西南交通大学学报

第 55 卷 第 2 期 Vol. 55 No. 2
JOURNAL OF SOUTHWEST JIAOTONG UNIVERSITY
2020 年 4 月 Apr. 2020

ISSN: 0258-2724 DOI：10.35741/issn.0258-2724.55.2.32

Research article

Computer and Information Science

COMPARATIVE ANALYSIS FOR DIFFERENT DIGITAL MODULATION

TECHNIQUES

不同数字调制技术的比较分析

Mohamed M. Al-Heeti a, Mohammed A. Fadhel b, *

a
University of Anbar
Anbar, Iraq, mohamed.muthanna@uoanbar.edu.iq
b
College of Computer Science and Information Technology, University of Sumer
Rifai, Dhi Qar, Iraq, Mohammed.a.fadhel@uoitc.edu.iq

Received: January 10, 2020 ▪ Review: February 26, 2020 ▪ Accepted: April 15, 2020

This article is an open access article distributed under the terms and conditions of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/4.0)

Abstract
In the paper, we present an analytical comparison for various modulation techniques that can be used
in a digital communication system. Digital modulation offers more powerful security, information, quality
communication, and great capacity. Therefore, digital modulation systems have a maximal and important
role in their capacity to transfer bigger amounts of data than analogue ones. There are more types of
digital modulation like multiple amplitudes, frequencies, and phase shift keys. The results examine the
comparisons of modulation techniques and the benefits of each system to determine which would be the
most suitable for a specific application. The paper also discusses the effectiveness of different bandwidth
and power efficiency factors.

Keywords: Digital Modulation, Bandwidth, Power Efficiency

摘要 在本文中，我们对可以在数字通信系统中使用的各种调制技术进行了分析比较。 数字调制可
提供更强大的安全性，信息，高质量的通信和强大的容量。因此，数字调制系统在传输比模拟传
输更大数量的数据的能力中具有最大和重要的作用。数字调制的类型更多，例如多个幅度，频率
和相移键。结果检查了调制技术的比较以及每个系统的优点，以确定哪种方法最适合特定应用。
本文还讨论了不同带宽和功率效率因素的有效性。

关键词: 数字调制，带宽，功率效率
 2 Al-Heeti and Fadhel / Journal of Southwest Jiaotong University / Vol.55 No.2 Apr. 2020

I. INTRODUCTION
The process of modulating (switching or
keying) analogue carries signals in a binary code
and are known as digital modulation techniques.
Basically, there are some criteria to demonstrate
the best modulation schemes that are dependent
on Signal-to-Noise-Ratio (SNR), Bit Error Rate
(BER), bandwidth availability, spectral efficiency,
quality of service, and cost efficiency. The
measurement of performance of each modulation
planner is estimated using a probability of error Figure 1. Polar representation of a sine wave
when they are operated with a channel of
Additive White Gaussian Noise (AWGN) [1]. Therefore, it is easy to accomplish the digital
Two techniques for classification of digital modulation with I/Q modulation, as it
modulation schemes in terms of the phase constellates all the data in several discrete points
recovery circuit at the receiver, which are in the I/Q plane (constellation diagram). In
coherent techniques and non-coherent techniques. addition, the use of I/Q modulation is
The coherent technique is to control the phase, advantageous, as it combines independent
not just the frequency, by a tolerance component composite signals into single composite signals
such as the Voltage Controller Oscillation (VCO) and then splits them, at the demodulator, into
[2]. Some studies applied the comparisons of separate parts [5].
Analysis of Digital Modulation in LTE 4G
Systems using MATLAB SIMULINK [3]. B. Binary Frequency Shift Keying (BFSK)
Moranchel et al. presented three different When an information signal has multiple
modulation techniques and compared them by states, BFSK will assign a frequency to each state.
applying them with modular multilevel In the case of BFSK, the information signal only
converters (MMCs). In this paper, the space- has two states, which means that there are only
vector modulation (SVM), the phase-shifted two different frequencies [6].
sinusoidal pulse width modulation (PS-SPWM),
and the nearest level modulation (NLM) were
studied [4]. This paper presents the main
modulation techniques and comparisons based on
the following factors: power, cost and bandwidth
efficiency.

II. METHODOLOGY
The main purpose of digital modulation is to
transfer streams of digital bits over an analog
channel. There are many methods used to utilize
digital modulation. Some examples are:
Figure 2. Block diagram of FSK modulator
A. Using In-Phase (I) and Quadrature
Projections (Q)
In this concept, the signals are described in
the rectangular form by its rectangular
projections. Figure 1 shows “the x-axis that
represents the amplitude of the In-phase (1) of
the signal and the y-axis that represents the
amplitude of the Quadrature (Q) of the signal”
[5].
Figure 3. Waveform of FSK

Each signal can be defined by [7]:

"𝑆1 (𝑡) = 𝐴𝑐cos(2𝜋𝑓1𝑡)" (1)
"𝑆2 (𝑡) = 𝐴𝑐cos(2𝜋𝑓2𝑡)" (2)
 3

The two frequencies can be written as 2πfct is taken as the orthonormal basis function,
deviations from the carrier frequency. as shown in the constellation diagram in Figure 4.
𝑓1 = 𝑓𝑙 = 𝑓𝑐 − ∆𝑓
𝑓2 = 𝑓ℎ = 𝑓𝑐 + ∆𝑓. (3)
The relative amplitude of each frequency is Figure 4. BASK signal constellation
given in Table 1.

Table 1.
The relationship between frequency and amplitude
Parameter UM Parameter
-1 0 1
+1 1 0

Figure 5. Block diagram of BASK system

This means that the BFSK signal can be
represented at any point by:
Figure 6 shows the binary amplitude shift to
(𝑡) = 𝐴ℎcos(2𝜋𝑓ℎ𝑡) + 𝐴𝑙cos(2𝜋𝑓𝑙𝑡) (4) keying the binary sequence (0101001) that
The advantages of using a similar modulation created the signal sequence. The carrier of
scheme to this are that it is low cost and easy to amplitude is keyed or switched by the binary
implement. It is also less susceptible to errors due signal m (t) and this is known as "On-Off Keying
to the receiver identifying the frequency of the (OOK)" [7].
signal rather than the voltage [8]. However, this
modulation scheme does have a greater
bandwidth requirement than other schemes,
which makes it unsuitable for applications with a
limited bandwidth.

C. Binary Amplitude Shift Keying (BASK)

Binary Amplitude Shift Keying is a
modulation technique that varies the amplitude of
the carrier wave according to the binary
amplitude of the digital information signal. The Figure 6. (a) Binary modulating signal, and (b) BASK signal
carrier amplitude should be null to convey a waveform
binary 0 and at the maximum level to convey a 1,
as shown in Figure 3. The receiver keeps track of Figure 7 shows the energy per bit [10]:
the amplitude variation of the received modulated Eb = P Tb, watts-second.
signal to extract the bit stream sent by the Null-to-Null RF transmission bandwidth:
transmitting end [9]. BASK can be defined by:
Bnull = (fo + fb) – (fo – fb) = 2fb = 2Rb =
0<t<T (5) 2/Tb (8)
where A is peak value, m(t) = 1; when a digit 1, Bandwidth with 95% of the total transmitted
and m(t) = when a digit is 0, where the bit power:
duration represented by (t) and the carrier B95% = 3fb (Hz), centered at fundamental
frequency represented by (fc). It has a power frequency [11].
defined by:

(6)
S (t) can be written:

(7)
Figure 7. Power spectral density of band pass BASK
The energy contained in a bit duration is
represented by E=PT, whereas φ1 (t) = 2/T cos
 4 Al-Heeti and Fadhel / Journal of Southwest Jiaotong University / Vol.55 No.2 Apr. 2020

Figure 8. MATLAB result of BASK modulated signal and Figure 9. BPSK simulation using MATLAB
BASK signal waveform
In addition, the M-ary phase modulation is
The transmission bandwidth for an ASK represented as M = ; where the number of bits
signal is given by: per transmitted symbol is represented by (n).
BASK = (1 + r)R (9) Figure 10 shows the signal point constellations
for M = 2, 4, and 8.
where r "(typically 0 < r < 1) is related to the From the PSK scheme, another technique can
technique," so the bandwidth for transmission is also be formed, which is QPSK or the Quadrature
established when the signal is filtered. As shown Phase Shift Keying. In this technique, the in-
in Figure 8, BASK has high spectral power phase carrier (I) changed its phase from 0° to
density and low bandwidth, which leads to poor 180° and the Quadrature carrier (Q) changed its
efficiency and poor performance. Binary phase from 90° to 270°. This results in two bits in
amplitude shift keying (BASK) is susceptible to each symbol that are modulated at once. Hence,
sudden amplitude variation due to noise, multi- there are 4 different phases (0°, 90°, 180°, and
path propagation, and interference. ASK is not 270°) in which each phase has a symbol that
used in a wireless transmissión. It requires low includes 2 bits of information [12].
bandwidth and is used only up to 1200 bps data
rate on voice-grade lines as in telemetry [12].

D. Phase Shift Keying (PSK)

In this modulation scheme, the phase of the
carrier signal is impressed with the message
signal over a communication channel. For
transmitting digital information over this scheme,
the carrier phases used are θ = 2 π.m/M; where m
= 0, 1, 2... M-1 and M is "M-ary" modulation.
Hence, the modulated signal by this scheme can
be represented by the following:

(10)
The simplest PSK scheme is a Binary PSK
when (M = 2), which is shown in Figure 9 that
only two carrier phases used: θ1 = 0 and θ2 = 2 π.

Figure 10. Signal constellations: (a) BPSK M = 2,

(b) QPSK M = 4, (c) 8PSK M = 8

E. Quadrature Amplitude Modulation (QAM)

"OAM" is a combination of the two digital
modulations, ASK and PSK. That means the
carrier signal is varied in both its amplitude and
phase. The common form of the QAM scheme
are illustrated by M-ary to be M-QAM, where the
 5

number of bits per symbol equals . Table  In 64-QAM, there are 6 bits transmitted
2 shows the number of symbols and bits in each per symbol.
type of QAM scheme. Hence, using a higher number of modulation
states can send more bits in each symbol using
Table 2. the same amount of data in the narrow RF
QAM types spectrum [11].
Number of
Number of
QAM types bits per C. Multi-Level Signaling
symbols
symbols This is used for transmitting higher data rates
“4_QAM 4 2
8_QAM 16 3 using the relationship M = , where the number
16_QAM 16 4 of bits transmitted per symbol is represented by
32_QAM 32 5 (n) [8]. For example, using 16-level signaling can
64_QAM 64 6 transmit 4 bits per symbol.
128_QAM 128 7
1024_QAM 1024 10”
D. Bandwidth Efficiency
This is the measure of how well the
Figure 4 shows the signal point constellations modulation scheme is utilized in a given
plot of M-array of QAM using M = 4, 16, and 64. bandwidth. Its unit is bit/sec/Hz and can be
mathematically expressed as [12]:

(11)

From equation 11, the ratio (C/B) is the

bandwidth efficiency and the ratio (Eb/No) is the
power efficiency or SNR per bit.
Table 3 shows how the bandwidth efficiency
Figure 11. QAM-4, QAM-16 and QAM-64 can be approximately expressed in each model of
the studied digital modulation schemes, where
III. PARAMETERS FOR COMPARISON M-ary signalling M = 2, 4, 8, 16, 32, 64, 128...
and so on [13], [14].
OF DIGITAL MODULATION
TECHNIQUES Table 3.
These main parameters are used as shown “Bandwidth efficiency of digital modulation”
below: Digital modulation
Bandwidth efficiency (C/B)
scheme
A. Information Transfer Rate “M-ASK
M-FSK
This is the speed of information (bits)
transferred from source to sink. The unit is M-PSK
bits/sec. It also can be defined as a frequency of M-QAM ”
the system hit stream. For example, if a radio
with 16-bit sampler that are sampling 12 KHz of E. Bit Error Rate
information, the information transmission rate This means the ratio of "number of bits error"
will be = 16 * 12 = 192 bits/seconds [10]. to "the total number of bits" that is transmitted
over a given time duration; this is normally
B. Symbol Rate (Baud Rate) represented as a percentage [15].
This is the information bit transfer rate
divided by the number of bits transferred in each F. Probability of Symbol Error
symbol. Its unit is symbol/sec or baud. It is an This is a determination of the probability of an
interesting parameter to consider since the signal error on bits received. It is an important
bandwidth for the needed communications characteristic since the presence of noise of any
channel depends on it. For example: data transmission on communication systems
 In BPSK, there is 1 bit transmitted per [16]. Table 4 shows the probability of symbol
symbol. error of each model of the studied digital
 In QPSK, there are 2 bits transmitted per modulation schemes where M-ary signaling M =
symbol. 2, 4, 8, 16, 32, 64, 128... and so on [17], [18].
 In 8-QAM, there are 3 bits transmitted Table 4.
per symbol. Probability of symbol error of digital modulation schemes
 6 Al-Heeti and Fadhel / Journal of Southwest Jiaotong University / Vol.55 No.2 Apr. 2020

Digital modulation
Probability of symbol error modulation techniques in terms of probability
scheme and the symbol error was plotted using
MATLAB and the database shown in Table 3.
M-ASK
The following figures show the performance
probability of symbol error versus the Fib/No of
M-FSK
4 different digital modulation schemes (using M
= 2, 4, 8, 16).

M-PSK

M-QAM

IV. RESULT AND DISCUSSION

According to equation 11 and the database in
Table 2, Figure 12 shows the results of
bandwidth efficiency that were simulated using
MATLAB of all models of the studied digital
modulation techniques (using M = 2, 4, 8, 16, 64) Figure 12. Performance probability of symbol error of M-
at a given probability of symbol error of . PSK by MATLAB
As discussed in a previous section, it is always
desirable to maximize bandwidth efficiency at a
minimal required Eb/No.

Figure 14. Performance probability of symbol error of M-

QAM
Figure 12. Performance in bandwidth efficiency
In M-ary of QAM, PSK, and ASK, the
It can be clearly seen that when the number of probability of symbol error and bandwidth
Multi-signaling (M-ary) increases, the bandwidth efficiency for a given value of Eb/No increases as
efficiency of ASK, PSK, and QAM also increases. M increases. Hence, the SNR per bit required to
However, there is an increase of SNR per bit attain a given symbol error probability decreases
because there is a cost for the higher data rate per while M-ary increases and the transmit power
unit of bandwidth. Hence, most communications between them needs to increase to maintain a
systems that have limited bandwidth are desired certain performance scale. This proves that ASK,
using the schemes (ASK, PSK, and QAM) since PSK, and QAM are bandwidth efficient but less
they are appropriate for them. power efficient and are suitable for
On the one hand, the figure shows that the communication channels that are bandwidth
bandwidth efficiency decreases in the FSK limited where to uphold an increase in M is
scheme when there is an increasing number of enough to get a high SNR.
M-ary signalling. However, the SNR per bit On the other hand, in M-FSK, the probability
decreases while M-ary increases. So, the FSK of symbol error and bandwidth efficiency for a
scheme is the most appropriate technique for the given value of Eb/No decreases as M increases.
desired communication systems that have power Hence, the SNR per bit required to achieve a
limited channels. given symbol error probability increases while it
In addition, the results presented the increases in M-ary and the transmit power
performance comparison between digital between them needs to decrease to maintain a
 7

given performance level. This proves that FSK is techniques for modular multilevel converters.
power efficient but less bandwidth efficient and Energies, 9 (12), 1091.
is suitable for power-limited channels that have [6] CLAESSENS, S., PAN, N.,
sufficiently large bandwidth to accommodate a SCHREURS, D., and POLLIN, S. (2019)
large number of signals. Multitone FSK modulation for SWIPT. IEEE
For a comparison of the performance between
Transactions on Microwave Theory and
M-PSK and M-QAM, the ratio below was used
[2]: Techniques, 67 (5), pp. 1665-1674.
[7] LANGTON, C. (2005) Intuitive
Guide to Principles of Communications.
[Online] Available from:
(12)
http://complextoreal.com/wp-
where M = 4 and the ratio G = 1 means they have content/uploads/2013/01/block.pdf
the same performance. However, for M > 4, the [Accessed 20/03/20].
ratio will be G > 1, which means that QAM has a [8] SCHNEIDER, E.M., ALLEN, D.J.,
better performance and SNR advantage than PSK.
SATOSKAR, A., ZANJANI, S.M.M.,
CHESNEY, B.P., TUCKER, J.C., and
V. CONCLUSION AMADI, C.J. (2019) Multi-path analog front
When designing a communication system, the
end and analog-to-digital converter for a
designer must consider the parameters of the
system and select a modulation scheme that suits signal processing system. U.S. Patent
it best. The three main parameters are whether Application 10/284,217.
the modulation scheme is power-, cost-, or [9] WANG, H. and FAPOJUWO, A.O.
bandwidth-efficient. For example, if a designer (2019) Design and Performance Evaluation
wanted a system that was cheap and had enough of Successive Interference Cancellation-
wide bandwidth available, Frequency shift Based Pure Aloha for Internet-of-Things
keying would be the appropriate scheme to use. Networks. IEEE Internet of Things Journal,
Every communication system has trade-offs due 6, pp. 6578-6592.
to the different priorities of each communication [10] MOLISCH, A.F. (2012) Wireless
system. Communications, Vol. 34. Hoboken, New
Jersey: John Wiley & Sons.
REFERENCES [11] GOLDSMITH, A. (2005) Wireless
Communications. Cambridge: Cambridge
[1] ANAND, P., GUPTA, A., and University Press.
BHAGAT, S. (2014) Bit Error Rate [12] OTUNG, I. (2001) Communication
Assessment of Digital Modulation Schemes Engineering Principles. Macmillan
on Additive White Gaussian Noise, Line of International Higher Education.
Sight and Non Line of Sight Fading Channels. [13] PUTTNAM, B.J., LUÍS, R.S.,
International Journal of Engineering Science RADEMACHER, G., SAKAGUCHI, J.,
Invention, 3 (8), pp. 6-10. AWAJI, Y., WADA, N., MARCIANTE, J.,
[2] SMITH, D.R. (2004) Digital KUO, B.P.P., and RADIC, S. (2018) High-
Transmission Systems. New York: Springer capacity transmission with homogeneous
Science & Business Media. multi-core fibers and wideband optical comb
[3] MCCUNE, E. (2010) Practical sources. In: Proceedings of the SPIE 10561,
Digital Wireless Signals. Cambridge: Next-Generation Optical Communication:
Cambridge University Press. Components, Sub-Systems, and Systems VII,
[4] NDUJIUBA, C.U., ONI, O., and San Francisco, California, January-
IBHAZE, A.E. (2015) Comparative analysis February 2018. International Society for
of digital modulation techniques in LTE 4G Optics and Photonics.
systems. Journal of Wireless Networking and [14] LI, G. and LI, J. (2019) A Pulse Shaping
Communications, 5 (2), pp. 60-66. Based Optical Transmission System of
[5] MORANCHEL, M., HUERTA, F., 128QAM for DWDM with N × 904 Gbps.
SANZ, I., BUENO, E., and RODRÍGUEZ, Applied Sciences, 9 (5), 988.
F.J. (2016) A comparison of modulation
 8 Al-Heeti and Fadhel / Journal of Southwest Jiaotong University / Vol.55 No.2 Apr. 2020

[15] SHARMA, K., MISHRA, A., and content/uploads/2013/01/block.pdf [20/03/20

SAXENA, R. (2010) Analog & Digital 访问]。
Modulation Techniques: An Overview. [8] SCHNEIDER，E.M.，ALLEN，D.J.，
International Journal of Computing Science SATOSKAR ， A. ， ZANJANI， S.M.M. ，
and Communication Technologies, 3 (1).
CHESNEY ， B.P. ， TUCKER ， J.C., 和
[16] LEE, E.A. and MESSERSCHMITT,
D.G. (2012) Digital Communication. AMADI，C.J.（2019）多路径模拟前端和
Springer Science & Business Media. 模数转换器，用于信号处理系统。美国专
[17] GUIMARAES, D.A. (2010) Digital 利申请 10 / 284,217。
Transmission: A Simulation-Aided [9] WANG，H。和 FAPOJUWO，A.O。
Introduction with VisSim/Comm. Berlin, （2019）基于物联网的基于连续干扰消除
Heidelberg: Springer.
的纯阿罗哈的设计和性能评估。电气工程
[18] KAUR, H., JAIN, B., and VERMA, A.
(2011) Comparative Performance Analysis of 师学会物联网杂志，6，第 6578-6592 页。
M-ary PSK Modulation Schemes Using [10] MOLISCH，A.F。（2012）无线通信，
Simulink. International Journal of 第一卷。34.新泽西州霍博肯：约翰·威利
Electronics & Communication Technology, 2 父子。
(3), pp. 204-209. [11] GOLDSMITH，A。（2005）无线通
信。剑桥：剑桥大学出版社。
[12] OTUNG，I。（2001）通信工程原理。
参考文: 麦克米伦国际高等教育。
[1] ANAND ， P. ， GUPTA ， A 。 和 [13] PUTTNAM ， B.J. ， LUIS ， R.S. ，
BHAGAT，S。（2014）关于加性高斯白 RADEMACHER ， G. ， SAKAGUCHI， J. ，
噪声，视线和非视线衰落信道的数字调制 AWAJI，Y.，WADA，N.，MARCIANTE，
方案的误码率评估。国际工程科学发明杂 J.，KUO，B.P.P., 和 RADIC，S.（2018）
志，3（8），第 6-10 页。 高均质多芯光纤和宽带光梳源的高容量传
[2] 史密斯 D.R.（2004）数字传输系统。 输。载于：斯皮 10561 的论文集，下一代
纽约：施普林格科学与商业媒体。 光通信：组件，子系统和系统 VII，加利
[3] MCCUNE，E.（2010）实用数字无线 福尼亚州旧金山，2018 年 1 月至 2 月。国
信号。剑桥：剑桥大学出版社。 际光学和光子学会。
[4] NDUJIUBA ， C.U. ， ONI ， O. 和 [14] LI，G. 和 LI，J.（2019）一种基于脉
IBHAZE，A.E.（2015）LTE 4G 系统中数 冲整形的 NQ 904 千兆位数字波分复用的
字调制技术的比较分析。无线网络和通信 128 质量管理光传输系统。应用科学，9
杂志，5（2），第 60-66 页。 （5），988。
[5] M.的 MORANCHEL，F。的 HUERTA， [15] SHARMA ， K. ， MISHRA ， A. 和
I。的 SANZ，E。的 BUENO 和 F.J.的 SAXENA，R.（2010）模拟与数字调制技
RODRÍGUEZ（2016）模块化多电平转换 术：概述。国际计算科学与通信技术杂志，
器 的 调 制 技 术 比 较 。 能 源 ， 9 （ 12 ） ， 3（1）。
1091。 [16] LEE，E.A。和 MESSERSCHMITT，
[6] CLAESSENS ， S. ， PAN ， N. ， D.G.（2012）数字通信。施普林格科学与
SCHREURS，D. 和 POLLIN，S.（2019） 商业媒体。
刷卡的多音 FSK 调制。电气工程师学会微 [17] GUIMARAES，D.A。（2010）数字
波理论与技术学报，67（5），第 1665- 传输：威盛/通讯的模拟辅助介绍。柏林，
1674 页。 海德堡：施普林格。
[7] LANGTON，C.（2005）通信原理的直 [18] KAUR，H.，JAIN，B. 和 VERMA，
观指南。[在线]可从以下网站获得： A.（2011）使用 Simulink 的玛丽·PSK 调
http://complextoreal.com/wp-
 9

制方案的比较性能分析。国际电子与通信
技术杂志，2（3），第 204-209 页。