized MSE (NMSE) and peak SNR (PSNR).
Moreover, with its exceptional noise generalization capability, ResCBDNet demonstrates
the superior MSE performance to other methods when SNR = 20 dB.

3.3.3. Application of RNN in Data-Driven Channel Estimation

RNN is not as widely used as other neural networks in channel estimation. However,
some researchers have recently delved into RNN-based methods to improve the channel
estimation accuracy. Given the time-dependent properties of RNN, it can be used for time-
varying channels to improve time-series information processing. Moreover, it is especially
well suited for channel estimation in high-speed mobile settings.
For high-speed mobile environments, channel estimation becomes problematic due to
the impact of multipath and Doppler effects, and conventional methods perform poorly
due to the limited number of available pilots. Consequently, Gizzini et al. [63] proposed a
frame-by-frame (FBF) channel estimation method based on bi-directional RNN (Bi-RNN)
for doubly-selective channels. The goal is to perform end-to-end 2D interpolation after
estimating the channel of pilot symbols, thereby obtaining the channel estimation of data
symbols. Simulation results demonstrate that the proposed Bi-RNN is robust and adaptive,
and the estimation accuracy is superior to some existing deep learning schemes. Although
the estimation accuracy is marginally lower than the conventional 2D-LMMSE algorithm,
the complexity is reduced by nearly 106 times, making it a viable alternative.
In 2021, Essai Ali et al. [64] estimated CSI using a variant of LSTM, namely a bi-
directional LSTM (Bi-LSTM) neural network. The proposed network model relies on
pilot assistance, does not require prior knowledge of channel statistical information, and
employs online training with offline deployment. Under the conditions of limited pilots
and uncertain channel prior statistics, this model is robust and has a superior symbol error
rate (SER) performance over the LSTM and the conventional algorithms (LS and MMSE).
The proposed Bi-LSTM can analyze massive data and establish relationships between
features, thus having excellent generalization capabilities and applicability to new datasets.
The gated recurrent unit (GRU) is a lightweight variant with potential developed
from LSTM. In 2022, Essai Ali et al. [65] designed an end-to-end channel estimation
method based on GRU to recover the transmitted signal from the received signal, thereby
estimating the CSI. Through experimental verification, the GRU-based method surpasses
the conventional estimation algorithms (LS and MMSE), DNN, and ReEsNet in terms of SER.
In 2023, Helmy et al. [66] combined LSTM and GRU to develop an LSTM-GRU estimator,
which is designed for channel estimation from the received signal. It consists of three parts:
the LSTM, the GRU, and the smoothing module. In this framework, the LSTM serves
as a minimum absolute filter that denoises the pre-processed received signals, and the
GRU module compensates for the information loss in the LSTM module as a compensator.
Finally, the model’s output is smoothed using batch normalization and convolutional layers.
Through experimental comparison, the proposed LSTM-GRU estimator has a superior
NMSE performance over the CNN and CGAN estimators. Additionally, the LSTM-GRU
trained with a pilot length of 8 is tested in scenarios of varying pilot lengths. The estimation
performance of this model does not degrade significantly in the short pilot length range,
indicating its outstanding generalization ability.

3.3.4. Application of GAN in Data-Driven Channel Estimation

In channel estimation, GAN has the ability to serve by generating additional data to
augment the required training dataset, which is particularly valuable in scenarios without
sufficient information about the actual CSI. Moreover, GAN can adaptively improve its
performance during training, thereby achieving more accurate estimation results through
the adversarial process between the generator and the discriminator.
Electronics 2023, 12, 4965 20 of 38

Motivated by ReEsNet in [58], Zhao et al. [67] developed a super-resolution GAN

(SRGAN) for channel estimation in 2021. However, the estimation results of ReEsNet have
lost high-frequency details and failed to match the desired fidelity at higher resolutions.
In the SRGAN scheme, the channel generated by the generator is closer to the actual
channel distribution. Meanwhile, the discriminator recovers more details, significantly
enhancing the estimation accuracy. Through experimental comparison, the SRGAN’s MSE
performance surpasses both the LS algorithm and the ReEsNet, and it approaches the
MMSE algorithm in high SNR.
For massive MIMO systems, Dong et al. [68] investigated a CGAN channel estimator
in the same year. The proposed CGAN learns how to map from quantized observations
to the actual channel and learns a suitable adaptive loss function for network training,
resulting in a highly robust model and a more realistic estimated channel. Even when the
model is tested on a different channel (i.e., a millimeter wave channel), its performance
remains almost unchanged, demonstrating its exceptional generalization capability. In
2023, Zhang et al. [69] also employed CGAN, but their focus was obtaining accurate CSI
for MIMO-OFDM systems in high-speed railway scenarios. Their scheme is executed in
two steps and regards the pilot signal as a 2D image. In the denoising step that employs
the U-Net framework, the received noisy pilot image is denoised using the Noise2Noise
(N2N) algorithm. In the estimation step, the CGAN completes the channel estimation by
learning the features of the denoised pilot image. Through experimental verification, the
N2N-CGAN has high estimation accuracy in high-speed mobile scenarios, exhibits strong
robustness in noisy environments, and has low computational complexity.
In the pilot-based channel estimation scheme, the estimation efficacy highly depends
on the pilot insertion pattern. To improve the estimation accuracy while reducing the
pilot overhead, Kang et al. [70] proposed a network model called CAGAN as a channel
estimation scheme for joint pilot design. The CAGAN comprises a CGAN and a concrete
autoencoder (concrete AE). First, the concrete AE is used to search for the most informative
location in the time–frequency grid and insert pilots here, thereby optimizing the pilot
design. Next, the optimized pilots are fed into the CGAN for channel estimation. Experi-
mental results show that the CGAN can exhibit outstanding estimation performance with a
limited number of pilots. In order to verify the generalization ability of noise, the CAGANs
trained with an SNR of 15 dB and trained at different SNRs are compared in a 0–30 dB SNR
range. The result demonstrates that, when the SNR is above 9 dB, the performance of these
two models is comparable.
Table 3 summarizes the reviewed data-driven deep-learning-based channel estimation
methods and their performance. This type of method exhibits high performance when deal-
ing with high-dimensional, complex data, and it can also show self-learning capabilities in
capturing channel characteristics. However, these methods also have glaring disadvantages.
They are highly reliant upon an extensive quantity of training data; if the training data
are insufficient or not well regularized, this could lead to overfitting, resulting in a decline
in generalization ability on new data. In addition, because the neural network structure
is often regarded as a “black box”, its internal working mechanism and decision-making
process are difficult to understand and explain, leading to the poor interpretability of the
data-driven approach.
Electronics 2023, 12, 4965 21 of 38

Table 3. Data-driven deep-learning-based channel estimation methods and their performance.

Reference Model Input Output Compared with Performance Complexity

Has better NMSE performance and can 2 ρ


Ma et al. [51] DNN Original channel Estimated channel SOMP algorithm O 2NBS
reduce pilot overhead

DNN Data subcarrier Channel estimation of LS algorithm Has better NMSE and BER performance \
Ge et al. [52] position index data subcarriers
The NMSE performance is close to the
Received signals and GOMP, MMSE, and burst
Zheng et al. [53] DNN Estimated channel GOMP and the burst LASSO algorithms CPU time = 0.338 ms
pilot symbols LASSO algorithms
while exceeding the MMSE algorithm
Received OFDM symbols The BER performance is close to the
DNN Parameters = 267.168 K,
Zhang et al. [54] and transmitted Estimated channel LS and MMSE algorithms MMSE algorithm, over 40% better than CPU time = 2.995 ms
pilot symbols the LS algorithm

LS, OMP, MMSE algorithms, Outperforms other methods in terms of

Li et al. [55] DAE-DNN Noisy received signal Estimated CSI BER and the MSE in impulse \
and DNN noise environments
At SNR below 20 dB, the MSE Floating-point operations per
ChannelNet Estimated channel at Ideal MMSE and ideal performance is superior to the ideal second (FLOPs) = 1533.583
Soltani et al. [57] (SRCNN + DNCNN) pilot positions Estimated whole channel ALMMSE algorithms MMSE algorithm but inferior to the ideal M, parameters = 130.754 K,
ALMMSE algorithm predict time = 4.422 ms
The MSE performance is superior to the FLOPs = 5.875 M,
Estimated channel at LS, MMSE algorithms, and parameters = 23.554 K,
Li et al. [58] ReEsNet Estimated whole channel LS algorithm and the ChannelNet,
pilot positions ChannelNet
comparable to the LMMSE algorithm predict time = 1.737 ms

Ideal MMSE algorithm and The MSE performance is superior to the

CENet Estimated channel at
Pradhan et al. [59] (SRCNN + CBDNet) pilot positions Estimated whole channel ChannelNet but inferior to the ideal \
ChannelNet MMSE algorithm
The MSE and the BER performance are
LS, LMMSE algorithms, superior to the LS algorithm and other FLOPs = 19.316 M,
LCET (LFEC + LAT) Estimated channel at Estimated whole channel deep learning networks and approach parameters = 21.970 K,
Li et al. [60] pilot positions ChannelNet, ReEsNet,
and SRGAN the LMMSE algorithm in predict time = 5.439 ms
multi-pilot scenarios

LS, MMSE algorithms, and Has the best NMSE performance and
Zhao et al. [61] ResNet-UNet Noisy pilot sequences Estimated channel exhibits strong robustness in \
deep CNN noisy environments
The NMSE and the PSNR performance
ResCBDNet Parameters = 536 K,
Rahman et al. [62] Noisy channel matrix Estimated channel DnCNN and FFDNet are significantly superior to the DnCNN inference time = 0.119 s
and the FFDnet
Electronics 2023, 12, 4965 22 of 38

Table 3. Cont.

Reference Model Input Output Compared with Performance Complexity

The BER performance is superior to the
Estimated channel at 2D-LMMSE algorithm, ChannelNet and the Bi-LSTM and is
O 2QKin + 4Q2


Gizzini et al. [63] Bi-RNN Estimated whole channel
pilot symbols ChannelNet, and Bi-LSTM slightly inferior to the
2D-LMMSE algorithm
Prediction matrix of features LS, MMSE algorithms, With limited pilots, the SER performance
Essai Ali et al. [64] Bi-LSTM Transmitted signal sequence extracted from the \
input sequence and LSTM is superior to all three

LS, MMSE algorithms, DNN, With limited pilots, the SER performance
Essai Ali et al. [65] GRU Received signals Transmitted signals \
and ReEsNet is superior to other methods

LSTM-GRU Computation time =

Helmy et al. [66] Received signals Estimated channel CNN and CGAN Has the best NMSE performance 23.34 ms
At low SNR, it provides the best MSE
FLOPs = 48.590 M,
Estimated channel at pilot LS, LMMSE algorithms, and performance; while at high SNR, it is parameters = 24.102 K,
Zhao et al. [67] SRGAN Estimated whole channel
positions ReESNet only slightly inferior to the predict time = 3.464 ms
LMMSE algorithm
Received signals and Computation time =
Dong et al. [68] CGAN Estimated channel U-Net and CNN Has the best NMSE performance 25.88 ms
pilot sequences
In terms of MSE, it is superior to the LS
LS, MMSE algorithms, algorithm and CGAN and only slightly
O N2


Zhang et al. [69] N2N-CGAN Noisy pilot sequences Estimated channel
and CGAN inferior to the MMSE algorithm at
high SNR
Outperforms the LS algorithm in terms
CAGAN (concrete LS, MMSE algorithms, of MSE and, under high SNR conditions, FLOPs = 548 K,
Kang et al. [70] Noisy channel Estimated channel parameters = 275 K
AE + CGAN) and ChannelNet approaches the ideal MMSE algorithm,
surpassing the ChannelNet
Electronics 2023, 12, 4965 23 of 38

3.4. Model-Driven Deep-Learning-Based Channel Estimation Methods

The model-driven approach is built on known mathematical models, and its central
idea is to combine deep learning with conventional algorithms to improve or expand
existing methods. Moreover, it does not depend on extensive labeled data to select the right
standard neural network, thus making deep learning more interpretable and predictable.
Figure 16 illustrates the components of the model-driven approach. Currently, most
research on model-driven channel estimation primarily unfolds on the basis of the LS
algorithm and the algorithms in compressive sensing (CS) that include the OMP algorithm
and the approximate message passing (AMP) algorithm. This section will discuss model-
driven deep-learning-based channel estimation methods based on these three algorithms.

Model-Driven
DL

Model Algorithm Network

Figure 16. Components of the model-driven approach.

3.4.1. Model-Driven Channel Estimation Combining LS Algorithm

As previously mentioned, the conventional LS channel estimation algorithm is com-
putationally simple but leads to poor performance since it ignores the impact of noise. The
ComNet proposed by Gao et al. [12] solves the problem using deep learning networks,
which provides a reference for subsequent research. As shown in Figure 17, the ComNet is
constructed by cascading the channel estimation and the signal detection subnets. In the
channel estimation subnet part, the authors utilized a DNN to refine the coarse estimation
of the traditional LS algorithm. Specifically, the DNN takes the LS coarse estimation as
input and learns from the discrepancies between this initial estimation and the actual
channel. Then, the DNN adjusts its parameters through training to minimize the prediction
error, resulting in a more accurate channel estimation.
yD
QAM
b¢
demodulation
yp hLS h xZF
LS LS ZF ZF
b
xp algorithm refinement algorithm refinement

Channel estimation subnet Signal detection subnet

DNN

h
hLS h
LS

Figure 17. ComNet structure [12].

In 2021, Jiang et al. [71] introduced a dual CNN structure to improve the LS algorithm’s
estimation accuracy with lower complexity than the general CNN-based methods. The
initial channel estimation obtained by the LS algorithm is used as the input for the dual
CNN, which consists of a spatial-frequency CNN (SFCNN) and an angle-delay CNN
Electronics 2023, 12, 4965 24 of 38

(ADCNN). In particular, the SFCNN effectively leverages the channel’s sparsity to process
white noise, while the ADCNN utilizes the channel correlation to reduce interference.
These two CNNs are connected through a discrete Fourier transform (DFT) process. Owing
to the dual CNN combining the advantages of SFCNN and ADCNN, it exhibits superior
NMSE performance to the single-domain CNN in experiments.
In 2023, Haq et al. [72] innovatively implemented channel estimation on a System on
Chip (SoC) based on deep learning. Specifically, the authors proposed a channel estimation
method called DNN-augmented LS (LSDNN) for the preamble-based OFDM physical layer.
This method employs a fully connected feedforward DNN to process the LS estimation results.
Specifically, the DNN uses the initial LS estimation as input, learning from its errors to improve
the initial estimation. It does this by minimizing the cost function, thereby refining the LS
estimation of the previous preamble. Finally, the proposed LSDNN, the LS algorithm, and
the LMMSE algorithm are mapped onto a Zynq multiprocessor SoC (ZMPSoC) platform
for extensive experimental comparison. Experimental results verify that the LSDNN has a
superior estimation accuracy over the LS and the LMMSE algorithms. However, there is still
room for improvement in resource utilization and power consumption.
To overcome the challenge of SNR mismatch in multipath time-varying channels,
Li et al. [73] introduced a cascaded network called NDR-Net for channel estimation. The
NDR-Net consists of a cascade of a noise level estimation subnet (NLE), a DnCNN, and
residual learning, and the channel matrix is regarded as an image in this scheme. Firstly,
the LS algorithm estimates the coarse value of the channel matrix, which is then inputted
into NLE to obtain the noise level estimation. Subsequently, the estimated noise level
and the initial noisy channel matrix image are inputted into DnCNN for noise reduction,
resulting in a pure noisy image. Finally, the noiseless channel matrix image is obtained by
residual learning. Through experimental comparison, the NDR-Net has better estimation
accuracy than conventional methods when the SNR is mismatched. In addition, it applies
to different Doppler shifts.

3.4.2. Model-Driven Channel Estimation Combining OMP Algorithm

As a greedy algorithm, the OMP algorithm offers optimal solutions to sparse recov-
ery problems. Leveraging the sparsity of channels, it facilitates channel estimation with
minimized pilot overhead. However, due to the grid mismatch issue, the OMP algorithm
may not provide satisfactory estimation results [74]. To achieve better channel estimation
performance, combining the OMP algorithm with deep learning is a viable approach.
In 2022, Li et al. [75] designed a model-driven ResNet-based scheme for orthogonal
time–frequency space (OTFS) systems. Considering the sparsity of the channel in the delay-
Doppler domain, they first employed the OMP algorithm to obtain a coarse estimation. Then,
this initial estimation was fed into the proposed ResNet for a more accurate result. During
the training process, the ResNet learns from the error between the OMP rough estimation
and the actual channel response. Then, this network continuously optimizes the parameters,
resulting in a more realistic channel estimation result. Compared to the conventional OMP
algorithm, the ResNet-based scheme has a superior NMSE performance. Furthermore, it
applies to various scenarios and displays outstanding robustness to Doppler spread.
In [76], Tong et al. introduced a two-step OMP-based channel estimation algorithm.
First, a composite convolution kernel function (CKF) is constructed based on the autocorre-
lation matrix, which then coarsely estimates the angles of arrival/departure (AoAs/AoDs)
for multipath channels. Second, a squeeze-and-excitation ResNet (SE-ResNet) utilizing the
Noise2Void (N2V) algorithm is proposed to refine the AoAs/AoDs estimation. Then, the
channel amplitude is estimated using the LS algorithm. Finally, the channel matrix is accu-
rately recovered based on all the results above. In the experiments, the proposed algorithm
exhibits a superior NMSE performance over the simultaneous weighted OMP (SW-OMP)
algorithm, the Newtonized OMP (NOMP) algorithm, and the channel estimation neural
network (CENN), while having computational complexity.
Electronics 2023, 12, 4965 25 of 38

To enhance the effectiveness of the conventional OMP algorithm in hybrid-field chan-

nel estimation, Nayir et al. [77] developed a novel scheme called OMP-CAE by incorporat-
ing the convolutional autoencoder (CAE) for massive MIMO channels. In this scheme, the
OMP algorithm first provides a rough channel estimation. Since the channel parameters
from the OMP rough estimation contain significant errors at low SNR, they are fed into
the CAE network. Then, the CAE network uses its strong denoising ability to improve
the accuracy of channel estimation. Through experimental verification, the OMP-CAE
surpasses the MMSE algorithm and the conventional OMP algorithm in terms of NMSE
and applies to different scenarios.

3.4.3. Model-Driven Channel Estimation Combining AMP Algorithm

The AMP algorithm is a significant iterative algorithm that excels at recovering sparse
signals. It can achieve channel estimation in sparse channel environments with low com-
putational complexity [78]. Nevertheless, the AMP algorithm’s performance is highly
dependent on certain preset parameters, and it presents a challenge to determine the opti-
mal parameters, leading to its limited estimation performance. Therefore, some researchers
have combined the AMP algorithm with deep learning. This approach allows the deep
learning network to learn and optimize the iterative steps of the AMP algorithm, forming a
novel and effective channel estimation method.
The learned AMP (LAMP) algorithm is a variant of the AMP algorithm that unfolds
and maps the AMP algorithm’s iterative steps directly into a DNN, thereby utilizing the
DNN’s ability to jointly optimize the coefficients of the linear transformations and the
parameters of the nonlinear shrinkage function. However, the existing LAMP networks
may not achieve the desired performance in beamspace channel estimation. In light of this,
Wei et al. [79] investigated an enhanced scheme based on a prior-assisted Gaussian mixture
LAMP (GM-LAMP). Specifically, the authors replaced the soft threshold shrinkage function
in the original LAMP network with a Gaussian mixture shrinkage function, which is able to
reflect more of the beamspace channel’s prior information. Compared to the OMP, the AMP,
and the LAMP algorithms, the GM-LAMP algorithm has a superior NMSE performance
with less pilot overhead.
The denoising AMP (DAMP) algorithm is also a variant stemming from the AMP
algorithm, replacing the shrinkage function with a denoiser. Pu et al. [80] deeply unfolded
the DAMP algorithm by replacing its original denoiser with a DnCNN. They developed
the model-driven learned denoising AMP (LDAMP) algorithm for channel estimation with
noisy channels in OTFS systems. Then, they predicted the theoretical NMSE performance
of the LDAMP algorithm using the state evolution (SE) equation. Owing to the integration
of the DAMP algorithm’s superior performance and the nonlinear fitting capability of deep
learning, the proposed algorithm exhibits effectiveness and high accuracy.
In 2022, Wang et al. [81] introduced an AMP-based multi-stage scheme with deep
learning for quasi-sparse channel environments. This scheme treats the entire system as
an end-to-end DNN, and each iterative process of the sensing matrix, noise introduction,
and AMP algorithm is regarded as a DNN layer. Firstly, the AMP sensing matrix is trained
to adapt to the quasi-sparse channel and learn the optimal sensing matrix. Secondly, the
AMP algorithm’s nonlinear shrinkage parameters and linear coefficients are optimized
layer-by-layer. Finally, all trainable parameters are jointly optimized, thereby optimizing
the entire system. Compared to the AMP, the LAMP, and the LDAMP algorithms, the
proposed scheme has a superior NMSE performance while reducing pilot overhead.
Table 4 summarizes the reviewed model-driven deep-learning-based channel esti-
mation methods and their performance. The model-driven approach utilizes some prior
knowledge of communication systems, which can enhance its interpretability. Based on
existing models in the physical layer, it decreases the quantity of parameters to be learned
and the sample complexity of the learning model. However, the model-driven approach
is less flexible, and the model could be negatively impacted if the prior knowledge is
inaccurate in practical scenarios.
Electronics 2023, 12, 4965 26 of 38

Table 4. Model-driven deep-learning-based channel estimation methods and their performance.

Reference Deep Learning Model Input Output Compared with Performance Complexity
The NMSE performance surpasses the
LMMSE, robust LMMSE
Dual CNN LS rough channel Accurate channel (RLMMSE) algorithms, RLMMSE algorithm, the SFCNN, and FLOPs = 3.7 M,
Jiang et al. [71] (SFCNN + ADCNN) estimation result parameters = 1764
estimation result SFCNN, and ADCNN
the ADCNN, while it is comparable to
the LMMSE algorithm
Has the best NMSE and BER
performances on SoC, the resource
LS rough channel Accurate channel Execution time = 0.0179 ms,
Haq et al. [72] DNN LS and MMSE algorithms utilization is lower than the MMSE SoC power = 2.849 W
estimation result estimation result algorithm but higher than the
LS algorithm
Has a better MSE performance when the
NDR-Net (NLE + DnCNN LS rough channel Accurate channel LS, MMSE algorithms, SNR is mismatched, with nearly 5–7 dB Parameters = 1231 K,
Li et al. [73] + residual learning) estimation result FLOPs = 67 K
estimation result and DnCNN of gain compared to the
MMSE algorithm
Significantly superior to the OMP
OMP rough channel Accurate channel algorithm in terms of NMSE, especially
Li et al. [75] ResNet OMP algorithm \
estimation result estimation result when the frame size is 128 × 16, and the
NMSE is 0.00173, the gain is close to 6dB
O((Kiter + 1) MNs Gr Gt
Coarsely estimated Finely estimated SW-OMP, NOMP Has better NMSE performance even with
Tong et al. [76] SE-ResNet log( Gr Gt ) +Kiter Nlayer
AoAs/AoDs AoAs/AoDs algorithms, and CENN low SNR and fewer training frames
Gr Gt Sfilter Cin Cout )
OMP rough channel Accurate channel Performs best in terms of NMSE, even at
Nayir et al. [77] CAE OMP and MMSE algorithms \
estimation result estimation result low SNR
Measurement signal vector OMP, AMP and Achieves better NMSE performance with
Wei et al. [79] GM-LAMP Estimated channel O( TMN )
and beam selection matrix LAMP algorithms lower pilot overhead
Has the best NMSE performance, with
Received signal vector and OMP, AMP algorithms,
the NMSE already less than 10−2 in the O L MN + M2 N 2



Pu et al. [80] LDAMP Estimated channel
transmitted signal matrix and ResNet
low SNR scenario with an SNR of 2dB
The NMSE performance outperforms the
AMP, LAMP, and AMP and the LAMP algorithms; it
Wang et al. [81] DNN Original channel Estimated channel \
LDAMP algorithms outperforms the LDAMP algorithm
when the SNR is above 8 dB
Electronics 2023, 12, 4965 27 of 38

According to the discussion in this section, we can learn that most existing studies on
channel estimation using deep learning mainly concentrate on the data-driven approach.
However, both data-driven and model-driven approaches have the potential to achieve or
even exceed conventional methods’ performance in the appropriate application scenarios.
Table 5 provides a detailed comparison of the advantages and disadvantages of these
two approaches.
Nevertheless, the studies mentioned above are limited to traditional communica-
tion scenarios and do not address future communication scenarios. As one of the critical
techniques for 6G networks, the research on RIS is in full swing. The introduction of
RIS improves the performance of communication systems but also increases their com-
plexity, making conventional techniques, including channel estimation, confront new
challenges. With the accelerated advancement in deep learning, numerous researchers
have adopted this technique to solve the aforementioned issue. The following section
will introduce channel estimation methods based on deep learning for RIS-aided wireless
communication systems.

Table 5. Comparison of the advantages and disadvantages of data-driven and model-driven approaches.

Characteristics Data-Driven Approach Model-Driven Approach

Self-learning/automatic feature extraction X
High accuracy X (With sufficient training data) X (When the model is well designed)
Advantages
Utilization of prior knowledge X
Interpretability X
Large data requirements X
Risk of overfitting X
Disadvantages
High computational complexity X
Inflexibility X

4. Channel Estimation Based on Deep Learning for RIS-Aided

Communication Systems
RIS is one of the most promising techniques for 6G communications, consisting of
numerous passive reflective metamaterial antenna elements that form a plane with low
cost, low power consumption, easy deployment, etc. RIS can reconfigure the wireless
propagation environment by intelligently adjusting the amplitude and phase to enhance
wireless communications’ spectrum and energy efficiency [82]. When combined with
wireless communication systems, RIS can modify or improve existing communication
concepts, thereby assisting in meeting the high standards of 6G communication. Figure 18
shows an example of RIS-aided communication systems. In this setup, the base station (BS)
is outfitted with M antennas and serves K users, each with a single antenna, and the RIS
is outfitted with a total of N passive reflective elements. Furthermore, the BS adjusts the
parameters of the reflective elements through the RIS controller. The direct link between
the BS and the users is blocked, making the BS unable to communicate properly with the
users. With the deployment of RIS, an additional link between them is established.
Let hd,k ∈ C M×1 represent the direct channel between the kth user and the BS, hr,k ∈
C N × 1 is the channel between the kth user and the RIS, and G ∈ C M× N is the channel
between the RIS and the BS. Then, the received signal at BS can be expressed as:

K
y= ∑ (hd,k + Gdiag(θ )hr,k )uk + v (12)
k =1

where θ = [θ1 , θ2 , . . . , θ N ]T denotes the reflection vector of the RIS, θn is the reflection
coefficient of the nth reflection element, uk is the symbol sent by the kth user, and v ∈ C M×1
is the received noise. In particular, θn can be further denoted by θn = β n e jφn , with β n ∈ [0, 1]
Electronics 2023, 12, 4965 28 of 38

and φn ∈ [0, 2π ] corresponding to the amplitude and phase of the nth reflection element,
respectively. Since passive RIS elements cannot actively transmit and receive signals and
perform additional signal processing, hr,k ∈ C N ×1 and G ∈ C M× N cannot be directly
estimated. Therefore, it is necessary to estimate the channel between the kth user and the
RIS and between the RIS and the BS as the cascaded channel Hk , Gdiag(hr,k ) ∈ C M× N .
In this instance, Equation (12) can be re-expressed as:

K
y= ∑ (hd,k + Hk θ )uk + v (13)
k =1

It is noteworthy that, in a TDD system, the downlink CSI can be obtained from the
uplink channel estimation because of the channel reciprocity [83].

Passive reflection element

RIS Controller
RIS

×
BS
Blockage K Users
Figure 18. RIS-aided communication system.

Channel estimation in RIS-aided systems presents more challenges compared to tra-

ditional systems. During signal reflection, RIS functions as a passive device without the
capability to receive or sample signals. Therefore, it can only carry out low-dimensional
signal sampling at the transmitter and receiver to estimate the high-dimensional CSI. When
the number of RIS reflective elements is substantial, the channel matrix that needs to be
estimated becomes larger in dimension, increasing the overhead and computational com-
plexity required for channel estimation. Suppose a few passive reflection elements of RIS
are replaced with active elements (i.e., semi-passive RIS architecture). In that case, the
sensing devices can estimate the CSI of the separate channels. Still, installing the active ele-
ments will diminish the benefit of RIS as a low-cost passive device. In response to the above
challenges, numerous researchers have started to work on utilizing deep learning’s strong
ability to estimate RIS-aided channels. Massive MIMO, millimeter wave (mmWave), and
multi-user (MU) systems are all popular areas of research in 5G and 6G communications.
Therefore, this section will concentrate on channel estimation for RIS-aided communication
systems based on deep learning related to them.

4.1. Application of Deep Learning in Channel Estimation for RIS-Aided Massive MIMO Systems
Massive MIMO systems use a lot of antenna elements to achieve a high level of spatial
multiplexing and beamforming, which improves the throughput and spectral efficiency
of wireless communications. These advantages make massive MIMO an integral part of
5G and subsequent technologies. However, it poses challenges such as low coverage, high
power consumption, and high cost. To mitigate these issues, RIS has been introduced as a
Electronics 2023, 12, 4965 29 of 38

promising solution. Nonetheless, the employment of RIS makes channel estimation more
difficult. Against this backdrop, the intervention of deep learning offers a novel approach
to addressing the complex issue of channel estimation.
In 2022, Mao et al. [84] utilized the cascaded channel’s quasi-sparse structure to
transform the channel estimation for RIS-aided massive MIMO systems into a CS problem.
Their scheme targeted the uplink, introducing a residual structure-based OMP (RS-OMP)
architecture. This architecture is designed to address the impacts of noise and grid mismatch
on the traditional OMP algorithm. Within this architecture, the ResNet uses a rectangular
convolutional kernel, which makes it more capable of learning the row or column sparsity.
Through experimental comparison, the proposed RS-OMP outperforms the DnCNN and
the traditional OMP algorithm in terms of NMSE.
In [85], Xie et al. investigated an estimation method based on deep compressive
sensing (DCS) for RIS-aided massive MIMO channels, with the aim of reducing the pilot
overhead. Specifically, they developed a ResU-Net by combining U-Net with the DCS
framework while introducing residual learning. The received low-dimensional pilot signals
are first transformed into high-dimensional signals by matrix multiplication. Then, the
ResU-Net with skip connections is utilized to compress the input data and recover the high-
dimensional cascaded channel. Experiments are conducted using varying pilot lengths
and quantization bits, and the results verify the ResU-Net’s excellent generalization ability
and robustness.
In 2023, Liu et al. [86] proposed a skip-connection attention (SC-attention) network for
accurate CSI acquisition in double-RIS-aided massive MIMO systems. In this scheme, they
modelled channel estimation as an image restoration-like denoising problem. Moreover,
they used self-attention layers to improve the model capacity and receptive field size of
the network, which allows for recovering the CSI matrix more accurately. Meanwhile, by
employing the skip-connection architecture, the estimation accuracy can be enhanced by
16% while reducing the computational complexity. Through experimental comparison,
the SC-attention network exhibits a superior NMSE performance over the CNN and the
conventional algorithms (LS and MMSE).

4.2. Application of Deep Learning in Channel Estimation for RIS-Aided mmWave Systems
Many researchers have pushed for the application of mmWave to enhance wireless
system performance, which can generate narrow, highly energy-concentrated information-
carrying beams. As one of the critical 6G techniques, mmWave offers higher transmission
rates and greater connection capacity for wireless communications due to its ultra-high-
frequency bandwidth. However, mmWave communication encounters challenges in cover-
age and signal quality. In addressing these issues, the emergence of RIS offers a promising
solution: through the precise control of signals and improvement of their transmission
properties, more excellent coverage can be achieved; additionally, by suppressing multipath
fading and interference, the signal quality can be significantly enhanced [87]. Nevertheless,
introducing RIS makes the higher-dimensional CSI need to be estimated. As an efficient tool,
deep learning holds promise for successfully coping with this high-dimensional complexity.
For RIS-aided channels, increasing users and reflective elements elevate the train-
ing overhead, making accurate CSI acquisition difficult. To deal with this issue,
Shtaiwi et al. [88] proposed a two-stage channel estimation scheme for entirely passive RIS-
aided mmWave systems. In the initial stage, a pilot-based method for uplink is designed,
where only randomly selected user terminals send known pilot sequences to the BS through
the RIS, reducing the active user count during the training period. In the subsequent stage,
a CNN-based spatial–temporal–spectral architecture (STS-CNN) is proposed to recover the
CSI of inactive user terminals in the first stage. Through experimental comparison, this
scheme outperforms the benchmark in terms of MSE. However, more research is needed
on the CNN-based estimation method, as its complexity can be further reduced.
For semi-passive RIS-aided mmWave systems with a few distributed active sensing
devices, Jin et al. [89] leveraged the sparsity of RIS-aided channels and used the image
Electronics 2023, 12, 4965 30 of 38

super-resolution method for channel estimation. In particular, their scheme used the
multi-scale deep super-resolution (MDSR) and the enhanced deep super-resolution (EDSR)
networks. First, the EDSR combines global feature fusion and residual blocks to obtain
CSI for single-scale sparse channels, simplifying the RIS hardware. Next, the MDSR is
employed to guarantee that the scale of active elements is adjustable to flexibly adapt to
multi-scale, low-sparse, and low-resolution channel scenarios. Experiments demonstrate
the scheme’s effectiveness, and the estimation performance improves as the sensing device
density and network size increase.
In 2023, Feng et al. [90] similarly transformed the channel estimation into an image
super-resolution problem. Specifically, they developed a cascade channel estimation scheme
for RIS-aided mmWave single-input–multi-output (SIMO) systems based on global atten-
tion ResNet (GARN). The RIS unit cells are first grouped in this scheme to reduce the pilot
overhead. Then, the initial channel estimation is obtained using the LS algorithm, serving as
the input to GARN. Since the GARN adopts the global attention mechanism, it can extract
multi-channel features and enhance information fusion, improving the accuracy of channel
estimation matrices. Experimental comparisons with multiple benchmark methods demon-
strate the proposed scheme’s excellent estimation performance and generalization ability.
In the same year, Abdallah et al. [91] developed two deep-learning-based schemes
for RIS-aided mmWave MIMO systems, i.e., data-driven frequency-flat cascaded channel
estimation (DD-FF-CE) and data-driven frequency-selective cascaded channel estimation
(DD-FS-CE). For the DD-FF-CE, the authors turned channel estimation into a CS problem
by leveraging the angular cascaded channel’s double-structured sparsity and solving it
using deep learning. In particular, they used two DnCNNs to find the row and column
supports and then the LS algorithm to estimate the angular cascade channel. In contrast to
the DD-FF-CE, the DD-FS-CE also utilizes the support information from each subcarrier to
cope with frequency selectivity. In the experiments, these two schemes exhibit excellent
NMSE performance and low computational complexity while having less pilot overhead.

4.3. Application of Deep Learning in Channel Estimation for RIS-Aided MU Systems

MU communication systems serve a vital role in modern wireless communications,
which improve the spectrum efficiency by enabling multiple users to share the same
spectrum resource simultaneously. By adjusting its surface units, RIS can modulate the
reflective properties of signals to optimize channel conditions, thereby enhancing the system
performance. However, accurately estimating the channels in RIS-aided MU systems
presents a new challenge. Utilizing the powerful learning capability of deep learning to
address this issue is a hot trend in current research.
In 2021, Ginige et al. [92] innovatively proposed a DNN based on the deep image
prior (DIP) in channel estimation for RIS-aided MU SIMO-OFDM systems. This network is
untrained and can optimize its parameters in real time. In this scheme, the LS algorithm
first estimates the effective channel, which is then denoised using the untrained DNN. After
denoising, the channel utilizes the information from the DFT-based reflection pattern matrix
to obtain the CSI of the direct and cascade channels. In simulations, the untrained DNN
exhibits a superior NMSE performance over the conventional methods (LS and MMSE),
while it has low complexity because no pre-training is required. Moreover, it can maintain
high accuracy in the case of hardware impairments and adjust flexibly to various scenarios
and channel settings.
In 2022, Liu et al. [93] modelled channel estimation for RIS-aided MU communication
(MUC) systems as a noise reduction problem and developed a flexible deep residual
learning (DReL)-based channel estimation framework. They derived an MMSE estimator
based on a deep residual network (DRN) using Bayesian theory under the MMSE criterion
within this framework. In addition, they introduced a CNN-based DRN (CDRN) designed
to recover the noisy channel matrix. The network features a CNN denoising module with an
element-by-element subtraction structure. This design enables it to simultaneously extract
the spatial properties of the noisy channel matrix and the additive characteristics of the
Electronics 2023, 12, 4965 31 of 38

noise. Experimental results demonstrate that this framework does not require additional
sensing devices in the RIS. Furthermore, its NMSE performance closely approximates
the optimal MMSE estimator, which requires deriving from the cascade channel’s prior
probability density function.
In RIS-aided MU MIMO-OFDM systems, the cascaded channel has high dimension-
ality and complex statistical properties. These factors make it challenging to achieve the
ideal MMSE algorithm with the best performance. In [94], Shen et al. applied the concept
of image super-resolution and developed an SRDnNet, which combines the properties of
SRCNN and DnCNN to obtain accurate CSI. The central idea of SRDnNet is to regard the
estimated channel of pilot positions as a low-resolution image. It uses the enhanced SRCNN
to extract input features, learns the appropriate interpolation method, and generates a
coarsely estimated channel matrix. After that, the DnCNN performs denoising to recover
the channel coefficients from the noisy channel matrix. Compared to the ChannelNet [57],
the SRDnNet can receive rawer inputs and requires just one division operation between the
received and transmitted pilots, leading to decreased computational and time consumption.
Without prior information on the direct and cascaded channels, the SRDnNet outperforms
the conventional LS, MMSE algorithms, and the ChannelNet, with NMSE gains exceeding
10 dB.
In 2023, Li et al. [95] proposed a DCSaNet based on dilated convolution and self-
attention to enhance the channel estimation accuracy of RIS-aided MUC systems. This
network uses dilated convolutional blocks to enlarge the receptive field and optimize
feature extraction. Additionally, it employs self-attention blocks to enable the system
to concentrate on critical channel features. To meet the requirements of the real-world
deployment of RIS, the authors further designed the lightweight DCSaNet-l network, which
reduces the parameters by a significant amount (nearly 83.7%) but performs almost as well
as DCSaNet. The simulation results demonstrate that these networks substantially improve
the NMSE performance and the training speed across various channel dimensions and
SNR levels.
This section focuses on deep-learning-based channel estimation methods for RIS-aided
communication systems. These methods cover massive MIMO, mmWave, and MU systems
and are summarized in Table 6. With deep learning, we can more accurately estimate
complex RIS channels. However, there may still be challenges in this area of research. First,
most existing studies on RIS focus on passive RIS. The passive reflection elements of RIS
introduce high-dimensional channel parameters, which increase the input dimensions of
the deep learning models. As a result, the complexity and training difficulty of the models
rise. Second, the deployment location of RIS and the surrounding environment impact
its reflective properties. Hence, the model must have a strong generalization capability,
ensuring its performance in various scenarios is as effective as the one it was trained
for. Given these, how to combine RIS technology with deep learning technology more
effectively is an important topic worth studying in the future.
Electronics 2023, 12, 4965 32 of 38

Table 6. Research on channel estimation based on deep learning for RIS-aided communication systems.

Reference System RIS Architecture Deep Learning Model Input Output Complexity
OMP rough channel Accurate channel O(KMNTt + ∑lD=1 MNSl
Mao et al. [84] RIS-aided massive MIMO Entirely passive ResNet
estimation result estimation result Cl −1 Cl )
CPU time = 95.7 ms,
Xie et al. [85] RIS-aided massive MIMO Entirely passive ResU-Net Received pilot signal Estimated channel
GPU time = 4.27 ms
Double-RIS-aided LS rough channel Accurate channel
Liu et al. [86] Entirely passive SC-attention network \
massive MIMO estimation result estimation result
Shtaiwi et al. [88] RIS-aided mmWave MIMO Entirely passive STS-CNN Partial CSI Entire CSI \
EDSR: average running
RIS-aided mmWave time = 7.8 ms
Jin et al. [89] Semi-passive EDSR + MDSR Partial CSI Entire CSI
massive MIMO MDSR: average running
time = 7.5 ms
GARN (ResNet + Grouped LS channel Reconstructed complete Parameters = 2.061 M,
Feng et al. [90] RIS-aided mmWave SIMO Entirely passive
global attention) estimation matrix channel matrix FLOPs = 1.807 G
O(UNBS Q + ULG NRIS

Q
Abdallah et al. [91] RIS-aided mmWave MIMO Entirely passive DnCNN Received pilot signal Residual noise
+ ULG 2QL2u + L3u
LS rough channel Accurate channel
Ginige et al. [92] RIS-aided MU SIMO-OFDM Entirely passive DNN \
estimation result estimation result
Liu et al. [93] RIS-aided MUC Entirely passive CDRN (CNN + DRN) Noisy channel matrix Denoised channel matrix computation time = 2.66 ms
RIS-aided MU SRDnNet Estimated channel at
Shen et al. [94] Entirely passive Estimated whole channel Predict time = 1.61 × 10−2 s
MIMO-OFDM (SRCNN + DnCNN) pilot positions
DCSaNet (dilated LS rough channel Accurate channel Execution time = 5.2 × 10−4 s,
Li et al. [95] RIS-aided MUC Entirely passive
convolution + self-attention) estimation result estimation result training time = 246 s
Electronics 2023, 12, 4965 33 of 38

5. Challenges and Future Research Trends

Deep learning techniques have demonstrated the capability to effectively process large
amounts of data and enhance the channel estimation performance. However, it is essential
to acknowledge that several challenges in this field remain unresolved. According to the
current advancements in wireless communications, the prospective research trends for
deep-learning-based channel estimation are outlined below.

5.1. Theoretical Exploration

In deep-learning-based channel estimation, strict mathematical proof is essential to
define its performance limits. However, this approach is not backed by a solid mathematical
foundation. Moreover, mathematical and physical theories explaining deep learning are still
exploratory, making it unclear why certain network structures could achieve a good channel
estimation performance [96]. Additionally, without advanced optimization theory, finding
the optimal policy and loss function during training is challenging. Hence, overcoming the
theoretical challenge is urgently needed and research-worthy.

5.2. Enhancing Estimation with Limited Data

The estimation accuracy of deep-learning-based methods highly depends on the
quantity and quality of training and testing data. However, obtaining channel data can be
challenging in some special communication scenarios. With limited or no data available,
the performance of these methods often suffers. Therefore, it is necessary to combine
multiple approaches in channel estimation algorithms. In this case, integrating transfer
learning [97] or meta-learning [98] with deep learning could be an effective strategy to
enhance the channel estimation accuracy.

5.3. Attention Mechanism in Noise Processing

In prior research, noise processing in channel estimation often relies on image super-
resolution algorithms or denoising neural networks. However, these approaches may
destroy the actual channel information [99]. The attention mechanism [100] is a significant
deep learning technique different from traditional neural networks, which treat all input
data equally important. It assigns varying weights to different parts of the data, focusing
more precisely on the most critical information. Therefore, adding the attention mechanism
to channel estimation could be a valuable direction for future research. By doing so, the
deep learning model would focus more on the effective information in the received signal
and reduce noise impact by adjusting feature weights, thereby achieving more accurate
channel estimation.

5.4. Adaptation to Dynamic Environments

A fixed channel estimation model may no longer maintain high accuracy in dynamic
wireless communication environments. If the model is retrained to adapt, it will consume
significant time and computational resources. This approach fails to meet the real-time and
low-cost requirements. Therefore, research on adaptive model design can be strengthened
in future work. Such a design would enable the model to adjust its parameters and
structure autonomously, adapting to complex and changing environments for stable and
high-performance channel estimation.

5.5. Channel Estimation for Other 6G Communication Scenarios

Beyond the RIS-aided communication systems we have mentioned, there are numer-
ous emerging scenarios for next-generation wireless communications. These scenarios
include those that employ terahertz (THz) communication technology [101] and visible
light communication (VLC) technology [102]. Among them, THz communications can bring
massive spectrum resources that meet the requirements of 6G. Additionally, VLC adopts
an ultra-high bandwidth for high-speed data transmission, offering a wide range of appli-
Electronics 2023, 12, 4965 34 of 38

cations in 6G. Given these, extending deep-learning-based channel estimation research to

these two communication scenarios is necessary and an important future research direction.

5.6. Lightweight Networks for Channel Estimation

From extensive previous research, it is evident that large-size deep learning models can
learn more complex data features due to more parameters and a deeper network structure.
These factors enable them to achieve higher channel estimation performance. However,
the performance improvement is often accompanied by higher computational complexity
and resource consumption, which may be a significant limitation in resource-constrained
practical applications. In light of this challenge, lightweight networks offer promising
prospects for channel estimation. These networks are characterized by optimized structures
and a reduced parameter count. Specifically, the design of these networks aims to balance
the high estimation performance with reduced complexity and resource demands.

6. Conclusions
As wireless communication technology advances rapidly, the outline of 5G becomes
increasingly clear. Meanwhile, with the advent of 6G on the horizon, integrating deep learn-
ing with traditional communication techniques has emerged as a focal point in academia.
In this paper, we reviewed recent research on deep-learning-based channel estimation
and classified it into data-driven and model-driven approaches. Meanwhile, we further
summarized the application of deep learning in channel estimation for RIS-aided commu-
nication systems, one of the future novel communication scenarios. Numerous studies
have showcased that deep-learning-based methods have the potential to match or surpass
conventional methods. However, most of the technical implementations are still in their
infancy, and their applications still face many challenges; thus, we also provided an out-
look on possible future research trends. We expect that this paper can assist researchers
interested in applying deep learning to channel estimation.

Author Contributions: Conceptualization, Z.L.; methodology; investigation, C.L.; writing—original

draft preparation, C.L.; writing—review and editing, Z.L.; supervision, Z.L.; project administra-
tion, Z.L.; funding acquisition, Z.L. All authors have read and agreed to the published version of
the manuscript.
Funding: This work was supported in part by the National Natural Science Foundation of China
under Grant (No. 61801319), the Sichuan Science and Technology Program under Grant (No.
2020JDJQ0061, No. 2021YFG0099), the Innovation Fund of Chinese Universities under Grant (No.
2020HYA04001), the Sichuan University of Science and Engineering Talent Introduction Project under
Grant (No. 2020RC33), the Postgraduate Innovation Fund Project of Sichuan University of Science
and Engineering (No. Y2023285).
Data Availability Statement: No new data were created or analyzed in this study. Data sharing is
not applicable to this article.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Henry, S.; Alsohaily, A.; Sousa, E.S. 5G is Real: Evaluating the Compliance of the 3GPP 5G New Radio System with the ITU
IMT-2020 Requirements. IEEE Access 2020, 8, 42828–42840. [CrossRef]
2. You, X.; Wang, C.X.; Huang, J.; Gao, X.; Zhang, Z.; Wang, M.; Huang, Y.; Zhang, C.; Jiang, Y.; Wang, J.; et al. Towards 6G wireless
communication networks: Vision, enabling technologies, and new paradigm shifts. Sci. China Inf. Sci. 2021, 64, 110301. [CrossRef]
3. Wang, C.X.; You, X.; Gao, X.; Zhu, X.; Li, Z.; Zhang, C.; Wang, H.; Huang, Y.; Chen, Y.; Haas, H.; et al. On the road to 6G: Visions,
requirements, key technologies and testbeds. IEEE Commun. Surv. Tutor. 2023, 25, 905–974. [CrossRef]
4. Cui, M.; Dai, L. Channel Estimation for Extremely Large-Scale MIMO: Far-Field or Near-Field? IEEE Trans. Commun. 2022,
70, 2663–2677. [CrossRef]
5. Kolliboina, S.S.; Teja, S.; Giridhar, K. Non-Parametric Adaptive Thresholding for Channel Estimation of OTFS-Based 6G
Communication Links. In Proceedings of the 2022 IEEE Globecom Workshops (GC Wkshps), Rio de Janeiro, Brazil, 4–8 December
2022; pp. 1561–1566. [CrossRef]
Electronics 2023, 12, 4965 35 of 38

6. Mao, Q.; Hu, F.; Hao, Q. Deep learning for intelligent wireless networks: A comprehensive survey. IEEE Commun. Surv. Tutor.
2018, 20, 2595–2621. [CrossRef]
7. Qin, Z.; Ye, H.; Li, G.Y.; Juang, B.H.F. Deep learning in physical layer communications. IEEE Wirel. Commun. 2019, 26, 93–99.
[CrossRef]
8. Huang, H.; Guo, S.; Gui, G.; Yang, Z.; Zhang, J.; Sari, H.; Adachi, F. Deep Learning for Physical-Layer 5G Wireless Techniques:
Opportunities, Challenges and Solutions. IEEE Wirel. Commun. 2020, 27, 214–222. [CrossRef]
9. Ye, H.; Li, G.Y.; Juang, B.H. Power of Deep Learning for Channel Estimation and Signal Detection in OFDM Systems. IEEE Wirel.
Commun. Lett. 2018, 7, 114–117. [CrossRef]
10. Aboulfotouh, A.; De Oliveira, T.E.A.; Fadlullah, Z.M. Channel Estimation in Cellular Massive MIMO: A Data-Driven Approach.
In Proceedings of the 2022 IEEE International Conference on Internet of Things and Intelligence Systems (IoTaIS), Bali, Indonesia,
24–26 November 2022; pp. 290–296. [CrossRef]
11. Thakkar, K.; Goyal, A.; Bhattacharyya, B. Deep Learning and Channel Estimation. In Proceedings of the 2020 6th International
Conference on Advanced Computing and Communication Systems (ICACCS), Coimbatore, India, 6–7 March 2020; pp. 745–751.
[CrossRef]
12. Gao, X.; Jin, S.; Wen, C.K.; Li, G.Y. ComNet: Combination of Deep Learning and Expert Knowledge in OFDM Receivers. IEEE
Commun. Lett. 2018, 22, 2627–2630. [CrossRef]
13. He, H.; Wen, C.K.; Jin, S.; Li, G.Y. Deep Learning-Based Channel Estimation for Beamspace mmWave Massive MIMO Systems.
IEEE Wirel. Commun. Lett. 2018, 7, 852–855. [CrossRef]
14. Zamanipour, M. A survey on deep-learning based techniques for modeling and estimation of massivemimo channels. arXiv 2019,
arXiv:1910.03390.
15. Boloursaz Mashhadi, M.; Gündüz, D. Deep learning for massive MIMO channel state acquisition and feedback. J. Indian Inst. Sci.
2020, 100, 369–382. [CrossRef] [PubMed]
16. Peng, G.; Zhu, Z.; Liang, Y. Application of Deep Learning in Wireless Networks for Channel Estimation: A Survey. J. Phys. Conf.
Ser. 2022, 2203, 012044. [CrossRef]
17. Gizzini, A.K.; Chafii, M. A Survey on Deep Learning Based Channel Estimation in Doubly Dispersive Environments. IEEE Access
2022, 10, 70595–70619. [CrossRef]
18. Vilas Boas, E.C.; e Silva, J.D.; de Figueiredo, F.A.; Mendes, L.L.; de Souza, R.A. Artificial intelligence for channel estimation in
multicarrier systems for B5G/6G communications: A survey. EURASIP J. Wirel. Commun. Netw. 2022, 2022, 116. [CrossRef]
19. Cui, T.J.; Qi, M.Q.; Wan, X.; Zhao, J.; Cheng, Q. Coding metamaterials, digital metamaterials and programmable metamaterials.
Light. Sci. Appl. 2014, 3, e218. [CrossRef]
20. Wei, X.; Shen, D.; Dai, L. Channel Estimation for RIS Assisted Wireless Communications—Part I: Fundamentals, Solutions, and
Future Opportunities. IEEE Commun. Lett. 2021, 25, 1398–1402. [CrossRef]
21. Oyerinde, O.O.; Mneney, S.H. Review of Channel Estimation for Wireless Communication Systems. IETE Tech. Rev. 2012,
29, 282–298. [CrossRef]
22. Dong, X.; Lu, W.S.; Soong, A.C. Linear Interpolation in Pilot Symbol Assisted Channel Estimation for OFDM. IEEE Trans. Wirel.
Commun. 2007, 6, 1910–1920. [CrossRef]
23. Coleri, S.; Ergen, M.; Puri, A.; Bahai, A. Channel estimation techniques based on pilot arrangement in OFDM systems. IEEE
Trans. Broadcast. 2002, 48, 223–229. [CrossRef]
24. Dowler, A.; Doufexi, A.; Nix, A. Performance evaluation of channel estimation techniques for a mobile fourth generation wide
area OFDM system. In Proceedings of the IEEE 56th Vehicular Technology Conference, Vancouver, BC, Canada, 24–28 September
2002; Volume 4, pp. 2036–2040. [CrossRef]
25. Kwak, K.; Lee, S.; Kim, J.; Hong, D. A New DFT-Based Channel Estimation Approach for OFDM with Virtual Subcarriers by
Leakage Estimation. IEEE Trans. Wirel. Commun. 2008, 7, 2004–2008. [CrossRef]
26. Jeon, W.G.; Paik, K.H.; Cho, Y.S. An efficient channel estimation technique for OFDM systems with transmitter diversity. In
Proceedings of the 11th IEEE International Symposium on Personal Indoor and Mobile Radio Communications. PIMRC 2000.
Proceedings (Cat. No.00TH8525), London, UK, 18–21 September 2000; Volume 2, pp. 1246–1250. [CrossRef]
27. Liu, Y.; Tan, Z.; Hu, H.; Cimini, L.J.; Li, G.Y. Channel Estimation for OFDM. IEEE Commun. Surv. Tutor. 2014, 16, 1891–1908.
[CrossRef]
28. Gong, Y.; Letaief, K. Low rank channel estimation for space-time coded wideband OFDM systems. In Proceedings of the IEEE
54th Vehicular Technology Conference. VTC Fall 2001. Proceedings (Cat. No.01CH37211), Atlantic City, NJ, USA, 7–11 October
2001; Volume 2, pp. 772–776. [CrossRef]
29. Luo, Z.; Huang, D. General MMSE Channel Estimation for MIMO-OFDM Systems. In Proceedings of the 2008 IEEE 68th
Vehicular Technology Conference, Calgary, AB, Canada, 21–24 September 2008; pp. 1–5. [CrossRef]
30. Hung, K.C.; Lin, D.W. Pilot-Based LMMSE Channel Estimation for OFDM Systems With Power–Delay Profile Approximation.
IEEE Trans. Veh. Technol. 2010, 59, 150–159. [CrossRef]
31. Rossi, P.S.; Müller, R.R.; Edfors, O. Linear MMSE estimation of time–frequency variant channels for MIMO-OFDM systems.
Signal Process. 2011, 91, 1157–1167. [CrossRef]
32. Yu, J.L.; Zhang, B.; Chen, P.T. Blind and semi-blind channel estimation with fast convergence for MIMO-OFDM systems. Signal
Process. 2014, 95, 1–9. [CrossRef]
Electronics 2023, 12, 4965 36 of 38

33. Shin, C.; Heath, R.W.; Powers, E.J. Blind channel estimation for MIMO-OFDM systems. IEEE Trans. Veh. Technol. 2007, 56, 670–685.
[CrossRef]
34. Kang, W.; Champagne, B. Subspace-based blind channel estimation: Generalization and performance analysis. IEEE Trans. Signal
Process. 2005, 53, 1151–1162. [CrossRef]
35. Tu, C.C.; Champagne, B. Subspace-Based Blind Channel Estimation for MIMO-OFDM Systems With Reduced Time Averaging.
IEEE Trans. Veh. Technol. 2010, 59, 1539–1544. [CrossRef]
36. Acar, T.; Petropulu, A. Blind MIMO system identification using PARAFAC decomposition of an output HOS-based tensor. In
Proceedings of the The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, Pacific Grove, CA, USA, 9–12
November 2003; Volume 1, pp. 1080–1084. [CrossRef]
37. Choqueuse, V.; Mansour, A.; Burel, G.; Collin, L.; Yao, K. Blind Channel Estimation for STBC Systems Using Higher-Order
Statistics. IEEE Trans. Wirel. Commun. 2011, 10, 495–505. [CrossRef]
38. Cirpan, H.; Tsatsanis, M. Maximum likelihood blind channel estimation in the presence of Doppler shifts. IEEE Trans. Signal
Process. 1999, 47, 1559–1569. [CrossRef]
39. Necker, M.; Stuber, G. Totally blind channel estimation for OFDM over fast varying mobile channels. In Proceedings of the 2002
IEEE International Conference on Communications. Conference Proceedings. ICC 2002 (Cat. No.02CH37333), New York, NY,
USA, 28 April–2 May 2002; Volume 1, pp. 421–425. [CrossRef]
40. Wu, C.Y.; Huang, W.J.; Chung, W.H. Low-Complexity Semiblind Channel Estimation in Massive MU-MIMO Systems. IEEE
Trans. Wirel. Commun. 2017, 16, 6279–6290. [CrossRef]
41. Kawasaki, H.; Matsumura, T. Semi-Blind Channel Estimation by Subspace Method for Orthogonal Precoded OFDM Systems. In
Proceedings of the 2022 IEEE 33rd Annual International Symposium on Personal, Indoor and Mobile Radio Communications
(PIMRC), Kyoto, Japan, 12–15 September 2022; pp. 451–456. [CrossRef]
42. Moon, T. The expectation-maximization algorithm. IEEE Signal Process. Mag. 1996, 13, 47–60. [CrossRef]
43. Obradovic, D.; Na, C.; Scheiterer, R.L.; Szabo, A. EM-based semi-blind channel estimation method for MIMO-OFDM communica-
tion systems. Neurocomputing 2008, 71, 2388–2398. [CrossRef]
44. Nayebi, E.; Rao, B.D. Semi-blind Channel Estimation for Multiuser Massive MIMO Systems. IEEE Trans. Signal Process. 2018,
66, 540–553. [CrossRef]
45. O’Shea, T.; Hoydis, J. An Introduction to Deep Learning for the Physical Layer. IEEE Trans. Cogn. Commun. Netw. 2017, 3, 563–575.
[CrossRef]
46. Khalid, W.; Yu, H.; Ali, R.; Ullah, R. Advanced physical-layer technologies for beyond 5G wireless communication networks.
Sensors 2021, 21, 3197. [CrossRef]
47. Kim, W.; Ahn, Y.; Kim, J.; Shim, B. Towards deep learning-aided wireless channel estimation and channel state information
feedback for 6G. J. Commun. Netw. 2023, 25, 61–75. [CrossRef]
48. LeCun, Y.; Boser, B.; Denker, J.S.; Henderson, D.; Howard, R.E.; Hubbard, W.; Jackel, L.D. Backpropagation Applied to
Handwritten Zip Code Recognition. Neural Comput. 1989, 1, 541–551. [CrossRef]
49. Sak, H.; Senior, A.; Beaufays, F. Long short-term memory based recurrent neural network architectures for large vocabulary
speech recognition. arXiv 2014, arXiv:1402.1128. [CrossRef]
50. Mirza, M.; Osindero, S. Conditional generative adversarial nets. arXiv 2014, arXiv:1411.1784. [CrossRef]
51. Ma, X.; Gao, Z. Data-Driven Deep Learning to Design Pilot and Channel Estimator for Massive MIMO. IEEE Trans. Veh. Technol.
2020, 69, 5677–5682. [CrossRef]
52. Ge, L.; Guo, Y.; Zhang, Y.; Chen, G.; Wang, J.; Dai, B.; Li, M.; Jiang, T. Deep Neural Network Based Channel Estimation for
Massive MIMO-OFDM Systems With Imperfect Channel State Information. IEEE Syst. J. 2022, 16, 4675–4685. [CrossRef]
53. Zheng, X.; Lau, V.K.N. Online Deep Neural Networks for MmWave Massive MIMO Channel Estimation with Arbitrary Array
Geometry. IEEE Trans. Signal Process. 2021, 69, 2010–2025. [CrossRef]
54. Zhang, Y.; Wang, H.; Li, C.; Chen, X.; Meriaudeau, F. On the performance of deep neural network aided channel estimation for
underwater acoustic OFDM communications. Ocean Eng. 2022, 259, 111518. [CrossRef]
55. Li, X.; Han, Z.; Yu, H.; Yan, L.; Han, S. Deep Learning for OFDM Channel Estimation in Impulsive Noise Environments. Wirel.
Pers. Commun. 2022, 125, 2947–2964. [CrossRef]
56. Mehlführer, C.; Colom Ikuno, J.; Šimko, M.; Schwarz, S.; Wrulich, M.; Rupp, M. The Vienna LTE simulators-Enabling repro-
ducibility in wireless communications research. EURASIP J. Adv. Signal Process. 2011, 2011, 29. [CrossRef]
57. Soltani, M.; Pourahmadi, V.; Mirzaei, A.; Sheikhzadeh, H. Deep Learning-Based Channel Estimation. IEEE Commun. Lett. 2019,
23, 652–655. [CrossRef]
58. Li, L.; Chen, H.; Chang, H.H.; Liu, L. Deep Residual Learning Meets OFDM Channel Estimation. IEEE Wirel. Commun. Lett. 2020,
9, 615–618. [CrossRef]
59. Pradhan, A.; Das, S.; Dayalan, D. A Two-Stage CNN Based Channel Estimation for OFDM System. In Proceedings of the
2021 Advanced Communication Technologies and Signal Processing (ACTS), Rourkela, India, 15–17 December 2021; pp. 1–4.
[CrossRef]
60. Li, J.; Peng, Q. Lightweight Channel Estimation Networks for OFDM Systems. IEEE Wirel. Commun. Lett. 2022, 11, 2066–2070.
[CrossRef]
Electronics 2023, 12, 4965 37 of 38

61. Zhao, J.; Wu, Y.; Zhang, Q.; Liao, J. Two-Stage Channel Estimation for mmWave Massive MIMO Systems Based on ResNet-UNet.
IEEE Syst. J. 2023, 17, 4291–4300. [CrossRef]
62. Rahman, M.H.; Chowdhury, M.Z.; Utama, I.B.K.Y.; Jang, Y.M. Channel Estimation for Indoor Massive MIMO Visible Light
Communication with Deep Residual Convolutional Blind Denoising Network. IEEE Trans. Cogn. Commun. Netw. 2023, 9, 683–694.
[CrossRef]
63. Gizzini, A.K.; Chafii, M. Deep Learning Based Channel Estimation in High Mobility Communications Using Bi-RNN Networks.
arXiv 2023, arXiv:2305.00208. [CrossRef]
64. Ali, M.H.E.; Taha, I.B. Channel state information estimation for 5G wireless communication systems: Recurrent neural networks
approach. PeerJ Comput. Sci. 2021, 7, e682. [CrossRef]
65. Essai Ali, M.H.; Rabeh, M.L.; Hekal, S.; Abbas, A.N. Deep Learning Gated Recurrent Neural Network-Based Channel State
Estimator for OFDM Wireless Communication Systems. IEEE Access 2022, 10, 69312–69322. [CrossRef]
66. Helmy, I.; Tarafder, P.; Choi, W. LSTM-GRU Model-Based Channel Prediction for One-Bit Massive MIMO System. IEEE Trans.
Veh. Technol. 2023, 72, 11053–11057. [CrossRef]
67. Zhao, S.; Fang, Y.; Qiu, L. Deep Learning-Based channel estimation with SRGAN in OFDM Systems. In Proceedings of the
2021 IEEE Wireless Communications and Networking Conference (WCNC), Nanjing, China, 29 March–1 April 2021; pp. 1–6.
[CrossRef]
68. Dong, Y.; Wang, H.; Yao, Y.D. Channel Estimation for One-Bit Multiuser Massive MIMO Using Conditional GAN. IEEE Commun.
Lett. 2021, 25, 854–858. [CrossRef]
69. Zhang, Q.; Dong, H.; Zhao, J. Channel Estimation for High-Speed Railway Wireless Communications: A Generative Adversarial
Network Approach. Electronics 2023, 12, 1752. [CrossRef]
70. Kang, X.F.; Liu, Z.H.; Yao, M. Deep learning for joint pilot design and channel estimation in MIMO-OFDM systems. Sensors 2022,
22, 4188. [CrossRef]
71. Jiang, P.; Wen, C.K.; Jin, S.; Li, G.Y. Dual CNN-Based Channel Estimation for MIMO-OFDM Systems. IEEE Trans. Commun. 2021,
69, 5859–5872. [CrossRef]
72. Haq, S.A.U.; Gizzini, A.K.; Shrey, S.; Darak, S.J.; Saurabh, S.; Chafii, M. Deep Neural Network Augmented Wireless Channel
Estimation for Preamble-Based OFDM PHY on Zynq System on Chip. IEEE Trans. Very Large Scale Integr. (VLSI) Syst. 2023,
31, 1026–1038. [CrossRef]
73. Li, Y.; Bian, X.; Li, M. Denoising Generalization Performance of Channel Estimation in Multipath Time-Varying OFDM Systems.
Sensors 2023, 23, 3102. [CrossRef] [PubMed]
74. Hu, Z.; Chen, Y.; Han, C. PRINCE: A Pruned AMP Integrated Deep CNN Method for Efficient Channel Estimation of Millimeter-
wave and Terahertz Ultra-Massive MIMO Systems. IEEE Trans. Wirel. Commun. 2023, 22, 8066–8079. [CrossRef]
75. Li, Q.; Gong, Y.; Meng, F.; Li, Z.; Miao, L.; Xu, Z. Residual Learning based Channel Estimation for OTFS system. In Proceedings
of the 2022 IEEE/CIC International Conference on Communications in China (ICCC Workshops), Foshan, China, 11–13 August
2022; pp. 275–280. [CrossRef]
76. Tong, W.; Xu, W.; Wang, F.; Shang, J.; Pan, M.; Lin, J. Deep Learning Compressed Sensing-Based Beamspace Channel Estimation
in mmWave Massive MIMO Systems. IEEE Wirel. Commun. Lett. 2022, 11, 1935–1939. [CrossRef]
77. Nayir, H.; Karakoca, E.; Görçin, A.; Qaraqe, K. Hybrid-Field Channel Estimation for Massive MIMO Systems based on OMP
Cascaded Convolutional Autoencoder. In Proceedings of the 2022 IEEE 96th Vehicular Technology Conference (VTC2022-Fall),
London, UK, 26–29 September 2022; pp. 1–6. [CrossRef]
78. Donoho, D.L.; Maleki, A.; Montanari, A. Message passing algorithms for compressed sensing: I. motivation and construction. In
Proceedings of the 2010 IEEE Information Theory Workshop on Information Theory (ITW 2010, Cairo), Cairo, Egypt, 6–8 January
2010; pp. 1–5. [CrossRef]
79. Wei, X.; Hu, C.; Dai, L. Deep Learning for Beamspace Channel Estimation in Millimeter-Wave Massive MIMO Systems. IEEE
Trans. Commun. 2021, 69, 182–193. [CrossRef]
80. Pu, X.; Liu, Y.; Song, M.; Chen, Q. Orthogonal Time Frequency Space Channel Estimation Based on Model-driven Deep Learning.
J. Electron. Inf. Technol. 2023, 45, 1–8. [CrossRef]
81. Wang, P.; Li, J.; Liu, X.; Wang, P. Multi-Stage Training Optimization for Pilot Compression and Channel Estimation in Massive
MIMO Systems Under Quasi-Sparse Channel Environment. IEEE Commun. Lett. 2022, 26, 3059–3063. [CrossRef]
82. Yuan, X.; Zhang, Y.J.A.; Shi, Y.; Yan, W.; Liu, H. Reconfigurable-Intelligent-Surface Empowered Wireless Communications:
Challenges and Opportunities. IEEE Wirel. Commun. 2021, 28, 136–143. [CrossRef]
83. Tang, W.; Chen, X.; Chen, M.Z.; Dai, J.Y.; Han, Y.; Jin, S.; Cheng, Q.; Li, G.Y.; Cui, T.J. On Channel Reciprocity in Reconfigurable
Intelligent Surface Assisted Wireless Networks. IEEE Wirel. Commun. 2021, 28, 94–101. [CrossRef]
84. Mao, Z.; Liu, X.; Peng, M. Channel Estimation for Intelligent Reflecting Surface Assisted Massive MIMO Systems—A Deep
Learning Approach. IEEE Commun. Lett. 2022, 26, 798–802. [CrossRef]
85. Xie, W.; Xiao, J.; Zhu, P.; Yu, C.; Yang, L. Deep Compressed Sensing-Based Cascaded Channel Estimation for RIS-Aided
Communication Systems. IEEE Wirel. Commun. Lett. 2022, 11, 846–850. [CrossRef]
86. Liu, M.; Li, X.; Ning, B.; Huang, C.; Sun, S.; Yuen, C. Deep Learning-Based Channel Estimation for Double-RIS Aided Massive
MIMO System. IEEE Wirel. Commun. Lett. 2023, 12, 70–74. [CrossRef]
Electronics 2023, 12, 4965 38 of 38

87. Liu, S.; Gao, Z.; Zhang, J.; Renzo, M.D.; Alouini, M.S. Deep Denoising Neural Network Assisted Compressive Channel Estimation
for mmWave Intelligent Reflecting Surfaces. IEEE Trans. Veh. Technol. 2020, 69, 9223–9228. [CrossRef]
88. Shtaiwi, E.; Zhang, H.; Abdelhadi, A.; Han, Z. RIS-Assisted mmWave Channel Estimation Using Convolutional Neural Networks.
In Proceedings of the 2021 IEEE Wireless Communications and Networking Conference Workshops (WCNCW), Nanjing, China,
29 March 2021; pp. 1–6. [CrossRef]
89. Jin, Y.; Zhang, J.; Zhang, X.; Xiao, H.; Ai, B.; Ng, D.W.K. Channel Estimation for Semi-Passive Reconfigurable Intelligent Surfaces
With Enhanced Deep Residual Networks. IEEE Trans. Veh. Technol. 2021, 70, 11083–11088. [CrossRef]
90. Feng, H.; Zhao, Y. mmWave RIS-Assisted SIMO Channel Estimation Based on Global Attention Residual Network. IEEE Wirel.
Commun. Lett. 2023, 12, 1179–1183. [CrossRef]
91. Abdallah, A.; Celik, A.; Mansour, M.M.; Eltawil, A.M. RIS-Aided mmWave MIMO Channel Estimation Using Deep Learning and
Compressive Sensing. IEEE Trans. Wirel. Commun. 2023, 22, 3503–3521. [CrossRef]
92. Ginige, N.; Shashika Manosha, K.B.; Rajatheva, N.; Latva-aho, M. Untrained DNN for Channel Estimation of RIS-Assisted
Multi-User OFDM System with Hardware Impairments. In Proceedings of the 2021 IEEE 32nd Annual International Symposium
on Personal, Indoor and Mobile Radio Communications (PIMRC), Helsinki, Finland, 13–16 September 2021; pp. 561–566.
[CrossRef]
93. Liu, C.; Liu, X.; Ng, D.W.K.; Yuan, J. Deep Residual Learning for Channel Estimation in Intelligent Reflecting Surface-Assisted
Multi-User Communications. IEEE Trans. Wirel. Commun. 2022, 21, 898–912. [CrossRef]
94. Shen, W.; Qin, Z.; Nallanathan, A. Deep Learning for Super-Resolution Channel Estimation in Reconfigurable Intelligent Surface
Aided Systems. IEEE Trans. Commun. 2023, 71, 1491–1503. [CrossRef]
95. Li, T.; Yang, Y.; Lee, J.; Qin, X.; Huang, J.; He, G. DCSaNet: Dilated Convolution and Self-Attention-Based Neural Network for
Channel Estimation in IRS-Aided Multi-User Communication System. IEEE Wirel. Commun. Lett. 2023, 12, 1139–1143. [CrossRef]
96. Dai, L.; Jiao, R.; Adachi, F.; Poor, H.V.; Hanzo, L. Deep Learning for Wireless Communications: An Emerging Interdisciplinary
Paradigm. IEEE Wirel. Commun. 2020, 27, 133–139. [CrossRef]
97. Qing, C.; Dong, L.; Wang, L.; Ling, G.; Wang, J. Transfer learning-based channel estimation in orthogonal frequency division
multiplexing systems using data-nulling superimposed pilots. PLoS ONE 2022, 17, e0268952. [CrossRef]
98. Kim, D.; Park, S.; Kang, J.; Kang, J. Block-Fading Non-Stationary Channel Estimation for MIMO-OFDM Systems via Meta-
Learning. IEEE Commun. Lett. 2022, 26, 2924–2928. [CrossRef]
99. Gao, W.; Zhang, W.; Liu, L.; Yang, M. Deep Residual Learning with Attention Mechanism for OFDM Channel Estimation. IEEE
Wirel. Commun. Lett. 2022, early access. [CrossRef]
100. Niu, Z.; Zhong, G.; Yu, H. A review on the attention mechanism of deep learning. Neurocomputing 2021, 452, 48–62. [CrossRef]
101. Serghiou, D.; Khalily, M.; Brown, T.W.C.; Tafazolli, R. Terahertz Channel Propagation Phenomena, Measurement Techniques
and Modeling for 6G Wireless Communication Applications: A Survey, Open Challenges and Future Research Directions. IEEE
Commun. Surv. Tutor. 2022, 24, 1957–1996. [CrossRef]
102. Sejan, M.A.S.; Rahman, M.H.; Aziz, M.A.; Kim, D.S.; You, Y.H.; Song, H.K. A Comprehensive Survey on MIMO Visible Light
Communication: Current Research, Machine Learning and Future Trends. Sensors 2023, 23, 739. [CrossRef]

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.