Channel Estimation in MIMO OFDM Systems with Tapped Delay Line Model

International Journal of Computer Networks & Communications (IJCNC) Vol.15, No.6, November 2023
DOI: 10.5121/ijcnc.2023.15605 97
CHANNEL ESTIMATION IN MIMO OFDM SYSTEMS
WITH TAPPED DELAY LINE MODEL
Ravi Hosamani1
, Yerriswamy T2
1
Department of Electronics and Communication Engineering,K.L.E . Institute of
Technology, Hubballi, Karnataka, India 580027
2
Department of Computer Science and Engineering. K.L.E. Institute of Technology
Hubballi, Karnataka, India 580027
ABSTRACT
The continuous increase in the user demands fornew-generation communication systems, is making the
wireless channel more complex and challenging for estimation, developing a simulation model for the
channel,and evaluating the performance of different MIMO systems. In this work, a simulation model
for multipath fading channels in wireless communication is performed. The model includes a selection
of typical Tapped-Delay-Line channel models that can be implemented to reproduce the effects of
representative channel distortion and interference. Based on the simulation results, the proposed method
exhibits accurate channel estimation performance for frequency-selective fading channels. The proposed
work employed LS, MMSE, and ML methods for channel estimation, using 16 and 32 pilots and fixed pilot
locations in each frame. Results are obtained for 4x4, 8x8, 16x16, 16x8, and 16x4 MIMO systems and
tapped delay line systems.
KEYWORDS
Channel estimation, MIMO systems, multipath fading, tapped-delay line, Mean square error,
1. INTRODUCTION
In the generation of mobile networks, the latest standard for wireless communication have
evolved substantially over the years. Current and future networks is to provide improved data
transfer speeds, reduced latency, and increased capacity compared to the preceding generations of
mobile networks[1]. These features are achieved through various advanced technologies like,
Multiple antenna technology, network slicing to deliver faster and more reliable connectivity [2]
mm wave frequencies. 5G networks are supposed to play a pivotal role in industrial internet of
things (IIoT) which supports continuous connectivity to large number of devices with minimum
latency and higher reliability. The 5G technology in combination with Industry 4.0 can bring a
huge transformation across industries, enabling better productivity, efficiency and advanced
automation [3, 4]. This makes a highly precise and efficient manufacturing as well as
constructive monitoring and control of equipment and systems [5].
The attractive features of 5G, comes with various challenges while implementing them. These
challenges are present at various levels or stages of communication systems. At physical layer,
modulation schemes at higher frequencies and impact of channel state on the data transmitted
using these techniques are the main thing which need to consider [6]. Various researchers are
exploring the effect of these modulation schemes and channel behavior. Based on the
observations, it is found that if channel state/behavior is previously known for certain scenario,

98
effective communication can be achieved. In these systems, advanced techniques such as
beamforming, massive MIMO, and network slicing heavily rely on accurate channel knowledge
in order to achieve high spectral efficiency and reliable communication [7].
Channel estimation is the process of estimating the characteristics of a communication channel
between a two points [8]. In wireless communication systems, channel estimation is essential to
optimize the transmission parameters and improve the quality of the communication [9].Channel
estimation can be done using various techniques, including pilot-based methods, blind estimation,
and channel prediction. Pilot-based methods involve transmitting pilot symbols, along with the
data symbols. These are used to estimate the channel behavior which includes the frequency
response, phase, and delay spread [10]. This information can then be used to optimize the
transmission parameters, such as the modulation scheme, coding rate, and transmit power. Blind
estimation, on the other hand, involves estimating the channel characteristics without using any
known reference signals. This can be done using statistical techniques, such as maximum
likelihood estimation or least squares estimation, based on the received data symbols [11].
The motivation for the work is the need for accurate channel estimation in wireless
communication systems. As the user demands for new communication systems, wireless channels
become more complex and challenging for estimation. The objective of the work is to develop a
simulation model for multipath fading channels in wireless communication that can accurately
estimate the channel parameters. The contribution of the work is the proposed simulation model
that includes typical Tapped-Delay-Line channel models, can be used to reproduce the effects of
channel distortion and interference. The simulation results show that the proposed system can
accurately estimate channel parameters for frequency-selective fading channels using LS,
MMSE, and ML methods. Additionally, tested the system using various MIMO systems and
tapped delay line systems to show its flexibility and applicability.
The article is organized into six chapters. The first chapter provides an introduction to the topic
and the significance of channel estimation in wireless communication systems. The second
chapter discusses the relevant literature on channel estimation and classical techniques that have
been developed over the years. The third chapter presents the proposed system and channel
models, including a discussion of the channel impulse response and its computation. The fourth
chapter focuses on the implementation of the Channel Estimator (CE), discussing the various
classical techniques used in channel estimation. The fifth chapter presents the simulation results
and analysis of the proposed system, including a discussion of their implications. The sixth
chapter provides a summary of the key findings of the study, limitations of the proposed system,
and recommendations for future research. Finally, the reference section contains a list of all the
sources cited in the article.
2. RELATED WORK
Y. Song et al. has proposed a channel estimation algorithm for MIMO-OFDM systems in time-
varying channels. The algorithm uses a time-domain training sequence that considers correlation
between antennas and subcarriers into account. Channel estimation is achieved through
maximum likelihood estimator and Kalman filter to track the time-varying channels. The method
is tested with existing algorithms, namely, the LS and the MMSE estimator. Results show that the
proposed algorithm outperforms the existing algorithms in terms of mean square error and bit
error rate. [12].
Y. Wang et al. proposed a low-complexity channel estimation algorithm for MIMO-OFDM
systems. The algorithm utilizes the inherent sparsity of the channel impulse response (CIR) to
reduce the number of pilot symbols required for accurate channel estimation. The algorithm

99
consists of two stages: 1) sparse CIR estimation using a Bayesian compressive sensing algorithm
and 2) CIR interpolation using a low-complexity Kalman filter. Results reveal that the proposed
algorithm outperforms the conventional LS and LMMSE channel estimation methods in terms of
both mean squared error (MSE) and bit error rate (BER) under various channel conditions.
Moreover, the proposed algorithm is shown to achieve comparable performance to the
computationally intensive AMP algorithm while requiring significantly lower complexity [13].
Z. Shen et al., proposed a new channel estimation method for MIMO-OFDM systems in time-
varying frequency-selective channels. The method uses a combination of a deep neural network
(DNN) and a channel prediction algorithm to estimate the channel. The DNN is used to extract
the channel features from the pilot signals, and the channel prediction algorithm predicts the
channel in the data transmission phase. The method was compared to several existing methods,
including the least squares (LS) method and the minimum mean square error (MMSE) method.
The study concludes that the proposed method is a promising solution for channel estimation in
MIMO-OFDM systems in time-varying frequency-selective channels [14].
Y. Cai et al., proposed a joint pilot decontamination and channel estimation algorithm for
massive multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing
(OFDM) systems in frequency-selective fading channels. The algorithm uses a low-rank matrix
completion (LRMC) method to eliminate the effect of pilot contamination and a sparse Bayesian
learning (SBL) algorithm to estimate the frequency-selective channel [15].
Rezaeiet.al., proposed a novel low-overhead channel estimation technique for frequency-selective
multiple-input multiple-output (MIMO) systems. The method employs a differential evolution
algorithm (DEA) to optimize the pilot pattern and estimate the channel response. The DEA-based
approach can efficiently estimate the channel in time-varying environments and alleviate pilot
contamination issues in MIMO systems. The method achieves high estimation accuracy with a
lower pilot overhead than conventional techniques, making it more efficient for practical
applications [16].
Ashraf et al. proposed a new frequency-selective channel estimation technique for 5G
communication systems. The method utilizes a sparse Bayesian learning (SBL) algorithm to
estimate the channel response in frequency-selective environments. The method is evaluated
using numerical simulations, and the results show that the SBL-based approach can achieve
higher accuracy in estimating the channel response compared to other existing methods. Also, the
method reduces the pilot overhead and enhance the spectral efficiency of the 5G system. The
authors also provide a comprehensive analysis of the performance of the proposed method under
different scenarios, such as different channel conditions, Doppler frequencies, and channel
coherence times [17].
H. Zhou, et.al., proposed a low-complexity channel estimation algorithm for OFDM-based
cooperative communication systems. The algorithm is based on the pilot-aided least squares
(PALS) algorithm, which uses the pilot symbols transmitted by the source and relay nodes to
estimate the channel frequency response. The algorithm uses a low-complexity scheme to reduce
the number of required pilot symbols, which is achieved by dividing the pilot symbols into
groups and using them in a cyclic manner. The algorithm provides a practical solution for low-
complexity channel estimation in OFDM-based cooperative communication systems, which can
effectively reduce the overhead of pilot symbols and improve the system performance [27].
3. SYSTEM AND CHANNEL MODELS
Wireless systems performance evaluation typically involves modeling or emulating the physical
radio channel. The enormous evolution is seen in channel models due to increase complexity and

100
various amendments and user demands in communication systems. The Channel Impulse
Response (CIR) contains crucial details about a channel that enable the examination of any
wireless transmission that traverses it. This is because, under certain circumstances, the radio
channel can be modeled as a linear filter. Eq (1) [18] can be employed to express the time-
varying impulse response of the multipath channel in baseband. This involves computing the
amplitudes and time delays of multiple waves that arrive at distinct intervals and adding them up
to derive the CIR.
ℎ(𝑡, 𝜏) = ∑ 𝑐𝑖𝑒𝑗2𝜋.𝑓𝑑𝑖𝑡
𝑁−1
𝑖=0
. 𝛿(𝜏
− 𝜏𝑖) (1)
At the receiver, every path i has an amplitude of ci, a delay of τi, and a Doppler shift of fdi. The
Doppler shift arises due to the motion of the mobile station and environmental changes, causing a
change in the frequency of the radio signal. The amount of Doppler shift is proportional to both
mobile speed and carrier frequency. The difference between the maximum and minimum
frequency shifts resulting from multipath is referred to as frequency spread. Depending on the
statistical distribution of the angle of arrival in each path. The CIR can be expressed using simple
Tapped-Delay-Line (TDL) or Cluster-Delay-Line (CDL) models [19]. More advanced models
and the latest trends in channel models fund in [20-23].
The TDL (tap delay line) model characterizes the wireless channel between a transmitter and
receiver by using statistical parameters, rather than incorporating the physical layout of the
environment. This model represents the time-domain CIR as a series of discrete taps, where each
tap has its own time-varying amplitude, delay, and coefficient values, as shown in eq(2):
ℎ(𝑡, 𝜏) = ∑ 𝑐𝑖(𝑡)
𝑁−1
𝑖=0
. 𝛿(𝜏 − 𝜏𝑖) (2)
The tap in the wireless channel is modeled as a Dirac delta function, and the overall time-varying
channel impulse response h (t, τ) is represented as a sum of delayed taps. The model assumes that
there are N paths in the channel, where ci(t) is a complex amplitude coefficient that varies with
time, δ is the Dirac delta function, and τi represents the delay of the ith path. Each tap is
associated with a Doppler spectrum that determines how the coefficients change over time.
The proposed model defines the channel impulse response as a finite sum of taps, which is
restricted by a maximum of N paths. The separation between two adjacent paths is dependent on
the system's bandwidth, and if the coefficients ci are stable, eq (3) can be utilized to express the
output signal of the channel. The block diagram of a tapped-delay channel model as shown in
Figure.1 illustrates components signal flow. The input signal represents the transmitted signal in
the wireless communication system. It can be a continuous-time analog signal or a discrete-time
digital signal. Taps represent the individual delay paths in the channel model. Each tap represents
a different propagation path or reflection, and it has associated attenuation and phase shift
parameters. The number of taps corresponds to the number of multipath components or echoes
considered in the model. Each tap is followed by a delay element that introduces a specific time
delay corresponding to the propagation delay of the associated path. The delays account for the
differences in arrival times of the multipath components. Attenuation elements are placed after
the delay elements to scale the signal strength of each tap. The attenuation factor represents the
loss or attenuation experienced by the signal in the specific propagation path. Phase shift
elements introduce the phase shifts experienced by the signal due to the propagation path. These
phase shifts are caused by differences in path lengths and reflections. The adders combine the

101
delayed and scaled tap signals to produce the overall channel output. The channel output
represents the received signal after passing through the tapped-delay channel model. It
incorporates the effects of multipath propagation, including delays, attenuations, and phase shifts.
𝑆(𝑡) = ∑ 𝑐𝑖(𝑡)
𝑁−1
𝑖=0
. 𝑠(𝜏 − 𝜏𝑖) + 𝑛(𝑡) (3)
Figure 1. Block diagram of a tapped-delay channel model.
However, not all of the models are suitable for real-time physical emulation and to ensure
accurate and reliable emulation, it is necessary to have specific information about the channel.
Therefore, in this study, we have chosen a TDL-based model that provides a detailed description
of the channel in each scenario. Consider the MIMO channel that experiences frequency-selective
block fading. The channel has L + 1 tap discrete-time channel impulse response (CIR) denoted as
𝐻 = 𝐻1, 𝐻2 − −, 𝐻𝐿−1, 𝐻𝐿 and it involves 𝑁𝑇 transmitters and 𝑁𝑅 receivers.
A block diagram of a MIMO (Multiple-Input Multiple-Output)OFDM (Orthogonal Frequency
Division Multiplexing) system represents the structure and signal flow of a communication
system that utilizes multiple antennas at both the transmitter and receiver, combined with OFDM
modulation as shown in Figure. 2 the Source Signals: Multiple source signals or data streams are
generated from different sources or data inputs. The source signals are then processed by a
MIMO encoder, which combines them with appropriate weights and spatial multiplexing
techniques to exploit the spatial diversity provided by multiple antennas at the transmitter [35].
The processed signals are then modulated using OFDM, which converts the data streams into
parallel subcarriers. An Inverse Fast Fourier Transform (IFFT) is applied to each subcarrier to
convert the frequency domain symbols into time-domain OFDM symbols. A cyclic prefix (CP) is
added to each OFDM symbol to mitigate the effects of multipath interference. : Known pilot
symbols are inserted into the transmitted OFDM symbols at regular intervals. The transmitted
OFDM symbols propagate through a wireless channel, which introduces fading, multipath
effects, and noise.
In a MIMO system, the wireless channel consists of multiple paths with different gains and
phases for each transmit-receive antenna pair [35]. At the receiver, the CP is removed from each
OFDM symbol.The received pilot symbols are used to estimate the channel conditions, including
channel gains, phases, and other parameters. The estimated channel information is interpolated or
extrapolated to obtain channel estimates for the entire OFDM symbol.The estimated channel

102
parameters are used to construct the channel matrix, which describes the relationship between
transmitted and received signals in a MIMO system. A Fast Fourier Transform (FFT) is applied
to each OFDM symbol to recover the frequency domain representation. The received signals
from multiple receive antennas are processed by a MIMO decoder, which separates and extracts
the transmitted data streams using spatial processing techniques. Each separated data stream may
undergo channel decoding to correct errors introduced during transmission. The channel-decoded
signals are further processed to recover the original source signals. The recovered source signals
represent the reconstructed data streams at the receiver's output[36].
Figure 2. MIMO OFDM System
In a MIMO OFDM system, the data (X) is transmitted simultaneously over multiple subcarriers
using orthogonal frequency division multiplexing. The subcarriers are orthogonal to each other,
which reduces the interference between them and allows more data to be transmitted over the
same bandwidth.
Consider that there are NT transmit antennas and NR receive antennas as shown in fig 2, the
channel matrix can be represented as eq (4). Each element Hl= hijrepresents the complex channel
gain between the ith
transmit antenna and the jth
receive antenna.
𝐻𝑙 = [
ℎ1,1(𝑙) ⋯ ℎ1,𝑁𝑇
(𝑙)
⋮ ⋱ ⋮
ℎ𝑁𝑅,1(𝑙) ⋯ ℎ𝑁𝑅,𝑁𝑇
(𝑙)
] , 𝑙
∈ [0, 𝐿] (4)
Channel estimation is necessary to estimate the frequency response of the multiple channels
between each transmit and receive antenna pair. The channel estimation process in MIMO
OFDM systems. Pilot symbols are transmitted from each transmit antenna to each receive
antenna.The pilot symbols are usually arranged in a grid pattern, called a pilot matrix that spans
the frequency-time grid of the OFDM symbols. The cyclic prefix 𝐶𝑃𝑛𝑡 is used to separate the
data and training symbols in the training sequence transmitted by the 𝑛𝑖
𝑡ℎ
antenna. The length of
𝐶𝑃𝑛𝑡 is set to L, and it is defined as [𝑆𝑛𝑡(𝑁𝑠 − 𝐿), . . . , 𝑆𝑛𝑡(𝑁𝑠 − 1)], where 𝑆𝑛𝑡represents the
transmitted signal and 𝑁𝑠is the total number of samples in the signal. The received training signal
on 𝑁𝑅 receive antennas. The pilot symbols are designed to be orthogonal to each other and to
the data symbols, so that they can be easily distinguished at the receiver as shown in Fig 2.
The received pilot symbols are used to estimate the channel frequency response for each
transmit-receive antenna pair. There are several methods for channel estimation in MIMO OFDM

103
systems. The vector [𝑦1(𝑛),𝑦2(𝑛) − − − − − −𝑦𝑁𝑅
(𝑛)]′represents the signals received by 𝑁𝑅
receive antennas at time n, while the complex noise matrix is denoted as E and defined as 𝐸 =
[𝑒(0), 𝑒(1),· · · , 𝑒(𝑁𝑠 − 1)], where e(n) is the vector of additive noise components for each
receive antenna, i.e 𝑒(𝑛) = [𝑒1(𝑛),𝑒2(𝑛), − − −𝑒𝑁𝑅
(𝑛) ]′
It is important to note that 𝑆𝑛𝑡
(𝑛) represents the pilot symbol transmitted by the 𝑛𝑡ℎ
transmit
antenna at time n, 𝑦𝑛𝑟
(𝑛) represents the signal received by the 𝑛𝑡ℎ
receive antenna at time n,
𝑒𝑛𝑟
(n) represents the additive noise component in 𝑦𝑛𝑟
(𝑛)and the expected signal-to-noise ratio
(SNR) on each receive antenna is calculated for n ∈ [0, Ns−1]. The use of multiple antennas at
the transmitter and receiver sides allows the system to use spatial diversity, which means that the
signals transmitted from different antennas can take different paths to the receiver, reducing the
effects of fading and improving the reliability of the system.
4. CHANNEL ESTIMATION (CE)
This section discusses the implementation of the Channel Estimator (CE), a fundamental
component in wireless communication systems that aids in the recovery of transmitted data by
estimating the state of the channel through which the data travels. Effective implementation of
the CE can significantly improve the reliability and accuracy of the received signals, enabling
better performance of the communication system as a whole.
Channel estimation is essential for reliable communication and plays a critical role in the
performance of the overall communication system. Classical channel estimation techniques have
been extensively researched and developed over the years, and this section will discuss some of
these classical techniques in detail. These techniques include Least Squares (LS), Maximum
Likelihood (ML), Minimum Mean Square Error (MMSE), Understanding these techniques is
vital for effectively implementing channel estimation in wireless communication systems.
LS (Least Squares) estimation is a method for estimating the unknown parameters of a
mathematical model by minimizing the sum of the squares of the differences between the
predicted values of the model and the actual observations [24]. It is a popular method in many
fields, including statistics, signal processing, and machine learning. In the context of wireless
communication, LS estimation can be used to estimate the channel response in a communication
system [25]. The LS estimate of the channel response is obtained by minimizing the sum of the
squared errors between the received signal and the predicted signal, using the estimated channel
response [26]. Specifically, the LS estimate of the channel response eq (5)
𝐻𝑙𝑠 = 𝑌𝑋′
𝑋𝑋′ (5)
⁄
𝐻𝑙𝑠is the LS estimate of the channel response, Y is the received signal, X is the transmitted signal
(known training symbols), and ' denotes the conjugate transpose. The LS estimate is simple to
compute and is commonly used in many communication systems, although it is sensitive to noise
and may not be optimal in some situations channel estimation and equalization using the least
squares error (LSE) method followed by zero-forcing (ZF) equalization[27].
The input variables to the are: The input are received symbols, pilot which is a pilot sequence,
positions of the pilot subcarriers, the number of transmit and receive antennas, the number of
subcarriers, number of channel taps, the noise variance. The input variables to the are: estimated
channel impulse response, estimated frequency response of the channel, then performs
equalization by looping over each subcarrier and calculating the inverse of the estimated channel

104
frequency response using the Moore-Penrose pseudoinverse of a matrix. The resulting equalized
symbols are stored in the variable.
MMSE (Minimum Mean Square Error) estimation is a method for estimating the unknown
parameters of a mathematical model by minimizing the expected value of the squared difference
between the predicted values of the model and the actual observations [28]. It is a popular method
in many fields, including signal processing, control theory, and machine learning. MMSE
estimation can be used to estimate the channel response in a communication system [29].The
MMSE estimate of the channel response is obtained by minimizing the expected value of the
squared error between the received signal and the predicted signal, using the estimated channel
response [30]. Specifically, the MMSE estimate of the channel response is given byeq (6):
𝐻𝑚𝑚𝑠𝑒 = 𝑅𝑥𝑥𝑆𝑛𝑡
(𝑛)′
+ sigma_n2
∗ eye(N𝑡)) (6)
Where 𝐻𝑚𝑚𝑠𝑒the MMSE estimate of the channel response is, 𝑅𝑥𝑥 is the autocorrelation matrix of
the known training symbols𝑆𝑛𝑡
(𝑛), sigma 𝑛2
is the noise variance, and eye(𝑁𝑡) is the identity
matrix of size 𝑁𝑡𝑥𝑁𝑡 in eq (4). The MMSE estimate takes into account the noise in the received
signal and is generally more robust to noise than the LS estimate [31].
The MMSE method offers advantages over LS and ML methods by providing improved
performance in noise, low-SNR scenarios, multipath fading environments, and achieving an
optimal trade-off between bias and variance in channel estimation.
Implement channel estimation and equalization using the minimum mean squared error (MMSE)
algorithm. Input are same as used for LS estimation, the output is the estimated channel is the
frequency response of the estimated channel. The equalization with MMSE is performed in the
following loop by computing the pseudo-inverse of the matrix H' * H + N0 * eye(Nt) ,the
estimated channel frequency response for each subcarrier, and N0 is the noise variance. The
result is multiplied by the received symbols to obtain the equalized symbols. The mean squared
error (MSE) between the equalized symbols and the original transmitted symbols is computed
and stored.
Maximum likelihood (ML) based channel estimation is a method used in communication systems
to estimate the channel coefficients of a communication channel. In this method, the receiver uses
the maximum likelihood principle to estimate the channel coefficients based on the received
signal [32]. The ML-based channel estimation assumes that the transmitted signal is known, and
the receiver measures the received signal at different time intervals. The received signal is
modelled as a linear combination of the transmitted signal and the channel coefficients, with
additive noise. The goal of the ML-based channel estimation is to estimate the channel
coefficients that maximize the likelihood function of the received signal [33].
To perform ML-based channel estimation, the receiver first constructs a likelihood function that
describes the probability of the received signal given the channel coefficients. The likelihood
function is usually expressed in terms of the channel coefficients and the received signal. The
receiver then computes the maximum of the likelihood function to obtain the estimated channel
coefficients implement channel estimation and equalization using the maximum likelihood (ML)
algorithm [34]. Input is same as used for LS estimation. It initializes and loops over all
subcarriers. For each subcarrier, it extracts the estimated channel coefficients H, and the received
symbols and computes the ML estimate of the transmitted symbol for that subcarrier using the
minimum distance criterion. The ML estimate is stored in the corresponding variable. Finally, it

105
computes the mean squared error (MSE) between the estimated symbols and the transmitted
symbols, and stores the result in the variable.
5. SIMULATION AND DISCUSSION
In the simulation 2 and 4 taps in each sub channel ℎ𝑛𝑡𝑛𝑟
𝑁𝑇= 4,8,16, 𝑁𝑅 = 4, 8, and N = 64,128
for the, additive noise, we assume elements of E are white complex Gaussian with unit variance.
For the underlying fading channel, Elements of H are independent complex Gaussian, These
parameters help to identify the performance of channel estimation with respect to SNR as shown
in table I.
Table 1. Simulation parameters
Parameter Value
FFT points 64,128
Cyclic prefix 32, 64
Nt 4,8,16
Nr 4,8,16
SNR 0-30 dB
Pilot in each frame 4
Modulation QPSK
No. of taps 2,4
The result compares channel estimation using the LS and MMSE techniques, employing 16 pilots
for channel estimation in each frame for MIMO 4x4. According to the results presented in
Figure.3, the LS method performs poorly at lower SNRs, at SNR=10, MSE of LS=2 while the
MMSE=0.5 method exhibits superior performance, the ML method demonstrates MSE as 1.3.
Figure 3. CE of MIMO (4x4), tap=2, pilots=16
The channel estimation in MIMO 8x8 systems, with pilots=32 is shown in Figure.4. The LS
method performs worse at lower SNRs, at 10dB its power of MSE=0.5 while the MMSE method

106
outperforms as MSE is 1.3. Furthermore, the ML method displays nearly constant MSE it is
around 1.8.
Figure.5 shows the channel estimation in MIMO 16x4 systems with pilots=32. The LS method
and the MMSE perform almost similar at SNR=10dB, it is power of MSE is .084 and 0.82
respectively. In addition, the ML method demonstrates almost constant MSE of 1.3 to 1.5 for
different SNR
.

107
In Figure.6, presents channel estimation of MIMO 16x8 systems with 32 pilots for channel
estimation in each frame. The results demonstrate that the LS and MMSE methods perform
similarly at higher SNR at 20dB, exhibiting little difference in performance. Additionally, the
ML method shows nearly constant MSEpower around 1.8 To 1.9 MSE.
In this study, we evaluated the performance of LS and MMSE techniques for channel estimation
in MIMO 16x16 systems, using pilots=32 with tap=2 for channel estimation in each frame. Our
analysis, as illustrated in Figure.7, indicates that the LS method underperforms at lower SNRs
compared to the MMSE method at SNR =10dB, MSE power is 8 and 3, the ML method shows
consistent MSE throughout the evaluation 1.9 of MSE power.

108
The purpose of our investigation was to compare the effectiveness of LS and MMSE techniques
in channel estimation MIMO 16x4 systems, utilizing 32 pilots and a tap of 4 for channel
estimation in each frame. Our findings, illustrated in fig. 8, demonstrate that the LS method
performs less optimally than the MMSE method at lower SNRs at SNR=10dB, power of MSE is
1.2 and 1.5 respectively. Which indicated increase number of tap effect on MSE which can
analysed from Figure.
Figure .9. CE of MIMO (16x8), tap=4, pilots=32

109
LS and MMSE techniques in channel estimation and detection for MIMO 16x4 systems,
utilizing 32 pilots and a tap of 4 for channel estimation in each frame. Based on our results
depicted in Figure. 9. Increase in tap in the systems, results indicates that the LS method performs
sub-optimally at lower SNRs in comparison to the MMSE method at SNR=10dB, which
demonstrates superior performance. Furthermore, the ML method maintains a consistent MSE
throughout the evaluation as in Figure.9 and Figure.6.
The performance of LS and MMSE techniques for channel estimation and detection in MIMO
16x16 systems, using 32 pilots and a 4-tap channel estimation per frame, was evaluated. The
results, presented in Figure.10 and Figure.7 indicate that the MMSE method outperforms the LS
method at lower signal-to-noise ratios (SNRs) for increased tap numbers. Additionally, the ML
method maintains a consistent mean squared error (MSE) across the evaluation of with different
taps.
Figure 11. MSE Vs MIMO (Nt, Nr), tap=2, pilots=32

110
Figure.11 depicts the MSE plotted against different antenna configurations in MIMO systems
with 2-channel taps and 128 pilots, with a fixed SNR of 15dB. The results indicate that the LS
method yields a higher MSE for the 16x16 MIMO system, whereas the MMSE method exhibits a
lower MSE for the 16x8 MIMO system. In contrast, the ML method performs consistently across
various MIMO sizes.
Figure 12. MSE Vs MIMO (Nt, Nr), tap=4, pilots=32
The MSE is plotted against different antenna configurations in MIMO systems with 4-channel
taps and 32 pilots in Fig. 12, with a fixed SNR of 15dB. The outcomes show that the LS method
results in a lower MSE for the 16x16 MIMO system than in Figure. 11. Conversely, the MMSE
method yields a lower MSE for the 16x8 MIMO system compared to Figure.11. However, the
ML method performs uniformly across various MIMO sizes and taps. From these results, we can
see that MMSE method performs better over LSE and ML methods for almost all the scenarios.
LSE method has higher MSE for lower SNRs, but as SNR increases it gives better performance
nearly same as MMSE method. The ML estimation technique aims to find the parameter values
that maximize the likelihood of the observed data given a specific model. In wireless
communication systems, ML estimation takes into account the statistical characteristics of the
received signals and noise to estimate the channel parameters accurately. ML estimation is
known to achieve the Cramer-Rao Lower Bound (CRLB) for parameter estimation, which
represents the minimum variance attainable for unbiased estimators. It achieves the lowest
possible estimation error in large sample sizes. ML estimation is optimal when the noise follows
a Gaussian distribution, which is a common assumption in many wireless communication
systems. In such cases, ML estimation provides the best estimate of the true channel parameters.
ML method can be used in low SNR regimes or the scenarios where we have a fixed error
tolerance as in all scenarios we receive nearly constant MSE in ML method.
6. CONCLUSION
The continuously increasing demand in communication network is enforcing adaptive channel
estimation techniques in current generation communication systems. In this paper, we present a
simulation performance of communication channel for multipath fading scenarios in wireless
communication. The model includes Tapped-Delay-Line channel models to reproduce the effects
of channel distortion and interference. Based on the simulation results, it can be concluded that
the proposed system is effective in accurately estimating the channel parameters that are
subjected to frequency-selective fading. This work employs LS, MMSE, and ML methods for
channel estimation and detection using 16, and 32 pilots and fixed pilots for channel estimation in
each frame. The MMSE method given better MSE performance in all the scenarios. The LS

111
method with specific configurations and taps exhibit small MSE for high-order antenna
performance. However, the MMSE and ML methods consistently outperform the LS method in
all configurations for both channel estimation in wireless communications system. The
simulation results obtained from our study provide valuable insights and serve as a foundation for
optimizing the performance of wireless communication systems in realistic channel
environments. They aid in system design, algorithm development, performance evaluation, and
optimization techniques, ultimately leading to improved system performance and efficiency so
proposed simulation model and methods can be used for evaluating and optimizing the
performance of wireless communication systems in realistic channel environments.
7. FUTURE WORK
This paper has examined classical estimation methods for different antenna array between
transmitter and receiver with different tap delay line. The proposed simulation model and
methods can indeed be applied to different types of wireless communication systems. While the
specific implementation may require some adjustments or customization to suit the characteristics
of the particular wireless system. However exact implementation and parameters may vary
depending on the specific system requirements, channel characteristics, and available resources.
Some adaptations or modifications may be necessary to ensure optimal performance in different
wireless communication scenarios. This study motivates the researchers to work on advanced
channel estimation techniques machine learning-based approaches and implementing the
proposed system in a real-world wireless communication system to identify and address practical
deployment challenges. These avenues for future research could help in advancing the field of
wireless communication and improving the performance of wireless communication systems
under various channel conditions.
CONFLICTS OF INTEREST
The authors declare no conflict of interest
ACKNOWLEDGMENT
The authors thank Visvesvaraya Technological University (VTU), Belagavi-590018, and K.L.E
Institute of Technology, Hubballi-580027, for providing a platform to carry out the research
work.
REFERENCE
[1] Aijaz, R. Gudipati, S. Muhaidat, and J. Rodriguez,(2017) "5G wireless networks: Opportunities and
challenges," Wireless Personal Communications, vol. 94, no. 2, pp. 795-823
[2] R. A. Shaikh, A. S. Sadiq, and S. Ahmed,(2019) "5G wireless networks: A comprehensive survey,"
IEEE Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2875-2902.
[3] P. Gupta and A. K. Singh,(2020) "5G-Enabled Industry 4.0: A Review" International Journal of
Engineering and Advanced Technology, vol. 9, no. 1, pp. 3535-3541,
A. Naseer, M. Hussain, and N. Javaid,(2021) "5G-Enabled Industry 4.0: A Survey on the Emerging
Trends and Research Challenges" IEEE Access, vol. 9, pp. 90579-90591.
[4] S. Garg, S. Tyagi, S. Singh, and S. Singh,(2021) "Industry 4.0 Enabled by 5G Networks: An
Overview, Challenges, and Opportunities" IEEE Consumer Electronics Magazine, vol. 10, no. 4, pp.
59-68, July 2021.
[5] H. Kim and J. Kim,(2020) "A Survey of 5G Communication Systems: Architecture, Radio Access,
and Core Network," Journal of Communications and Networks, vol. 22, no. 4, pp. 330-343.

112
[6] T. Li, Q. Zhang, and G. Li,(2021) "5G and Beyond: Recent Advances and Future Outlook," IEEE
Network, vol. 35, no. 1, pp. 8-15.
[7] R. Hosamani and T. Yerriswamy,(2022) "CSI based Performance Investigation of Single Cell-Multi
user MaMIMO System," 2nd Asian Conference on Innovation in Technology (ASIANCON), IEEE,
2022.
A. J. Goldsmith(200), Wireless Communications. Cambridge University Press, 2005.
[8] G. Kaddoum and A. O. Boudraa, (2020) "Time-frequency based channel estimation techniques for
OFDM-based systems: A survey," Digital Signal Processing, vol. 97, pp. 102671
[9] S. Zhang and Y. Chen, (2021)"Channel estimation and performance analysis for wireless sensor
networks in frequency-selective fading channels," IEEE Access, vol. 9, pp. 17934-17946
[10] Y. Song and Y. Li,(2007) "Channel Estimation for MIMO-OFDM Systems in Time-Varying
Channels," in IEEE Transactions on Broadcasting, vol. 53, no. 4, pp. 884-892,doi:
10.1109/TBC.2007.908704.
[11] Y. Wang, L. Dai, Z. Wang, and S. Zhang, (2012)"A Low-Complexity Channel Estimation Algorithm
for MIMO- OFDM Systems," in IEEE Transactions on Wireless Communications, vol. 11, no. 9,
pp. 3226-3235, doi: 10.1109/TWC.2012.072012.111243.
[12] Z. Shen et al.,(2019) "Channel Estimation for MIMO-OFDM Systems in Time-Varying Frequency-
Selective Channels," IEEE Transactions on Vehicular Technology, vol. 68, no. 5, pp. 4589-4601,
DOI: 10.1109/TVT.2019.2901674."
[13] Y. Cai et al., (2019)"Pilot Decontamination and Channel Estimation for Massive MIMO-OFDM
Systems in Frequency-Selective Fading Channels," IEEE Journal on Selected Areas in
Communications, vol. 37, no. 4, pp. 898-912, DOI: 10.1109/JSAC.2019.2906802"
[14] M. H. Rezaei and S. S. Agaian(2021) "Low-Overhead Channel Estimation for Frequency-Selective
MIMO Systems using Differential Evolution Algorithm" (IEEE Transactions on Communications).
[15] M. Ashraf, K. T. K. Teo, Y. Li, Z. Yang, and Y. Zhang, (2020) "Frequency Selective Channel
Estimation for 5G Communication Systems," IEEE Transactions on Wireless Communications, vol.
19, no. 2, pp. 1323-1335, DOI: 10.1109/TWC.2019.2956407.
[16] S. Salous,(2013) Radio Propagation Measurement and Channel Modelling.Hoboken, NJ, USA: Wiley
[17] A.F. Molisch, (2012)Wireless Communications, vol. 34, 2nd ed. Hoboken, NJ,USA: Wiley
[18] P. Almers, E. Bonek, N. Czink, M. Debbah, A. Burr, G. Matz, V. Degli-Esposti, H. Hofstetter, P.
Kyösti, D. Laurenson, A. F. Molisch, C. Oestges, and H. Özcelik,(2007) ``Survey of channel and
radio propagation models for wireless MIMO systems,'' EURASIP J. Wireless Commun. Netw., vol.
2007, no. 1, p. 19.
[19] P. Ferrand, M. Amara, S. Valentin, and M. Guillaud, (2016) ``Trends and challenges in wireless
channel modeling for evolving radio access,'' IEEE Commun. Mag., vol. 54, no. 7, pp. 93_99
[20] Guidelines for evaluation of radio transmission technologies for IMT- 2000,'' Int. Telecommun.
Union, Geneva, Switzerland, Tech. Rep. Rec.ITU-R M.1225. Question ITU-R 39/8, 1997.
[21] Guidelines for evaluation of radio interface technologies for IMTadvanced,(2009)'' Int. Telecommun.
Union, Geneva, Switzerland, Tech. Rep. M.2135-0 (2008)
[22] P. Hong and Y. Hwang,(2020) "Low Complexity Channel Estimation for OFDM Systems using Least
Squares Algorithm with Joint Spatial-Frequency Correlation Exploitation," in IEEE Access, vol. 8,
pp. 219580-219593, doi: 10.1109/ACCESS.2020.3047438.
[23] S. Chen, C. Yang, and Q. Liu,(2020) "Low-Complexity Channel Estimation Based on Recursive
Least Squares Algorithm for OFDM Systems," in IEEE Access, vol. 8, pp. 222444-222455, doi:
10.1109/ACCESS.2020.3049336.
[24] J. Liu, Y. He, H. Zhu, and X. Zhang,(2021) "Improved Channel Estimation for UFMC Systems Based
on Variable Step Size Least Squares Algorithm," in IEEE Access, vol. 9, pp. 44425-44436, doi:
10.1109/ACCESS.2021.3068419.
[25] H. Zhou, Y. Wang, and Q. Yang,(2021) "Low-Complexity Channel Estimation for OFDM-Based
Cooperative Communication Systems using Pilot-Aided Least Squares Algorithm," in IEEE Access,
vol. 9, pp. 47674-47685, doi: 10.1109/ACCESS.2021.3067845.
[26] M. A. Elshennawy, A. El-Keyi, and A. K. Elmaghraby, (2019)"Channel estimation in MIMO-OFDM
systems using MMSE estimation technique," in 15th International Computer Engineering Conference
(ICENCO), 2019, pp. 18-22.
[27] S. M. Hossain, M. S. Islam, and M. H. Rahman,(2019) "MMSE-based channel estimation for multi-
antenna systems with pilot contamination," in IEEE International Conference on Advanced
Communication Technologies and Networking (CommNet), pp. 1-5.

113
[28] X. Zhang, F. Jiang, and H. Liu,(2021) "Joint user activity detection and channel estimation for grant-
free massive MIMO with MMSE," in 2021 IEEE International Conference on Communications
Workshops (ICC Workshops), pp. 1-6.
[29] K. N. Choudhury, M. A. Islam, and A. Bhattacharjee,(2022) "Adaptive MMSE Channel Estimation
for Massive MIMO Systems," in IEEE International Conference on Electronics, Computing and
Communication Technologies (CONECCT), pp. 1-5.
[30] Y. Zhao, X. Zhang, and M. Elkashlan, (2019)"Channel Estimation for Hybrid Millimeter Wave
Massive MIMO Systems via Maximum Likelihood Estimation," in 2019 IEEE Global
Communications Conference (GLOBECOM), pp. 1-6.
[31] W. Li, H. Cao, and X. Li, (2021) "Joint Channel Estimation and Equalization for Underwater
Acoustic OFDM Systems Using Maximum Likelihood Estimation," in IEEE International
Conference on Communications (ICC), pp. 1-6.
[32] S. A. Basit, S. A. Malik, and S. G. M. Amin, (2022)"A Maximum Likelihood Approach for Joint
Estimation of DOA and Channel Parameters for Multipath Propagation," in 2022 IEEE 17th Annual
Wireless and Microwave Technology Conference (WAMICON), pp. 1-5.
[33] E. Stergiou, C. Angelis, and S. V. Margariti, "Evaluation Methodology of MIMO Networks
Performance over Rayleigh Fading," International Journal of Computer Networks & Communications
(IJCNC), vol. 12, no. 1, pp. 37, January 2020, doi: 10.5121/ijcnc.2020.12103.
[34] P. Poornima, G. Laxminarayana, and D. Srinivas Rao, "Performance Analysis of Channel Capacity
and Throughput of LTE Downlink System," International Journal of Computer Networks &
Communications (IJCNC), vol. 9, no. 5, pp. 55, September 2017, doi: 10.5121/ijcnc.2017.9505
AUTHORS
Ravi Hosamani received the B.E degree in Electronics and Communication Engineering
and his M.Tech from in VLSI Design and Embedded System from Visvesvaraya
Technological University (VTU) , Belgaum, Karnataka in 2006 and 2010 respectively and
currently pursuing Ph.D. in Electronics and Communication Engineering in ECE from KLE
Institute of Technology affiliated to VTU, Belgaum. He is working as assistant professor in
Electronics and Communication Engineering. His research interests includes wireless
communication, VLSI, digital signal processing, He published papers in various national and international
journals and conference and he is associated member of SMIEEE, LMISTE, MIEI, and IAENG.
Yerriswamy T received her B.Tech in Electronics and CommunicationEngineering,
M.Tech in Digital Communication and Networking from Visvesvaraya Technological
University (VTU) , Belgaum Karnataka in 2000 and 2005 respectively. He completed his
Ph.D. in 2013. He is currently working as professor in the dept of CSE KL Institute of
Technology, Hubballi. He published papers in various national and international journals
and conference. He is an active participant in both academic and research activities His
research interest are in the areas of array antenna Signal processing, Artificial Intelligence and 5G
Communication. He is member of IEEE, ISTE.

Channel Estimation in MIMO OFDM Systems with Tapped Delay Line Model

More Related Content

Similar to Channel Estimation in MIMO OFDM Systems with Tapped Delay Line Model (20)

More from IJCNCJournal (20)

Recently uploaded (20)

Channel Estimation in MIMO OFDM Systems with Tapped Delay Line Model