Highlights in Science, Engineering and Technology ESAC 2025

Volume 134 (2025)

Performance Analysis of QAM Modulation and Demodulation

Yukun Chai *
School of Information, Beijing University of Posts and Telecommunications, Beijing, China
* Corresponding Author Email: [email protected]
Abstract. The following study will investigate the performance of Multiple Quadrature Amplitude
Modulation(M-QAM) in baseband modulation systems and further analyze the bit error rates (BER)
performance under different conditions of communication. The error performance of several M-QAM
schemes has been extensively simulated and analyzed under ideal Additive white Gaussian noise
(AWGN) channel conditions. The influences of some common channel impairments, such as
frequency offset and partial band interference, on the performance of the system are also taken into
consideration to assess the reliability of the system in depth. The results in this paper will show both
the potential and challenges of M-QAM in practical communication systems by comparing the BER
under different channel models. Notably, there are usually inevitable frequency offset and
interference in wireless communications. Therefore, a study on its effect on the performance may be
very necessary for the further enhancement of robustness and interference resilience in modern
communication systems. This work will provide enlightenment on insights into the design of the
baseband modulation system and support theoretical improvements to enhance communication
quality and reliability.
Keywords: M-QAM, BER, AWGN, Frequency Offset, Partial Band Interference.

1. Introduction
1.1. Research Background
The demand for high data rates in modern communication systems has led to the development of
several modulation schemes, among which M-QAM is one of the most popular techniques. M-QAM
can send more than one bit per symbol by changing both the amplitude and phase of the carrier signal.
This results in higher spectral efficiency, thus making it suitable for high-throughput applications
such as wireless communications, broadband internet, and digital television.
However, despite such merits, M-QAM systems remain susceptible to various channel
impairments, most notably AWGN that is usually modeled in analytical work. In practical
environments, further channel distortions, such as frequency offset and partial band interference, can
also be present. The former is due to a mismatch between the transmitter and receiver frequencies,
which will potentially cause symbol misalignments. Partial band interference involves interference
in some frequency bands that could degrade the signal quality and increase the BER [1,2]. With the
evolution in the communication system, it has been of prime importance to know the performance of
M-QAM under these impairments.
1.2. Research Focus and Objectives
This research focuses on the performance evaluation of M-QAM systems in the presence of
AWGN, frequency offset, and partial band interference. The simulator models the behavior of a
system subjected to these degradations by calculating the BER as the main performance metric.
Different M-QAM schemes, such as 16QAM [3] and 64QAM, are considered to observe how
increasing the modulation order affects system performance in noisy and distorted channels [4,5,6,7].
While it seeks to determine the performance of channel impairments in M-QAM systems and more
importantly, how higher-order modulations, which will yield higher data rates, necessarily trade off
against increased sensitivity to noise and interference, this paper attempts to be practical in the choice
of modulation schemes in realistic environments [8].

114
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

2. System Model and Basic Theory

2.1. System Introduction
The whole system structure is presented in Fig.1. First, the system generates raw bits and
modulates them into corresponding symbols according to different QAM constellations. The
modulated signal will enter the channel, where the signal will be added to the AWGN, frequency
offset, and interference signal. For exploring the effect on the signal, the signal is demodulated at the
receiving end to recover the bits. We compare the returned bits with the original bits and can calculate
the BER of the system [9].

Figure 1. The block diagram of QAM systems

2.2. QAM modulation theory
The theoretical expression for the minimum Euclidean distance between points in a constellation
diagram is given by Equation (1)
6log2 ⁡ M⋅Eb
d=√ . (1)
M−1

Equation (1) is the theoretical expression for the minimum Euclidean distance between points in
the constellation diagram. In the above formula, M is the number of symbols in the constellation and
log 2 ⁡ M indicates the number of bits per symbol. 𝐸𝑏 is the energy per bit, which is one of the main
parameters characterizing system performance in terms of power efficiency. 𝑀 − 1 in the
denominator normalizes for the average energy of the constellation. This distance d has a direct
impact on system performance since a higher value of d decreases the probability of symbol errors in
an additive noise environment. Therefore, the formula presents the trade-off between constellation
size M and the system's robustness to noise as a basic measure for the study of modulation schemes
such as M-QAM.
2.3. Constellation Diagrams for QAM Modulation of Different Orders
Modulation is done on the binary bits. It is basically the process of mapping the binary data onto
a specific modulation constellation, such as QAM. During this process, each group of binary bits is
mapped to a complex symbol, which corresponds to a point in the modulation constellation on the
complex plane. This process effectively transforms the original binary data into complex symbols
suitable for wireless transmission. The system reshapes the modulated symbols into the length of the
frame and the number of symbols, respectively, in three-dimensional matrix form for further
processing. Each frame has a certain number of symbols; the whole sequence of signals is rearranged
in that structure for further transmission or processing tasks. Finally, the system visualizes the
modulated signal by plotting a scatter diagram, showing the distribution of the symbols on the
complex plane. This allows for a visual inspection of the modulation constellation, where the symbol
spacing and phase differences can be analyzed.
This article uses constellation diagrams generated by the models for QPSK, 16QAM, and 64QAM
modulation as examples which can be seen in Fig. 2. Upon comparison, it is evident that as the
modulation order increases, the distance between points in the constellation diagram becomes smaller,
which also implies a weaker ability to resist channel noise and distortion.

115
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

(a) QPSK

(b) 16QAM (c) 64QAM

Figure 2. Constellation maps of different M-QAM

3. System Performance Analysis

3.1. AWGN
3.1.1. Calculation formula of BER
R In an M-QAM system, the symbol error rate (SER) can be approximately calculated using the
following formula
4 1 3E log ⁡ M
Pe,MQAM ≈ log (1 − ) ∑√M/2
i=1
b 2
Q ((2i − 1)√ (M−1)N ), (2)
2⁡ M √M 0

where M is the modulation order, log 2 ⁡ M represents the number of bits per symbol, Q(x) is the
Q-function that characterizes the tail probability of the Gaussian distribution, Eb denotes the energy
per bit, and N0 is the noise power spectral density. The formula consists of three parts: a scaling
4 1
factor log (1 − ) , a summation term related to the constellation points ∑√M/2
i=1 Q(⋅) , and a
2⁡ M √M
b 2 3E log ⁡ M
coefficient √ (M−1)N which reflects the relationship between the SNR and system performance.
0
This formula provides a theoretical estimation of the SER for M-QAM modulation in an AWGN
channel, effectively capturing the trade-offs between modulation order, signal energy, and noise
strength.

116
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

3.1.2. Constellation Diagrams under Different QAM Modulation Schemes with Added Noise
QPSK (Quadrature Phase Shift Keying), 16QAM (16-Quadrature Amplitude Modulation), and
64QAM (64-Quadrature Amplitude Modulation) are commonly used modulation schemes in digital
communication systems, each offering different trade-offs between data rate and robustness [10].
QPSK is a phase modulation technique that has each symbol carrying 2 bits of information by
using four different phase shifts to encode the signal. This will provide better spectral efficiency
compared to BPSK, with fairly good noise immunity, hence it is suitable for low SNR environments.
16QAM combines both amplitude and phase modulation, where each symbol represents 4 bits of
information attained by mapping 16 different symbols on the complex plane. It gives higher data rates
compared to QPSK; a factor that makes it effective in applications where bandwidth is limited.
However, 16QAM has a greater vulnerability to noise, hence a higher BER under poor channel
conditions.64QAM extends this further, where in this case, each symbol represents 6 bits of
information, mapped across 64 distinct points on the complex plane. While 64QAM offers the highest
data rate, it demands a high SNR for reliable communication, as the symbols are closely spaced,
increasing the likelihood of errors in noisy environments [11,12].
In order to investigate the impact of noise with different signal-to-noise ratios (SNR) on various
QAM modulation schemes we add the Gaussian white noise with SNR ranging from -10 dB to 20 dB
to the channel. The constellation diagrams for QPSK, 16QAM and 64QAM modulation schemes are
recorded accordingly. The noise levels with SNR of -10 dB, 0 dB, 10 dB and 20 dB are used as
representatives, as shown in the Fig.3 below.

QPSK Images with a SNR of -10, 0, 10, 20

16QAM Images with a SNR of -10, 0, 10, 20

64QAM Images with a SNR of -10, 0, 10, 20

Figure 3. Constellation maps of different M-QAM with added noise
It can be seen that, for the same noise level, e.g., SNR = 10 dB, the constellation diagram for
64QAM modulation is far more cluttered and disordered as compared to that of the 16QAM
modulation. For 64QAM constellations, the symbols are very compact in the complex plane, with

117
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

almost negligible distance between them. This makes the system vulnerable to noise, as even small
variations in the signal can result in the misinterpretation of a symbol, thus increasing the BER. In
contrast, the 16QAM constellation has more widely spaced symbols, thus greater separation between
them. The increased spacing of the symbols, in turn, provides improved error resilience since noise-
induced deviations are less likely to cause symbol misclassification.
This observation points to one of the fundamental trade-offs in modulation schemes: while the
modulation order increases, more bits are transmitted per symbol, which results in a higher data rate.
However, this also means that the signal's sensitivity to noise increases. The reduced distance between
adjacent symbols in higher-order modulation schemes like 64QAM means that the system's ability to
resist channel noise is compromised, particularly in environments with lower signal-to-noise ratios.
In contrast, lower-order schemes like 16QAM, while offering lower data rates, tend to be more robust
against noise and, therefore, are preferable in conditions where the signal quality is less stable or the
SNR is lower. Thus, the choice of modulation scheme must carefully balance the need for high data
rates with the system's robustness to noise, depending on the specific requirements of the
communication system and the quality of the transmission channel.
3.1.3. BER under the Influence of Noise
The theoretical BER of the system can be computed using Equation (2). By storing the actual
simulation data of SNR values ranging from -10 to 20 and plotting the results, we can see the BER
trends for different QAM modulation orders. We then compare the theoretical BER, derived from the
formulas, with the actual simulated BER for QPSK, 16QAM, and 64QAM modulation schemes on
the same linear graph Fig. 4.

Figure 4. BER of M-QAM under the influence of noise

The results clearly indicate a uniform variation pattern, and this can be clearly depicted in the
figure. The theoretical BER values are found to agree rather well with the actual simulation results,
thereby justifying the accuracy of the analysis. One can also find that with the same SNR condition,
a larger modulation order corresponds to a higher BER value. The immediate observation which can
be very well made based on the higher-order modulation schemes is that with the increase in
modulation order, the system becomes more vulnerable to noise. As a result, the noise resistance
performance of the system deteriorates, leading to the increased possibility of errors. This happens
because higher-order modulations require finer decision boundaries that are more susceptible to
distortions caused by noise. These findings underscore the trade-off between data rate and error
performance in M-QAM systems, where achieving higher data throughput may come at the cost of
reduced robustness to noise.

118
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

3.2. Impact of Frequency Offset on Performance

3.2.1. Normalized Frequency Offset
Frequency offset in communication systems refers to the difference in carrier frequency between
the received signal and the transmitted signal. Frequency offset typically causes phase changes in the
signal, which can affect signal demodulation and subsequently impact the BER.
The normalized frequency offset x represents a point's shift relative to the previous point, where
the shift is 𝑥 × 2𝜋. The specific formula can be expressed simply as the following equation
Txsignal 2 (t) = Txsignal 1 (t) ⋅ ej2πfoffset t . (3)
3.2.2. Changes in Constellation Diagrams under Different Modulation Schemes
The normalized frequency offset manifests in the constellation diagram as a rotation of each point
around the center by an angle of. To observe this effect more intuitively, we simulated 1024 × 1000
symbols and 1024 ×10000 symbols and selected two frequency offset angles: a relatively large value
of 0.125 and an extremely small value of 0.0000001. By comparing the changes in the constellation
diagrams under different modulation schemes, we can analyze how the modulation order influences
the impact of frequency offset. This can be seen in Fig 5.
QPSK Number of Symbols 1024000

Normalized Frequency Offset 0.125 Normalized Frequency Offset 0.0000001

QPSK Number of Symbols 10240000

Normalized Frequency Offset 0.125 Normalized Frequency Offset 0.125

119
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

16QAM Number of Symbols 1024000

Normalized Frequency Offset 0.125 Normalized Frequency Offset 0.0000001

16QAM Number of Symbols 10240000

Normalized Frequency Offset 0.125 Normalized Frequency Offset 0.01

64QAM Number of Symbols 1024000

Normalized Frequency Offset 0.125 Normalized Frequency Offset 0.0000001

120
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

64QAM Number of Symbols 10240000

Normalized Frequency Offset 0.125 Normalized Frequency Offset 0.0000001

Figure 5. Constellation maps of different Frequency Offset
Through comparison, we can observe that the impact of symbol numbers on frequency offset is
multifaceted. For higher-order modulation, the impact of frequency offset is usually more significant
because the signal points are closely packed; hence, they are prone to phase shifts [13]. On the other
side, the lower-order modulation will be more robust to frequency offset and can tolerate a certain
amount of frequency deviation.
3.3.3. Analysis and Conclusion
From the images, it can be observed that with an increase in the number of phase shifts, their
cumulative effect also grows with the number of transmitted symbols, particularly in long
transmissions. Systems with fewer symbols or shorter transmission times are less affected since the
phase errors accumulate more slowly. Thus, while higher-order modulation increases spectral
efficiency, it is more sensitive to frequency offset and requires more precise frequency
synchronization. In a similar way, the systems with longer transmission duration or more symbols
become more sensitive to the accumulation of frequency offset. Balancing modulation order, symbol
count, and frequency offset tolerance is crucial in system design and optimization.
3.3. Effect of Partial Band Interference
3.3.1. Generation of Band Interference Using QPSK Modulation
1) Bit Stream Generation and QPSK Modulation
The code first generates a random bit stream and then maps it to the QPSK symbols. QPSK
modulation is one of the most usable modulation schemes in which every symbol caries 2 bits. The
bit stream is first modulated using the function qammod to get complex symbols. To normalize the
modulated symbols in such a way that they have an average unit power, it normalizes the system by
setting the parameter UnitAveragePower=true. This normalizes the symbols to unit average power so
that the signal is comparable at different power settings.
2) Power Control
In order to simulate how different interference power affects the system, the simulation code
implements power control. Here, the power of the interference signal is changed based on the ratio
between the powers of the interference signal and the effective signal, referred to as the Jamming-to-
Signal Ratio (JSR) in decibels. That is, the JSR value in dB defines the logarithmic ratio of the
interference signal power to the effective signal power. It can be evaluated by using the following
expression
P
JSR⁡(dB)=10log10 ⁡ ( Psignal ).
jammer
(4)

121
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

The goal of JSR is the scaling of the amplitude of the interference signal by using a gain factor. In
order to apply the JSR, the code uses the following formula to calculate the gain factor and adjust the
interference signal power
JSR
. (5)
Power⁡Adjustment⁡Factor=10 20
This formula converts the JSR value from dB to a linear gain. By performing element-wise
multiplication on the modulated interference symbol matrix, the amplitude of the interference signal
is scaled by the corresponding multiple of the original effective signal's amplitude. Particularly,
the gain factor scales the interference signal's power due to a given JSR value. For instance, if the
JSR is 10 dB, the gain factor will be around 3.16, which means that the power of the interference
signal will be 3.16 times the effective signal power. In the case where the JSR is -10 dB, the gain
factor will be approximately 0.316, meaning that the interference signal power will be 0.316 times
the effective signal power. This approach allows the code to simulate interference with varying
strengths and adjust the interference signal's impact on the system as needed.
3) Frequency Domain Mapping and Inverse Transformation
The program arranges the modulated interference signal in a frequency domain matrix in a certain
way and then applies the Inverse Fast Fourier Transform to this matrix. The inverse transform serves
to convert the frequency domain signal back into the time domain, generating the final interference
waveform. During this process, the signal is appropriately normalized to ensure that the resulting
time-domain signal has the correct amplitude.
4) Frequency Domain Mapping and Inverse Transformation
Finally, the interference signal, after undergoing the IFFT process, is stored in the interference
variable and returned as the output. At this point, the interference signal is ready in the time domain
for further analysis or for combination with other signals. The whole process is shown in the Fig 6.

Figure 6. The block diagram of generating interference signals

3.3.2. Interference Factors
1) Centralized Partial Band Interference
Interference Signal Bandwidth: The interference signal mentioned here occupies a portion of the
effective signal bandwidth. In communication systems, interference signals refer to those signals that
do not belong to the valid communication signal but still reside within the same frequency band.
Centralized partial band interference refers to interference signals that are concentrated within a
specific frequency band of the effective signal's bandwidth, rather than being evenly distributed
across the entire frequency band.
Interference Bandwidth: Let the operational bandwidth of the effective signal be denoted as 𝐵𝑠 ,
and the bandwidth range of the interference signal be 𝐵𝑖 . This implies that the interference signal
occupies only a portion of the effective signal's bandwidth, rather than the entire frequency band.
2) Interference Factor
The Interference Factor (𝛾): The interference factor γ is used to quantify the proportion of the
interference signal relative to the effective signal's bandwidth. It is defined as
B
γ = Bi , (6)
s

Where:
Bi is the bandwidth of the interference signal, i.e., the width of the frequency band occupied by
the interference signal.

122
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

Bs is the operational bandwidth of the effective signal, i.e., the bandwidth used by the
communication system to transmit the valid signal.
Interference factor γ is one of the important parameters that characterizes the share of the frequency
the spectrum occupied by the interference signal. It is directly involved in system performance
regarding interference strength, BER, JSR, and spectral efficiency. A higher interference factor γ
indicates that the interference signal has a greater impact on the system and is typically associated
with degradation in communications performance, such as an increased bit error rate. By analyzing
γ, the performance of the system in interference can be evaluated and optimized. Particularly in
various modulation schemes, like QPSK, 16QAM, and 64QAM.
3) The BER Results of Different QAM Under Different JSR
250
Experiments were conducted with the parameter⁡r = 1024 = 24.4%and without other impairments.
The BER results of M-QAM under different JSR conditions are shown in Fig 7.

Figure 7. The BER results of different QAM under different JSR

4) image analysis
To begin with, Higher-Order QAM Exhibits Weaker Interference Resilience to Partial Band
Interference:
The results indicate that higher-order QAM modulation schemes (e.g., 16QAM, 64QAM) exhibit
weaker resistance to partial band interference compared to lower-order QAM schemes (such as
QPSK). This is because higher-order modulation schemes carry more bits per symbol, leading to
higher symbol density and smaller distances between constellation points. Therefore, when part of
the frequency band is interfered with, the symbols in these higher-order modulation schemes are more
likely to be misinterpreted, resulting in higher bit error rates (BER). Overall, higher-order QAM
modulation schemes perform worse in such interference environments, exhibiting weaker
interference resilience.
Furthermore, The Impact of JSR (Jamming-to-Signal Ratio) on BER:
As the JSR increases (i.e., the interference signal becomes stronger relative to the noise signal),
the relative strength of the signal increases, and the system's BER performance improves. This is
because a higher JSR implies that the signal is more prominent compared to the interference, reducing
the impact of interference on the signal, which in turn allows the system to detect the signal more
accurately, thereby lowering the BER. This aligns with intuition: when the interference signal is
stronger, the system is better able to receive the effective signal clearly, reducing the bit error rate.
Besides, The Effect of Interference Factor:
With the interference signal occupying 24.4% of the bandwidth, the system is subject to
interference over a significant portion of the spectrum. For higher-order QAM modulation schemes,
this frequency overlap results in more substantial performance degradation, as the constellation points
in higher-order QAM are denser, with smaller distances between symbols. This makes the signal

123
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

more susceptible to interference, leading to an increased bit error rate. Therefore, under the same
interference conditions, the performance degradation in higher-order QAM systems is more
pronounced, while lower-order QAM schemes (e.g., QPSK) exhibit relatively stronger resistance to
interference.

4. Conclusion
4.1. Summary of Research Findings and Conclusions
This study investigates the performance of baseband modulation systems utilizing M-QAM, with
a particular focus on the impact of higher modulation orders on system performance. The results
indicate that as the order of QAM increases, the distance between constellation points decreases,
leading to reduced resistance to noise and other channel impairments. Although higher-order QAM
schemes, such as 64QAM, provide higher data transmission rates, their performance under noise,
frequency offset, and band interference significantly deteriorates, resulting in a substantial increase
in BER. Thus, selecting the appropriate modulation order requires balancing data transmission speed
with system robustness. Advanced algorithms, such as carrier synchronization, symbol timing
recovery, channel estimation, and equalization, will also be necessary in practical communications to
mitigate channel impairments.
4.2. Implications and Significance
These results have an important meaning for the design and optimization of communication
systems. The modern tendency in communication systems is the continuous increase in data
transmission rates. In this context, the most important challenge is the need to find an appropriate
balance between high-speed transmission and the possibility of the system's resistance to noise and
interference. In such conditions, the analysis of the M-QAM system with different channel conditions
gives significant insight into the choice of modulation schemes for various applications. While
higher-order M-QAM modulations yield increased data rates, it makes them more susceptible to noise
and impairments, therefore, requiring sophisticated compensation techniques. It reassures that this
study emphasizes how this understanding of the trade-off between transmission speed and reliability
may guide the design of robust communication systems in realistic environmental conditions.
4.3. Future Research Directions
In this respect, future research work should further consider the performance of M-QAM systems
under more complex and dynamic channel conditions. As communication technologies are
continuously evolving, traditional modulation schemes and compensation algorithms may struggle to
handle new types of channel impairments introduced by high-speed mobility or multi-path fading. It
thus follows that future work needs to shift attention to algorithm development, especially deep
learning-based adaptive channel estimation, real-time modulation scheme optimizations, and hence
enhance robustness in demanding environments. Hybrid modulation techniques, combined with
MIMO, shall also be another interesting topic for further investigations, regarding their application
on emerging 5G/6G networks, hence opening further perspectives for enhancement of the efficiency
and reliability of a communication system.

References
[1] Breed, Gary. "Bit error rate: Fundamental concepts and measurement issues." High Frequency Electronics
2.1 (2003): 46 - 47.
[2] Jeruchim, Michel. "Techniques for estimating the bit error rate in the simulation of digital communication
systems." IEEE Journal on selected areas in communications 2.1 (1984): 153 - 170.
[3] El-Nahal, Fady. "Coherent 16 quadrature amplitude modulation (16QAM) optical communication
systems." Photonics letters of Poland 10.2 (2018): 57 - 59.

124
Highlights in Science, Engineering and Technology ESAC 2025
Volume 134 (2025)

[4] Hong, Sungnam, et al. "Frequency and quadrature-amplitude modulation for downlink cellular OFDMA
networks." IEEE journal on selected areas in communications 32.6 (2014): 1256 - 1267.
[5] Park, Sung-Joon. "Performance analysis of triangular quadrature amplitude modulation in AWGN
channel." IEEE communications letters 16.6 (2012): 765 - 768.
[6] Pappi, Koralia N., Athanasios S. Lioumpas, and George K. Karagiannidis. "θ-QAM: A parametric
quadrature amplitude modulation family and its performance in AWGN and fading channels." IEEE
Transactions on Communications 58.4 (2010): 1014 - 1019.
[7] RANHOTRA, SARVRAJ SINGH. "Analytical performance evaluation of OFDM in presence of channel
estimation errors and carrier frequency offset." (2015).
[8] Delmas, J-P., and Habti Abeida. "Cramer-Rao bounds of DOA estimates for BPSK and QPSK modulated
signals." IEEE Transactions on Signal Processing 54.1 (2005): 117 - 126.
[9] Ali, Irfan. "Bit-error-rate (BER) simulation using MATLAB." International Journal of Engineering
Research and Applications 3.1 (2013): 706 - 711.
[10] Singya, Praveen Kumar, et al. "A survey on design and performance of higher-order QAM constellations."
arXiv preprint arXiv:2004.14708 (2020).
[11] Youssef, Tamer, and Eman Abdelfattah. "Performance evaluation of different QAM techniques using
Matlab/Simulink." 2013 IEEE Long Island Systems, Applications and Technology Conference (LISAT).
IEEE, 2013.
[12] Zhendong, Chou, et al. "Modulation recognition based on constellation diagram for M-QAM signals."
2013 IEEE 11th International Conference on Electronic Measurement & Instruments. Vol. 1. IEEE, 2013.
[13] Singya, Praveen Kumar, et al. "A survey on higher-order QAM constellations: Technical challenges,
recent advances, and future trends." IEEE Open Journal of the Communications Society 2 (2021): 617 -
655.

125