Secured Communication of Speech Signal Using the Discrete Cosine
Transform Based on Hyperchaos-System
Noureddine Aissaoui
1 a
, Fethi Demim
2 b
, Abdenebi Rouigueb
3 c
, Hadjira Belaidi
6 d
,
Ali Zakaria Messaoui
4 e
, Kahina Louadj
5 f
, Abdelkrim Nemra
2 g
, Ahmed Allam
7
, Yasmine Saidi
7
,
Said Sadoudi
1
and Mohamed Salah Azzaz
1
1
Laboratoire Syst
`
emes Electronique et Num
´
eriques, Ecole Militaire Polytechnique, Bordj El Bahri, Algiers, Algeria
2
Laboratory of Guidance and Navigation, Ecole Militaire Polytechnique, Bordj El Bahri, Algiers, Algeria
3
Laboratory of Artificial Intelligence and Virtual Reality, Ecole Militaire Polytechnique, Bordj El Bahri, Algiers, Algeria
4
Complex Systems Control and Simulators Laboratory, Ecole Militaire Polytechnique, Bordj El Bahri, Algiers, Algeria
5
Laboratoire d’Informatique, Mathmatiques, et Physique pour l’Agriculture et les For
ˆ
ets, Universit
´
e de Bouira, Algreia
6
Signals and Systems Laboratory, Institute of Electrical and Electronic Engineering,
University M’Hamed Bougara of Boumerdes, Algeria
7
Ecole Nationale Polytechnique, Algiers, Algeria
Keywords:
Chaos-Based Cryptography, Secure Chaotic Communication, Speech Signal, Discrete Cosine Transform,
Encryption-Based Diffusion.
Abstract:
This paper proposes a novel approach that combines chaos-based encryption and Discrete Cosine Transform
(DCT) to ensure high-level speech security and robustness against attacks. In this approach, the encryption
process is based on Lorenz’s hyperchaotic system, which utilizes the One Time Pad approach to encrypt the
speech DCT coefficients. The effectiveness of this approach has been validated through experiments on two
PCs interconnected via real-time serial communication links (USB-RS232), which showed that the original
speech is effectively hidden, and the proposed solution is highly resistant to possible attacks. Moreover, the
proposed solution can be implemented in real-time applications using technologies such as FPGA.
1 INTRODUCTION
Speech-based communication has taken an important
place in many fields, such as civil applications like
trade, military defence, telephone banking and tele-
conferences, etc. This type of communication is in
continuous evolution with the continuous investiga-
tion for better speeds, improved mobility and espe-
cially high confidentiality. The exchange of impor-
tant confidential information via unsecured commu-
nication provides easy access to secret information
a
https://orcid.org/0000-0002-5328-2414
b
https://orcid.org/0000-0003-0687-0800
c
https://orcid.org/0000-0001-5699-2721
d
https://orcid.org/0000-0003-2424-626X
e
https://orcid.org/0000-0001-5753-5776
f
https://orcid.org/0000-0002-4203-6357
g
https://orcid.org/0000-0001-9237-9449
and facilitates the hacking process. Today’s secu-
rity devices are mainly based on encrypted systems
that guarantee specific requirements. On one hand,
the current encryption methods such as Rivest Shamir
Adleman (RSA), Data Encryption Standard (DES)
systems, and Rivest Cipher 4 (RC4) have already been
defeated and are therefore no more secured. In this
context, it is necessary to provide using other meth-
ods that are secure and have not yet been broken with
the generation of microprocessors, such as Cipher 5
or the Advanced Encryption Standard (AES). Thus,
chaotic cryptography is one of the alternatives created
in the last ten years, particularly quantum cryptogra-
phy, even though it cannot be opened due to its char-
acteristics and obscures the data from its original form
into difficult-to-understand data.
In 1990, L.M. Pecora et al. established the practi-
cal feasibility of synchronization between two identi-
Aissaoui, N., Demim, F., Rouigueb, A., Belaidi, H., Messaoui, A., Louadj, K., Nemra, A., Allam, A., Saidi, Y., Sadoudi, S. and Azzaz, M.
Secured Communication of Speech Signal Using the Discrete Cosine Transform Based on Hyperchaos-System.
DOI: 10.5220/0012091600003543
In Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2023) - Volume 2, pages 69-76
ISBN: 978-989-758-670-5; ISSN: 2184-2809
Copyright © 2023 by SCITEPRESS Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
69
cal chaotic systems, thus proving the exploitation and
efficient application of chaotic systems in communi-
cations (Pecora and Carroll, 1990). In 1993, K.M.
Cuomo et al. proved that for some chaotic systems,
Lorenz for example, the synchronization property, ac-
cording to the technique of L.M. Pecora et al. is ro-
bust despite small perturbations in the coupling sig-
nal. This property allowed these researchers to pro-
pose the first scheme of a chaotic communication sys-
tem based on the principle of Chaotic Addition Mask-
ing (Cuomo and Oppenheim, 1993).
In 2003, A.S. Dmitriev developed a new ap-
plication base-chaos in communications called Di-
rect Chaotic Communication (DCC) (Dmitriev et al.,
2003). This new chaotic-based technology has
opened a new field of application of chaos, namely
ultra-wideband communications. In 1997, T. Yang et
al. published the first work presenting the concept of
chaotic encryption systems (Yang and Chua, 1997).
In this context, they proposed to integrate cryptogra-
phy into chaotic communication systems to improve
the security degree of the latter. Chaotic pseudo-
random sequences have desirable cryptographic prop-
erties, such as good randomness, deterministic dy-
namics, structure complexity, and sensitivity to ini-
tial conditions. Therefore, cryptanalysts have adopted
several cryptographic standards based on chaos the-
ory.
Subsequently, P.G. Vaidya et al. proposed chaotic
cryptography with chaotic timing in 1998 (Vaidya and
Ronge, 1998). The chaos-based digital cryptography
system, based on pseudo-random number generators,
does not depend on the chaos timing; it uses the initial
conditions and control parameters as the secret key
(Li et al., 2001). The one-dimensional logistic map
is often used as a Pseudo-Random Number Generator
(PRNG) in encryption, for example M.H.A. Samah
et al. introduced a method where the logistic map is
used to scatter the samples and the one-dimensional
circular map is used to confuse the samples (Samah
and Eihab, 2013). E. Mosa et al. (Mosa et al., 2009)
presented cryptography in the transformation domain,
based on the two-dimensional Baker map. The three-
dimensional Lorentz map as a pseudo-random num-
ber generator is discussed by B.S. Sattar (Sadoudi and
Azzaz, 2009).
In 2009, S. Sadoudi and M.S. Azzaz, developed
a hardware implementation of the Rossler chaotic
system to secure communication (Sadoudi and Az-
zaz, 2009) (Sattar and Rana, 2015), followed by a
new auto-switched chaotic system and its Field Pro-
grammable Gate Array (FPGA) implementation is
presented in (Azzaz et al., 2013b), as well as some
work based on synchronized hybrid chaotic genera-
tors is proposed in (Azzaz et al., 2013a). In the same
year, S. Sadoudi et al. developed a wireless hyper-
chaotic communication system for secure real-time
image transmission (Sadoudi et al., 2013). Subse-
quently, in the same year, they proposed an FPGA
real-time implementation of Chen’s chaotic system
to secure chaotic communications (Sadoudi et al.,
2009). Several works are proposed founded on real-
time FPGA implementation of Lorenz’s chaotic gen-
erator, Duffing’s chaotic attractor and experimental
synchronization technique for chaotic communica-
tions (Azzaz et al., 2009), (Sadoudi et al., 2014) and
(Sadoudi et al., 2015).
Currently, from 2020 to 2022, several researchers
have been working on implementing FPGA-based
real-time chaos secure encryption systems, as well
as chaos-based video encryption algorithms (Azzaz
et al., 2019)-(Hadjadj et al., 2022). Secure commu-
nication is crucial in the digital age, with data pri-
vacy and confidentiality being paramount concerns.
One promising method for encrypting speech sig-
nals is the use of the Discrete Cosine Transform
(DCT) based on hyperchaos-system. Hyperchaos-
system employs hyper-chaotic systems, which are
even more unpredictable than traditional chaotic sys-
tems, adding an extra layer of security to the encryp-
tion process (Zghair et al., 2021). Combining the
DCT with hyperchaos-system allows efficient encryp-
tion of speech signals while ensuring accurate recon-
struction at the receiver’s end. This novel approach
presents an effective solution for secure speech sig-
nal communication (Sathiyamurthi and Ramakrish-
nan, 2022).
In this paper (Abdallah and Meshoul, 2022), au-
thors propose multilayer cryptosystems for audio
communication encryption by continuously fusing
audio signals with speech signals without silent pe-
riods. Three levels of encryption (fusion, substitu-
tion, and permutation) are considered, and the pro-
posed approach shows increased security compared to
one-dimensional logistic map-based encryption tech-
niques.
Cryptography is a mathematical discipline allow-
ing one to perform operations on an intelligible text
to ensure some security properties of the informa-
tion. This work aims to secure speech communication
via an encryption system based on the chaotic PRNG
technique using a transmission channel-based USB-
RS232. Hybrid chaotic PRNG-based cryptography
is secure and has powerful confusion and diffusion
properties necessary for strong encryption. Each unit
in the confusion and scattering area wants to block
the derivation of the secret write key or the possible
prevention of the original message. While scattering
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
70
increases the repetition of the clear text over the en-
crypted text key to make it obscure, confusion is used
to make the invalid encrypted text. Only confusion
can make stream encryption work; otherwise, stream
encryption and block encryption require to scatter.
This paper is organized as follows: Section II de-
scribes hyper-chaotic systems and presents a hyper-
chaotic system based-speech encryption using a Dis-
crete Cosine Transformation. While, in Section III,
the speech encryption process is proposed. Then, in
Section IV, the simulations and experiment results are
presented. At last, Section V concludes the paper and
gives some future works.
2 HYPERCHAOTIC SYSTEM
BASED-SPEECH ENCRYPTION
A dynamical system can generate hyperchaos if it has
at least four state variables. It is modeled by at least
four nonlinear differential equations. The mathemati-
cal model of a nonlinear dynamical system represents
a four dimensional hyper chaotic system as follows:
x
t
= hy ax + yz
y
t
= hx by xz
z
t
= dz + xy + w
2
w
t
= xy + cw
(1)
where (x, y, z, w) are state variables and (a, b, c, d, h)
represent system parameters. The initial states and the
eventual control parameters show that this system is
chaotic as: x(0) = 0.8, y(0) = 0.3, z(0) = 10.1, w(0) =
4.5, a = 5, b = 0.5, c = 2.2, d = 1, h 4.85. The fi-
nal state of a chaotic system is extremely sensitive to
small changes in the initial state. With chaotic attrac-
tor, the phase diagram is the set of trajectories that are
the solutions of the hyper-chaotic system as shown
in Eq. (1), represented in the phase plane initiated
from different initial conditions. A chaotic switching-
rule allows the specified chaotic system to change its
behaviour, while still producing complicated chaotic
attractors. The properties of the chaotic system are
so interesting for data encryption. Thus, we have
developed an encryption system based on a hyper-
chaotic function. The hyperchaotic system is used as
a PRNG. Scattering of speech samples in this domain
is performed only by Discrete Cosine Transformation
(DCT) using PRNG, as shown in Figure 1. Before
the transmission of the encrypted speech signal using
DCT, it is converted into the time domain by the In-
verse Discrete Cosine Transformation (IDCT). In the
reception phase,, we will apply the inverse operation
of encryption to recover the original speech signal. It
is useful to note that the generation of the key stream
encryption is done by the same initial values, as well
as the same control parameters.
Figure 1: Encryption process flowchart.
Nowadays, telecommunication technologies are
involved everywhere. The large-scale use of mod-
ern speech communication technologies via the In-
ternet or mobile telephony requires securing informa-
tion exchanges. The concept of security involves per-
forming transformations on the original speech signal
to mask it, making it unintelligible to unauthorized
users, while ensuring the possibility of reconstituting
it for authorized users. To meet these requirements,
several methods of processing the speech signal have
been proposed in the literature.
In the case of key management for streaming
encryption, it has been generated using a four-
dimensional hyper-chaotic system as seen in Figure
2 which represents the simulation results. This figure
shows the change of variables over time, all variables
of the states whose starting point is the point men-
tioned in the initial conditions. These initial condi-
tions and parameters of the system are the key func-
tions, which are given as input to the hyper-chaotic
system. At the output of the PRNG, the state vari-
ables are given random values, and the normalization
operation of each key is presented as follows:
¯x =
x x
min
x
max
x
min
(2)
where x
min
is the minimum value of the generated key
and x
max
is the maximum value of the generated key
stream encryption. The normalized key is converted
to a 16-bit integer as follows:
˜x
i
= (round( ¯x
i
× 10
15
))mod 2
16
(3)
where i = {1, 2, 3, . .. , N
s
} and N
s
is the number of
samples of the speech signal.
Secured Communication of Speech Signal Using the Discrete Cosine Transform Based on Hyperchaos-System
71
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-10
-5
0
5
10
15
Normalized vectors
Normalized X-vector
Normalized Y-vector
Normalized Z-vector
Normalized W-vector
Figure 2: Normalized key to different variables x, y, z and w.
3 SPEECH ENCRYPTION
PROCESS
To perform this process, we have chosen an audio
recording of 4.5 seconds with a sampling rate of
8000 HZ. In the example illustrated in Figure 3 (a),
the original speech signal is recorded and then con-
verted to the Wav format using a time domain rep-
resentation. By definition, the audiogram graphically
represents the evolution of the capacity or intensity
of a speech to see its temporal envelope. A Dis-
crete Cosine Transform (DCT) describes a finite se-
quence of data points in terms of the sum of cosine
functions varying in frequency. It is used in signal
and image processing and especially in data compres-
sion. Indeed, it has an exceptional energy aggregation
property, and the low-frequency coefficients carry the
information. In this case, Figure 3 (b) shows the
original speech signal in transformation mode (Azzaz
et al., 2019). The chaotic signals in the process of
masking the speech signal are based on the streaming
process. At the output of the PRNG, the keys stream
encryption has given random values, followed by a
normalization operation for each key; after this op-
eration, each normalized key will be converted into
a 16-bit coded integer and the encryption operation
is performed by the XOR operator between the key
stream encryption and the transformed speech scram-
bles to completely encrypt the speech signal. How-
ever, before this operation, the transformed speech
scrambles are converted into equivalent decimal val-
ues using 16-bit quantization. The samples are pro-
cessed in a fixed-size block, where the block size de-
pends on the random key size (Sadoudi et al., 2015)
(Kaibou et al., 2021). The following process presents
the diffusion using the XOR function between each
number of samples of the intercepted speech signal
and the key converted to a 16-bit integer, as follows:
˜e
i
= ˜m
i
+ ˜x
i
(4)
where ˜x
i
presents the key converted to a 16-bit inte-
ger, ˜m
i
defines the samples of the original speech, and
˜e
i
presents the encrypted samples. If the correlation
Table 1: Correlation coefficient and SNR results.
Key Correlation
r
m,e
SNR (dB)
x(t) 0.000003713 -263.9514
y(t) 0.0003711 -263.9649
z(t) -0.002972 -263.9557
w(t) 0.003713 -263.9682
coefficient is zero, the original and encrypted signals
are different. The smaller the values of the correla-
tion coefficient, the better the cryptographic process.
It can be calculated as follows:
r
m,e
=
cov(m, e)
σ
m
σ
e
(5)
where cov(m, e) is the covariance between the original
signal m and the encrypted signal and σ
m
, σ
e
are the
standard deviations of the signal m and signal e.
The determination of the key is based on the best
performance in terms of the correlation coefficient.
The analysis of the correlation coefficient measures
the quality of the encryption system. By analysis, if
the correlation coefficient is equal to zero, the origi-
nal signal and the encrypted signal are considered to
be completely different. When the Signal-to-Noise
Ratio (SNR) is a negative value in dB, it means that
the signal intensity is lower than the noise intensity,
making the encrypted data not detectable, and the key
represents a good correlation (Hadjadj et al., 2022).
In this case, we compute the noise between the orig-
inal speech signal and the encrypted speech signal as
follows:
SNR = 10 × log10
N
s
i=1
m
2
i
N
s
i=1
(m
i
e
i
)
2
(6)
4 SIMULATION RESULTS AND
DISCUSSION
Several experimental analyses are performed to test
the effectiveness of the speech-based cryptography
technique. In this analysis, we will measure the qual-
ity of our cryptographic system through the analysis
of the coefficient. The latter is given by the correlation
between the similar segments in the original speech
signal and the encrypted one. Table 1 shows different
correlation coefficients for the different keys, and we
notice that the key x(t) converges to zero, which indi-
cates a good correlation. These results are close to
each other, but we choose the key having the small-
est negative value, which is x(t). Self-correlation is
a mathematical method to analyze a periodic signal
using the cross-correlation of a signal with itself (see
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
72
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-1
-0.5
0
0.5
1
Magnitude (v)
Input speech
(a) Audiogram of the original speech.
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-15
-10
-5
0
5
10
15
Magnitude (v)
Transformed speech
(b) Original speech using a discrete cosine transformation.
Figure 3: Original speech and its DCT over time.
Figure 4: Self-correlation analysis of the key x(t).
Figure 4). The proposed key flow gave better corre-
lation values. After selecting the encryption key, the
next step is to use this key in masking the speech sig-
nal using the diffusion process. An example of a tem-
poral representation of an encrypted signal is shown
in Figure 5, which is almost impossible to identify the
transmitted speech signal. The obtained results from
the simulation demonstrate the disappearance of the
speech signal completely, which proves the success
of the function of the encryption process. The advan-
tages of chaotic masking based on diffusion are the
simplicity of its realization and the possibility of us-
ing it to mask analogue or digital signals with high
SNR channels. As far as the security analysis of the
speech signal is considered, we will test the efficiency
and robustness of our cryptographic system for the
encryption of a speech signal by analyzing the two
signals presented in Figure 6. It is a measure of the
similarity of the signal to itself over time. Typically,
the correlation is maximal for a zero-time delay. The
principle of correlation is to extract future predictive
information from the predicted signal values. In this
part, we tested the efficiency and robustness of our
adopted encryption system. From Figure 7, we can
notice that the autocorrelation of the encrypted signal
is like white noise converging to zero. Therefore, pre-
dicting the signal from its predicted state is not possi-
ble. Figure 8 illustrates the diffusion of information in
the crypto-system. The histogram is a statistical study
tool that is used to illustrate the efficiency of the dif-
fusion of information in the encrypted data brought
by the used cryptosystem. As for the original signal
history, it is presented in the form of a stripe, as well
as with a uniformly presented encrypted signal. In
the case of the spectrogram analysis signal, the anal-
ysis of the time-frequency distribution of the original
speech signals is shown in Figure 9, along with an en-
Table 2: Perceptual evaluation of speech quality based tech-
nique.
Original
speech
Encrypted
speech
Decrypted
speech
4.176 1.260 3.952
crypted signal that has a uniform frequency behavior
over time. In the encryption process, the encrypted
signal in the transformation domain is converted into
the time domain by the IDCT and in the decryption
process step, the speech signal is recovered at the re-
ceiver. The latter must generate the same key stream
encryption with the same initial values and control pa-
rameters to restore the original speech signal. The
received signal is transformed into a frequency rep-
resentation using DCT. As for the diffusion opera-
tion, it is performed between the samples of the en-
crypted signal and its counterpart of the encryption
key. These encrypted samples are converted to equiv-
alent 16-bit decimal values. Note that the key is a 16-
bit encoded integer, and the XOR operation between
the encrypted samples and the encryption key yields
the original signal in its frequency representation, as
shown in Figure 10 (a). To recover the original sig-
nal, a transformation from the frequency to the time
domains is necessary using the inverse DCT, as shown
in Figure 10 (b).
In the decryption process step, the signal recov-
ery at the receiving end is still with a scattering op-
eration, but this time it is performed between the en-
crypted signal samples and its counterpart of the en-
cryption key. These encrypted samples are converted
into equivalent 16-bit decimal values.
Figure 11 is the recovered speech signal in the
time domain. According to this example, the recon-
structed speech signal corresponds well to the original
transmitted original signal. We can see that the orig-
inal speech signal has been recovered successfully in
terms of audio quality with its appropriate character-
istics, as seen in Figure 3. Table 2 presents the Percep-
tual Evaluation of Speech Quality (PESQ) based eval-
uation technique. The results for the PESQ are almost
similar to the original speech, which shows the perfor-
Secured Communication of Speech Signal Using the Discrete Cosine Transform Based on Hyperchaos-System
73
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
0
1
2
3
4
5
6
7
Magnitude (v)
10
4
Encrypted speech
(a) Encrypted speech using DCT.
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-1
0
1
2
3
4
5
6
7
Magnitude (v)
10
4
Encrypted speech transformation
(b) Encrypted speech using IDCT.
Figure 5: Encrypted speech presentation using transformation process.
0 1 2 3 4
Algorithm iterations
10
5
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Normalized magnitude
Original speech signal
0 1 2 3 4
Algorithm iterations
10
5
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Normalized magnitude
Encrypted speech signal
Figure 6: Speech encryption normalization.
-5 0 5
Delay time (s)
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Autocorrelation
Speech autocorrelation
-5 0 5
Delay time (s)
-0.2
0
0.2
0.4
0.6
0.8
1
Autocorrelation
Speech autocorrelation
Figure 7: Presentation of speech signal autocorrelation.
Original speech
-1 -0.5 0 0.5 1
Magnitude
0
500
1000
1500
2000
2500
3000
3500
4000
Number of samples
Speech distribution probability
Normal distribution
Encrypted speech
-1 -0.5 0 0.5 1
Magnitude
0
100
200
300
400
500
600
700
800
900
Number of samples
Speech distribution probability
Normal distribution
Figure 8: Histogram analysis.
Original speech spectrogram
2 4 6 8
Time (s)
0
5
10
15
20
Frequency (Hz)
-140
-120
-100
-80
-60
-40
Magnitude (dB)
Encrypted speech spectrogram
2 4 6 8
Time (s)
0
5
10
15
20
Frequency (Hz)
-140
-120
-100
-80
-60
-40
Magnitude (dB)
Figure 9: Spectrogram analysis presentation.
mance of our proposition. We used the PESQ method
for evaluation in our simulation. Based on PESQ
analysis, the results shown in Table 2 depict the rate
of loss of the frames. The PESQ scale (PESQ > 3),
which offers good speech quality and high accuracy,
preserves the intelligibility of the signal. The PESQ
score obtained between the original and the encrypted
speech is lower than the value of 2, which is consid-
ered satisfactory.
As for the realization of secure speech com-
munication, speech encryption is done directly in
real-time using a USB-RS232 serial communication
converter. Figure 12 illustrates this implementa-
tion via this transmission channel. In this part,
an encryption device is used to transmit speech via
two PCs via a graphical interface which is con-
sidered as a transmission device for recording and
processing the speech signal which will be inter-
cepted by a decryption device with a reverse oper-
ation, as shown in Figure 13. In the emission pro-
cess, we used an HP-PC with Pentium (R) Dual-core
CPU T 4500 @ 2.30 GHz, 2.00 GB RAM and 64-bit
operating system with windows10, while in the recep-
tion process, we used a DELL-PC with Pentium (R)
Dual-core CPU T 4500 @ 2.30 GHz, 2.00 GB RAM
and 32-bit operating system with windows7.
5 CONCLUSIONS
This study advances secure communication through
hyper-chaotic systems, leveraging chaotic behavior
for effective speech signal encryption using logistic
and sinusoidal maps. Validation via stringent trans-
mission tests on encrypted speech using RS232 pro-
tocol between two PCs demonstrates strong encryp-
tion efficiency, signal-to-noise ratio, and bit error rate
performance, confirming the method’s robustness in
securing speech communication.
Our work’s utilization of hyper-chaotic systems
and its successful implementation of real-time en-
cryption signify a compelling step towards enhanc-
ing speech communication security. The integration
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
74
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-1
0
1
2
3
4
5
6
7
Magnitude (v)
10
4
Encrypted speech transformation using DCT
(a) Encrypted speech using DCT.
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-15
-10
-5
0
5
10
15
Magnitude (v)
Decrypted speech transformation using IDCT
(b) Decrypted speech using IDCT.
Figure 10: Encrypted and Decrypted speech presentation using transformation process.
Figure 13: Graphical interface for transmission and reception of encrypted speech.
0 0.5 1 1.5 2 2.5 3 3.5 4
Algorithm iterations
10
5
-1
-0.5
0
0.5
1
Magnitude (v)
Final step with decrypted speech
Figure 11: Final step with the decrypted speech signal.
Figure 12: Experimental validation of the proposed ap-
proach of secured speech.
of chaotic systems in this context paves the way for
further advancements and applications in secure com-
munication methods. This research opens up exciting
possibilities for developing novel and practical solu-
tions to safeguard communication in various domains.
In addition, future work could extend the proposed
approach beyond speech signals to other types of sig-
nals, such as images and videos. Adapting the algo-
rithm for different signal types would allow for ad-
dressing security concerns in various domains and ap-
plications.
Moreover, implementing the algorithm on differ-
ent hardware platforms, such as FPGAs and micro-
controllers, will be a key focus to achieve real-time
performance and ensure portability for practical de-
ployments. This will allow the method to be inte-
grated into various systems and devices, enhancing
its usability in real-world scenarios.
Additionally, exploring the design space of hyper-
chaotic systems is planned to further optimize the pro-
posed algorithm. Evaluating its performance in terms
of security and computational complexity under vari-
ous conditions will help in understanding its strengths
and limitations. This evaluation will be crucial in as-
sessing the potential advantages of the method com-
pared to existing solutions.
To facilitate a comprehensive evaluation, per-
formance metrics and execution times will be pro-
vided. These measurements will enable researchers
and practitioners to assess the efficiency and effec-
tiveness of the proposed method for various applica-
tions and scenarios. Making informed decisions about
the best implementation approach will be essential for
successful deployment. Investigating error correction
codes and compression techniques will enhance the
system’s robustness and signal integrity during trans-
mission and storage.
REFERENCES
Abdallah, H. and Meshoul, S. (2022). A multilayered audio
signal encryption approach for secure voice commu-
nication. pages 1–18.
Secured Communication of Speech Signal Using the Discrete Cosine Transform Based on Hyperchaos-System
75
Azzaz, M., Tanougast, C., and Benssalah, A. (2019). An ef-
ficient and lightweight multi-scroll chaos-based hard-
ware solution for protecting fingerprint biometric tem-
plates. pages 1–13.
Azzaz, M., Tanougast, C., Sadoudi, S., and Bouridane,
A. (2013a). Synchronized hybrid chaotic generators:
Application to real-time wireless speech encryption.
pages 2035–2047.
Azzaz, M., Tanougast, C., Sadoudi, S., and Dandache, A.
(2009). Real-time fpga implementation of lorenz’s
chaotic generator for ciphering telecommunications.
In Proceedings of Joint IEEE North-East Workshop
on Circuits and Systems and TAISA Conference, pages
1–4.
Azzaz, M., Tanougast, C., Sadoudi, S., Fellah, R., and Dan-
dache, A. (2013b). A new auto-switched chaotic sys-
tem and its fpga implementation. pages 1792–1804.
Cuomo, K. and Oppenheim, A. (1993). Circuit implementa-
tion of synchronized chaos with applications to com-
munications. In Physical Review Letters.
Dmitriev, A., Kyarginsky, B., Panas, A., and Starkov, S.
(2003). Experiments on direct chaotic communica-
tions in microwave band. pages 1495–1507.
Hadjadj, M., Sadoudi, S., Azzaz, M., Bendecheche, H., and
Kaibou, R. (2022). A new hardware architecture of
lightweight and efficient real-time video chaos-based
encryption algorithm. pages 1–14.
Kaibou, R., Azzaz, M., Benssalah, M., Teguig, D., Hamil,
H., Merah, A., and Akrour, M. (2021). Real-time
fpga implementation of a secure chaos-based digital
crypto-watermarking system in the dwt domain using
co-design approach.
Li, S., Mou, X., and Cai, Y. (2001). Pseudo-random Bit
Generator Based on Couple Chaotic Systems and Its
Applications in Stream-Cipher Cryptography. Lecture
Notes in Computer Science book series, London.
Mosa, E., Messiha, N., and Zahran, O. (2009). Chaotic
encryption of speech signals in transform domains.
In Proceedings of IEEE International Conference on
Computer Engineering and Systems, Cairo, Egypt.
Pecora, L. and Carroll, T. (1990). Synchronization in
chaotic systems. In Physical Review Letters.
Sadoudi, S. and Azzaz, M. (2009). Hardware implementa-
tion of the rossler chaotic system for securing chaotic
communication. In Proceedings of 5th International
Conference of Sciences of Electronic, Technologies of
Information and Telecommunications.
Sadoudi, S., Azzaz, M., Djeddou, M., and Benssalah, M.
(2009). An fpga real-time implementation of the
chen’s chaotic system for securing chaotic communi-
cations. pages 467–474.
Sadoudi, S., Azzaz, M., and Tanougast, C. (2014). Novel
experimental synchronization technique for embed-
ded chaotic communications. In Proceedings of IEEE
International Conference on Control, Decision and
Information Technologies, pages 669–672.
Sadoudi, S., Tanougast, C., and Azzaz, M. (2015). Journal
of engineering science and technology review.
Sadoudi, S., Tanougast, C., Azzaz, M., and Dandache, A.
(2013). Design and fpga implementation of a wireless
hyperchaotic communication system for secure real-
time image transmission. pages 1–18.
Samah, M. and Eihab, B. (2013). Speech scrambling based
on chaotic maps and one time pad. In Proceedings of
IEEE International Conference on Computing Electri-
cal and Electronics Engineering, Khartoum, Sudan.
Sathiyamurthi, P. and Ramakrishnan, S. (2022). Speech
encryption using hybrid-hyper chaotic system and bi-
nary masking technique. pages 6331–6349.
Sattar, B. and Rana, S. (2015). A proposed voice encryp-
tion based on random lorenz map with dct permuta-
tion. page 90.
Vaidya, P. and Ronge, H. (1998). Implementation of chaotic
cryptography with chaotic synchronization. In Physi-
cal Review Letters.
Yang, T. and Chua, L. (1997). Impulsive stabilization for
control and synchronization of chaotic systems: The-
ory and application to secure communication.
Zghair, H. K., Mehdi, S., and Sadkhan, S. (2021). Speech
scrambler based on discrete cosine transform and
novel seven-dimension hyper chaotic system. pages
1–14.
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
76