Phys. Scr. 99 (2024) 125008 https://doi.org/10.

1088/1402-4896/ad87c5

PAPER

Image encryption algorithm based on matrix projective combination-

RECEIVED
18 August 2024
combination synchronization of an 11-dimensional time delayed
REVISED
4 October 2024
hyperchaotic system
ACCEPTED FOR PUBLICATION
16 October 2024
PUBLISHED
Jyotsna Kumari Bharti1 , P Balasubramaniam2 and K Murugesan1,∗
4 November 2024 1
Department of Mathematics, National Institute of Technology, Tiruchirappalli, Tiruchirappalli, 620015, Tamil Nadu, India
2
Department of Mathematics, The Gandhigram Rural Institute (Deemed to be University), Gandhigram, Dindigul, 624302, Tamil Nadu,
India
∗
Author to whom any correspondence should be addressed.
E-mail: [email protected], [email protected] and [email protected]

Keywords: hyper-chaotic system, time delay, matrix projective combination-combination synchronization, nonlinear active controller,
image encryption

Abstract
Secure image transmission is critical to protect sensitive data from unauthorized access, especially in
an era of increasing digital threats. Chaotic systems with their inherent complexity and unpredict-
ability, provide a promising solution for enhancing encryption security. To contribute to this field, we
investigate a new 11-dimensional hyperchaotic system by taking advantage of its complex dynamical
properties to strengthen security. The high dimensional of the system intensifies chaotic behaviors
such as stability, attractors and sensitive to initial conditions, making it particularly suitable for
encrypted transmission. Time delay is an important factor to be considered affecting the control and
synchronization in nonlinear system. Additionally, time delays include the effects of past states,
further increasing the unpredictability of the system. To explore these dynamics, we analyze the
Lyapunov exponents, stability of equilibrium points, symmetry and dissipation. A matrix projective
combination-combination synchronization scheme is proposed to synchronize four identical
11-dimensional hyperchaotic systems with time delay. Nonlinear active controllers designed based on
Lyapunov stability theory are used to achieve this synchronization. This work advances an important
idea for encryption and decryption algorithms, which is the secure transmission of images using affine
encryption. In the affine encryption algorithm, the key is based on the solution of synchronized
chaotic delayed systems and the private message of the sender and receiver. This proposed encryption
and decryption algorithms have been applied on plain images. Numerical simulations and security
analysis including key space, histogram, information entropy and correlation analysis are conducted
to validate the theoretical results and encryption algorithm. Experimental analysis and comparisons
with existing literature confirm the effectiveness and security of the proposed approach for
cryptographic purposes.

1. Introduction

Chaos control and synchronization are important research areas due to the unique challenges of chaotic
systems, which exhibit sensitivity to initial conditions, boundedness and unpredictable behavior. Since Pecora
and Carroll [1] first observed chaotic synchronization in 1990, it has been applied in areas such as
communication encryption, engineering and statistical prediction.
Synchronization occurs when nonlinear systems match their outputs by adjusting dynamic parameters with
types such as complete synchronization [2], anti-synchronization [3], generalized synchronization [4], lag
synchronization [5] and projective synchronization [6]. Various techniques such as active control [7], sliding
mode control [8], adaptive control [9], backstepping control [10], linear and nonlinear feedback control [11]

© 2024 IOP Publishing Ltd. All rights, including for text and data mining, AI training, and similar technologies, are reserved.
Phys. Scr. 99 (2024) 125008 J K Bharti et al

and time delay feedback [12] have been successfully employed in chaotic systems, especially in secure
communication and cryptography. Additionally, nonlinear systems present considerable challenges in dynamic
control, especially when time delays are involved, since these delays create dependencies on both present and
past conditions. This makes them important considerations in communications, electronics, hydraulics and
other fields. The inherent non linearity of these systems increases their sensitivity to initial conditions and
introduces non periodicity which are key factors in the development of effective control techniques both
theoretically and practically [13]. Many nonlinear systems exhibit chaotic behavior making chaos theory
particularly useful for addressing these challenges.
Despite these complexities, chaotic systems have proven highly effective for image encryption, where their
unpredictability and complexity make them ideal for cryptographic applications. In the digital age, protecting
sensitive information in digital images is important due to their susceptibility to unauthorized access [14, 15].
Chaos theory has enabled secure image encryption, ensuring that images are securely transmitted and stored
through sequences that are resistant to decryption attacks [16, 17]. This breakthrough has stimulated significant
research interest , as chaotic encryption provides both speed and security , allowing encrypted images to resist
attacks while enabling fast and efficient transmission.
Among the various types of chaotic synchronization, projective synchronization has recently attracted
significant attention due to its effectiveness in secure communication systems, especially in image encryption.
Projective synchronization introduced by Mainieri and Rehaceh [6], synchronizes chaotic systems with different
dynamics increasing the security in the transmitted images via a scaling factor α between master and slave
systems. Different forms of projective synchronization such as hybrid projective synchronization [18], function
projective synchronization [19], quasi projective synchronization [20] and matrix projective synchronization
[21]. Many studies [22–27] have been carried out on different forms of synchronization achieved through
different control strategies particularly in the context of secure communication and cryptography applications.
An adaptive fuzzy control method for synchronizing uncertain fractional order chaotic systems with input
saturation and external disturbances is achieved in [22] using a Takagi-Sugeno fuzzy model. The approach
guarantees parameter convergence and bounded closed loop signals, which is not suitable for fractional order
chaotic systems with time delay. The exponential synchronization of quaternion-valued neural networks with
time varying delays is examined in [23] using matrix measure techniques and state feedback controllers. Stability
criteria are established without assuming differentiable activation functions and numerical examples confirm
the theoretical findings and demonstrate quaternion-valued neural networks capability to restore real color
image patterns. Research by Razmara et al [24] involved a novel audio cryptography system using sliding mode
synchronization which enables secret voice messages through unsecured communication channels. In [25],
a 3D chaotic system with strong nonlinear behavior, high sensitivity and stability was introduced. They also
designed a new synchronization technology and a secure information communication method using a masking
method. In [26], Yao et al found a memristor based sixth order chaotic circuit focusing on dynamic analysis and
sliding mode synchronization. They applied this system to image security, demonstrating its effectiveness and
anti interference properties. In [27], the authors explore triple compound synchronization, a new type of
synchronization among chaotic systems with external disturbances. This complex synchronization enhances
communication security through strong signals, using nonlinear control functions and Lyapunov stability
theory. These works demonstrate the potential of synchronization in enhancing encryption schemes and
ensuring secure data transmission.
Matrix projective synchronization extends the technique by expanding the scaling factor into a matrix,
thereby increasing the complexity and security of the synchronization process. This method is particularly useful
in image encryption, making it more difficult for attackers to decode encrypted images and providing resilience
for secure image transmission in chaotic cryptographic systems. Moreover, earlier research mainly focused on
the master slave synchronization model, where a single master system is synchronized with a single response
system. To address the need for higher security, combination synchronization was developed. It enables the
synchronization of multiple drive and response systems, allowing for the simultaneous transmission of multiple
encrypted image signals while isolating and modulating each signal independently, thus increasing security and
flexibility. Building on this, combination-combination synchronization was introduced, synchronizing
multiple combinations of drive systems with multiple combinations of response systems. This approach further
increases the complexity of the system, making it highly effective in securing image encryption against attacks.
Matrix projective combination-combination synchronization combines the strengths of matrix projective
synchronization and combination-combination synchronization. It synchronizes multiple combinations of
drive and response systems using matrix scaled parameters, providing robust, high security image encryption.
This technique enables fast and secure transmission of encrypted images while maintaining strong resistance to
attacks, making it ideal for high security applications.
Various studies have been carried out based on the projective synchronization, matrix projective
synchronization, combination synchronization and combination-combination synchronization of chaotic

2
Phys. Scr. 99 (2024) 125008 J K Bharti et al

systems, many of which focus on their application in image encryption. Some of these studies are discussed as
follows: a new scheme for dual combination synchronization of fractional order complex chaotic systems using
Lyapunov stability theory to synchronize four master and two slave systems is presented in [28]. It shows that
synchronization is achieved more quickly when accessing fractional order from standard order. Pan et al [29]
explored multi-switching sliding mode combination synchronization of fractional order non-uniform chaotic
systems with unknown parameters and double stochastic disturbances. They proposed fractional order sliding
surfaces, multi-switching adaptive controllers and multi-switching adaptive update laws to synchronize drive
system state variables with response system state variables. They showed that multi-switching adaptive
controller is effective and robust because the error variables converge quickly to the equilibrium point. In [30],
authors demonstrates the potential of optimizing synchronization control in combination-combination
schemes with minimal control functions, offering practical implementation advantages and cost effectiveness.
In [31], a three-dimensional chaotic system with quadratic non linearity is analyzed using various methods.
Also, the effectiveness of the system is demonstrated through the study of triple compound combination
synchronization and Lyapunov stability theory. In [32], an adaptive control for synchronizing time delayed
chaotic systems, establishing conditions for adaptive projective synchronization is explored. It designs an
adaptive synchronization controller using Lyapunov-Krasovskii stability theory, showcasing error reduction
towards zero for global and asymptotic synchronization. A full states pseudo-random projective
synchronization scheme, using pseudo random sequence as scaling factors and demonstrates its application
with the hyper-chaotic Lorenz system is proposed in [33]. A novel secure communication algorithm based on
full states pseudo-random projective synchronization is developed offering high security, robustness and
scalability for encrypting various information signals, including color images, text, audio and videos. The finite
time and fixed time full state projective synchronization of cellular neural networks with time delays and noise
establishing criteria and estimating synchronization times using inequality techniques and appropriate
controllers is addressed in [19]. In [34], fixed time stabilization and fixed time projective synchronization for
distributed delay in inertial neural networks were investigated and two feedback controllers and useful
Lyapunov functions were introduced to directly achieve the desired goal. In [35], a class of complex network
models with general structures were studied, including external projection synchronization and finite time
synchronization. A mixed control strategy with an adaptive law was presented, thereby simplifying complex
synchronization problems. In [36], Yadav et al presented a unique class of chaotic/hyperchaotic systems,
showing how projective and hybrid projective synchronization schemes can synchronize them. They used
hyperchaotic Lorenz fourth order and chaotic Pehlivan third order systems, demonstrated flexibility, adaptable
control and smooth synchronization curves. The matrix projective synchronization between chaotic time delay
systems with external disturbance and hyperbolic non linearity, achieving a synchronization between these
systems is studied in [21]. The fixed time adaptive time varying matrix projective synchronization in time
delayed different dimensional chaotic system with unknown parameters was investigated by Zheng et al [37].
An adaptive controller and laws of updating parameters are designed, which allow advance estimation of the
settlement time in fixed time synchronization. Despite the complexity and unpredictability of dual function
matrix projective synchronization, a new active control scheme is proposed in [38] to implement dual function
matrix projective synchronization between two pairs of fractional order hyperchaotic systems that enhance
information security and signal privacy. The principle of hybrid projective combination difference
synchronization for hyperchaotic complex time delay systems has been designed in [39] using adaptive control
techniques. The system synchronizes between two identical Lü master and Lü slave systems, stabilizes error
states and delay differential equations and offers exceptional communication and image encryption
applications.
Recent advances in chaotic systems have led to the development of numerous innovative image encryption
algorithms. Ongoing research continuously brings forth new strategies that use the specific feature of chaos to
enhance security in image transmission [40–43]. In [40], researchers presented a robust and efficient encryption
algorithm for grayscale images, combining fractional order four dimensional hyperchaotic system, two
dimensional henon chaotic map permutation and knight tour algorithm. This approach enhanced security by
minimizing pixel correlation and providing a large key space, making it highly resistant to brute force attacks.
The algorithm demonstrated strong computational efficiency; however, its application was limited to grayscale
images. In [41], Yousfi et al investigated a novel linear control approach for synchronizing fractional-order
Halvorsen circulant systems using the adomain decomposition method algorithm. They explored the potential
of this synchronization technique for image encryption, emphasizing its high security, key sensitivity, robustness
and compatibility. In [42], Li et al developed a two dimensional hyperchaotic map and two dimensional
enhanced logistic modular map. Its simple structure, wide range of hyperchaos and excellent chaotic
performance make it particularly suitable for image encryption. They introduced chaotic image encryption
scheme dynamic vector enhanced map, a novel encryption scheme consisting of eight steps, including key
stream generation, hash value stacking, dynamic binary diffusion and dynamic binary scrambling. They

3
Phys. Scr. 99 (2024) 125008 J K Bharti et al

demonstrated plaintext sensitivity and superior encryption efficiency. In [43], Feng et al developed the multi
image encryption algorithm using a fractional order three dimensional Lorenz chaotic system and two
dimensional sinusoidally constrained polynomial hyperchaotic map. This algorithm enhances security and
expands key space and includes chaotic sequence generation, multi channel fusion, plaintext related parameter
generation and two rounds of substitution, diffusion and scrambling.
In the context of securing image encryption algorithms, it is necessary to evaluate their resistance to various
types of attacks. One way to assess this resistance is through cryptanalysis. In cryptanalysis research [44, 45], it
has been indicated that while this algorithm has shown some effectiveness in various tests, it remains vulnerable
to plaintext attacks due to its limited key space and lack of key variability. Additionally, the deterministic nature
of the algorithm may allow attackers to exploit patterns, further compromising its security.
Our focus is to address security challenges to create more secure environment for image transmission,
ultimately enhancing the data integrity and confidentiality of the data. To achieve this, we propose a strong
image encryption algorithm by combining the solution of matrix projective combination-combination
synchronization of hyperchaotic system and the private message of the sender and receiver. Through in depth
security analysis, we ensure that this approach provides better protection against potential vulnerabilities. The
aim of this work is to contribute valuable insights and practical solutions for the development of secure image
encryption algorithms. The major contributions outlined in this paper are as follows:

• High dimensional hyperchaotic systems exhibit diverse dynamical behaviors and emergent patterns due to
their nonlinear dynamics and high dimensionality, making them attractive subjects for chaos theory and
nonlinear dynamics research. The high dimensional hyperchaotic systems are used in secure communication,
including image encryption, video encryption, text encryption and voice encryption due to its large key space
and enhanced security features. As a result, a new 11-dimensional hyperchaotic system has been constructed
by integrating complex variables into a 6-dimensional hyperchaotic system. Adding complex variables to a
6-dimensional hyperchaotic system can increase the complexity and unpredictability of the system,
potentially leading to more robust encryption schemes and increased resistance against decryption attacks.
• Next, we are incorporating time delay into our 11-dimensional hyperchaotic system, allowing past states to
influence current dynamics. This addition introduces a temporal dimension, resulting in more complex
attractive structures longer transient behaviors and increased resilience against disturbances. Time delays
enrich the dynamics of the system, increasing its potential for secure communications and encryption
applications.
• This study proposed four 11-dimensional identical chaotic system with time delay, where two systems act as
drive system and two as response system. A matrix projective combination-combination synchronization
strategy is used to synchronize these systems, ensuring that the response system tracks the state of the drive
system despite the time delay. A nonlinear active controller is designed that generates control signals for
synchronization and the Lyapunov stability theory guarantees system stability and robustness. This strategy is
promising for applications in secure communications, chaos based cryptography and chaotic control systems.
• The matrix projective combination-combination synchronization of hyperchaotic systems with time delay
enhances the resistance against brute force and statistical attacks, making the encryption algorithm more
robust for image encryption. An affine cipher based image encryption strategy that takes advantage of these
synchronizing system is proposed, which involves a combination of individual private messages and
synchronized chaotic system solutions. This affine cipher exhibits high security due to its sensitivity to initial
chaotic conditions and unique key combination.

1.1. Paper organization

The remainder of the article is structured as follows: section 2 begins with the design and dynamic analysis of an
11-dimensional hyper-chaotic system. Section 3 covers the problem formulation, key definition, main theorems
and the synchronization mechanism used in the study. Section 4 presents numerical simulations to demonstrate
the effectiveness of the matrix projective combination-combination synchronization strategy. Section 5
highlights the numerical results and discussion obtained from the conducted experiments. Section 6 presents
the results of applying matrix projective combination synchronization to image encryption and the security
analysis of the encryption algorithm. Finally, section 7 concludes the study by summarizing the key findings and
implications.

4
Phys. Scr. 99 (2024) 125008 J K Bharti et al

1.2. Notation
In this entire article, n represents an n-dimensional vector space over the real numbers. It is the set of all n
tuples of real numbers, essentially an n-dimensional column vector. n ´ n represents the space of n × n matrices
with real numbers as entries. It is the set of all n × n matrices where each element is a real number. ||. || represents
euclidean norm. The superscripts m1, m2, s1 and s2 represent the first master system, second master system, first
slave system, and second slave system, respectively. X tmi = X mi (t - t ), X tsi = X si (t - t ), i = 1, 2 and
Eτ = E(t − τ), τ is the time delay. The superscript  denotes the transpose of matrix. diag () represents the
diagonal matrix.

2. Design and dynamical investigation of a novel 11-dimensional hyperchaotic system

In this section, we extend the 6-dimensional hyperchaotic system introduced by Sabaghian et al [46] by
incorporating complex variables, transforming it into an 11-dimensional hyperchaotic system, which is
described as follows:

x = a ( y - x ) + u ,
y = cx - y - xz + v ,
z = - bz + xy ,
u = du - xz ,
v = - hy + w ,
w = - k1 x + k2 y , (1)

where x, y, z, u, v and w are state vectors and (a, b, c, d, h, k1, k2) are constant parameters of system (1). System (1)
exhibits hyperchaotic behavior when a = 10, b = 8/3, c = 28, d = 2, h = 8.8, k1 = 1 and k2 = 0.9.
System (1) can be written in complex form as follows:

x ¢ = a ( y ¢ - x ¢) + u¢ ,
y ¢ = cx ¢ - y ¢ - x ¢z ¢ + v ¢ ,
1
z ¢ = - bz ¢ + 2 (x ¢y ¢ + x ¢y ¢) ,
u ¢ = du¢ - x ¢z ¢ ,
v ¢ = - hy ¢ + w ¢ ,
w ¢ = - k1 x ¢ + k2 y ¢. (2)

where x ¢ , y ¢ , z ¢ , u¢ , v ¢ are state vectors, x ¢ = x1 + ix2, y ¢ = x3 + ix4, u¢ = x 6 + ix7, v ¢ = x8 + ix 9 and

w ¢ = x10 + ix11 are complex variables and z ¢ = x5 is real variable and i = -1 .
By separating real and imaginary parts, we get the real version of system (2) which is the new 11-dimensional
chaotic system given as:

x1 = a (x3 - x1) + x 6,

x2 = a (x 4 - x2) + x7,
x3 = cx1 - x3 - x1 x 5 + x 8,
x 4 = cx2 - x 4 - x2 x 5 + x 9,
x 5 = - bx 5 + x1 x3 + x2 x 4,
x 6 = dx 6 - x1 x 5,
x7 = dx7 - x2 x 5,
x 8 = - hx3 + x10,
x 9 = - hx 4 + x11,
x10 = - k1 x1 + k2 x3,
x11 = - k1 x2 + k2 x 4, (3)

where X = (x1, x2, x3,¼,x11) Î 11 are state variables. The parameter values are taken as a = 10, b = 3 , c = 28,
8

d = − 2, h = 8.8, k1 = 1, and k2 = 0.9, with the initial condition ( − 1, 1, − 1, − i, 1, − 1 + i). System (3) exhibits
chaotic behavior, as shown in figures 1 and 2. Figure 3 denotes the time response of the state vectors
x1, x2, x3, K, x11 for system (3).
Remark 1. We have plotted the time response and phase portraits for both system (1) and system (3) to
demonstrate the effect of introducing complex variables. The system (1), as presented in [46], serves as the basis
for comparison. A visual comparison of the time response and phase portrait of systems (1) and (3) indicates that
system (3) exhibits greater complexity and unpredictability. These differences are highlighted in figure 4.

5
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 1. Phase portrait of the 11-dimensional hyperchaotic system (3) with parameters (a, b, c, d, h, k1, k2) = (10, 8/3, 28, −2, 8.8,
1, 0.9).

2.1. Lyapunov exponent and Kaplan Yorke dimension

Lyapunov exponents are essential for characterizing the chaotic behavior of dynamical systems. It measures the
rate of divergence or convergence of nearby trajectories, with at least one positive exponent indicating sensitivity
to initial conditions, a defining characteristics of chaos. The fractal dimension, a characteristic feature of chaotic
systems, is calculated using the Lyapunov exponents through the Kaplan Yorke dimension [47] which is given as
m
Le i
K=m+ å ∣Le ,
i=1 m + 1∣

m+1
where each Lei denote Lyapunov exponent, m is the largest integer satisfying å i = 1 Le i  0 and å i = 1 Le i < 0.
m

The Lyapunov exponents for 11-dimensional system (3) are calculated as Le1 = 0.2033, Le2 = 0.1146,
Le3 = − 0.0124, Le4 = 0.0050, Le5 = − 0.0029, Le6 = − 0.3308, Le7 = − 1.4608, Le8 = − 1.9183, Le9 = − 3.1288,
Le10 = − 9.8466 and Le11 = − 36.3038. The system described by (3) has three positive Lyapunov exponents and
eight negative Lyapunov exponents is shown in figure 5. Therefore, the system (3) is hyperchaotic because it has
more than two positive Lyapunov exponents. By looking at the values of eleven Lyapunov exponents, the value
of m = 3 can be determined. The Kaplan Yorke dimension is calculated as

(0.2033 + 0.1146 + 0.0050)

K=3+ » 3.9761.
∣0.3308∣

Considering the Kaplan Yorke dimension, the value of fractal dimension is the fractional value. Therefore,
system (3) becomes hyperchaotic in nature.
Remark 2. System (1) exhibited four positive lyapunov exponents, confirming its hyperchaotic nature. In
contrast, system (3) formed by adding complex variables to a 6-dimensional system (1) exhibited three positive

6
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 2. Phase portrait of the 11-dimensional hyperchaotic system (3) with parameters (a, b, c, d, h, k1, k2) = (10, 8/3, 28, −2, 8.8,
1, 0.9).

lyapunov exponents. The retention of hyperchaotic behavior when extended to higher dimensions depends on
the specific parameter values and initial conditions. Although both systems (1) and (3) exhibited hyperchaotic
properties, the number of positive lyapunov exponents can vary with different parameter settings, as is typical in
chaotic systems.

2.2. Equilibrium points

The equilibrium points of the 11-dimensional system (3) will be obtained by solving the following equations:

a (x3 - x1) + x 6 = 0,
a (x 4 - x2) + x7 = 0,
cx1 - x3 - x1 x 5 + x 8 = 0,
cx2 - x 4 - x2 x 5 + x 9 = 0,
- bx 5 + x1 x3 + x2 x 4 = 0,
dx 6 - x1 x 5 = 0,
dx7 - x2 x 5 = 0,
- hx3 + x10 = 0,
- hx 4 + x11 = 0,
- k1 x1 + k2 x3 = 0,
- k1 x2 + k2 x 4 = 0. (4)

The system (3) has two equilibrium points as E1 and E2. Here, E1 = (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) is an isolated
equilibrium point and E2 is the whole circle of equilibria in (x3, x4) space given as
abdk1 (k2 - k1)
x 32 + x42 = .
k22

7
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 3. Time trajectory of the 11-dimensional hyperchaotic system (3).

Figure 4. Comparison of 6-dimensional system (1) in blue color and 11-dimensional system (3) in red color with parameters (a, b, c, d,
h, k1, k2) = (10, 8/3, 28, 2, 8.8, 1, 0.9): (a) Time trajectories (b) Chaotic attractors.

8
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 5. Trajectory of Lyapunov exponents for the 11-dimensional hyperchaotic system (3).

This equation represents a circle with centre (0, 0) and radius R2 = abdk1 (kk22 - k1) . Let x3 = Bc and x4 = Bs, where
2
Bc = R cos q and Bs = R sin q, θ ä [0, 2π], then we get the non-isolated fixed points as
E2 = (ABc, ABs, Bc, Bs, B, CBc, CBs, DBc, DBs, hBc, hBs) where A = kk2 , B = ad (k2k - k1) , C = a (k2k- k1)
1 2 1

and D = (1 - ad ) k1 +
k1
(ad - c ) k2
.
The Jacobian matrix of system (3) is obtained as follows:
⎡ -a 0 a 0 0 1 0 0 0 0 0⎤
⎢ 0 - a 0 a 0 0 1 0 0 0 0⎥
⎢c - x5 0 -1 0 - x1 0 0 1 0 0 0⎥
⎢ 0 c - x5 0 -1 - x2 0 0 0 1 0 0⎥
⎢ x x4 x1 x2 -b 0 0 0 0 0 0⎥
⎢ 3 ⎥
J = ⎢ -x5 0 0 0 - x1 d 0 0 0 0 0 ⎥.
⎢ 0 -x5 0 0 - x2 0 d 0 0 0 0⎥
⎢ ⎥
⎢ 0 0 -h 0 0 0 0 0 0 1 0⎥
⎢ 0 0 0 -h 0 0 0 0 0 0 1⎥
⎢ - k1 0 k2 0 0 0 0 0 0 0 0⎥
⎢ 0 - k1 0 k2 0 0 0 0 0 0 0⎥
⎣ ⎦

To examine the stability of trivial equilibrium point E1, the characteristic polynomial |J − λI| at E1 is

86 10 3601 9 118723 8 2050733 7 25210352 6 201962099 5

l11 + l - l - l + l + l + l
3 15 15 75 75 300
9633038 4 28525646 3 1947776 2 9364 32
+ l - l + l + l+ .
75 75 25 5 3

The characteristic equation is written as follows:

86 10 3601 9 118723 8 2050733 7 25210352 6 201962099 5

l11 + l - l - l + l + l + l
3 15 15 75 75 300 (5)
9633038 4 28525646 3 1947776 2 9364 32
+ l - l + l + l+ = 0.
75 75 25 5 3

On solving the above equation (5), we get the following eleven eigenvalues

l1 = 11.3453, l2 = 11.3453, l3 = 0.3497,

l4 = 0.3497, l5 = - 0.0111, l 6 = - 0.0111,
l7 = - 2.0000, l 8 = - 2.0000, l 9 = - 2.6667,
l10 = - 22.6839, l11 = - 22.6839.

9
Phys. Scr. 99 (2024) 125008 J K Bharti et al

2.3. Symmetry
The system (3) exhibits rotational symmetry about x5- axis under the coordinate transformation:

( - x1, - x2, - x3, - x 4, x 5, - x 6,¼, - x11) = (x1, x2, x3, x 4, x 5, x 6,¼,x11).

2.4. Dissipative analysis

Let ψ(t) be the domain in the smooth surface 11 and let ξ(t) be the volume of ψ(t).
dx (t )
dt
= òy (t ) (D.x ) dx1 dx2 dx3 dx 4 dx5 dx 6 dx7 dx8 dx 9 dx10 dx11. (6)

From equation (6), the dissipation of the chaotic system is calculated as follows:

¶x1 ¶x2 ¶x3 ¶x4 ¶x5 ¶x6 ¶x7 ¶x8 ¶x9 ¶x10
 ¶x
D.x = + + + + + + + + + + 11
x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11
b
= - 2a - 2 - b + 2d < 0 if d < 0 or d < a + 1 + .
2

8 -86
For a = 10, b = 3 and d = − 2, D.x = -2a - 2 - b + 2d = 3
< 0.
From equation (6), we get

dx (t ) - 86
= x (t ). (7)
dt 3

-86
( )
Solving (7), x (t ) = exp 3 x (0), if Δ. ξ < 0 then system (3) is dissipative. Hence, system (3) satisfies the
principle of hyperchaotic state.

2.5. Sensitivity analysis

We have investigated the sensitivity to initial conditions of the 11-dimensional hyperchaotic system (3) by
performing time series analysis. The time response of all state vectors is plotted under two different initial
conditions (ICs),
IC1 = [-1 0.1 1 0.1 -1 0.1 -1 1 0.1 -1 1] and
IC2 = [-1 0.5 1 0.5 -1 0.5 -1 1 0.5 -1 1], as shown in figure 6. Comparison of these time
responses demonstrates the sensitivity of the system to chaotic behavior, which is a hallmark of chaotic behavior,
further reinforcing the hyperchaotic nature of system (3).

3. Matrix projective combination-combination synchronization scheme

For practical purposes, system (3) can be rewritten as:

X (t ) = AX (t ) + F (t , X (t )) , (8)

where X (t ) = (x1 (t ), x2 (t ),¼,x11 (t )) Î 11 is the state vectors, A Î 11 ´ 11 is the linear parameter matrix and F
(t, X(t)) is the nonlinear term.Since, many real world systems are inherently associated with delays due to limited
signal transmission time, processing time or other lag effects. In nature, the deterministic model (8) can be
affected by delay effects, which can be captured by incorporating the model into delay differential systems
involving time delay. Hence, studying the hyperchaotic system by incorporating time delays may lead to more
accurate representations of such real world systems, thereby increasing the relevance and applicability of
theoretical studies. In the proposed work, the following delay differential equation is considered to analyze the
synchronization of the 11-dimensional hyperchaotic system. Let τ > 0 be the delay term.

X (t ) = AX (t ) + BX (t - t ) + F (t , X (t ) , X (t - t )) , (9)

where X(t) is the state vector of the system (9), A, B Î 11 ´ 11 are the linear parameter matrices. To clarify the
concept of matrix projective combination-combination synchronization, we consider system (9) as drive system
and rewrite it as follows:

X m1 (t ) = AX m1 (t ) + BX tm1 (t ) + F1 (t , X m1 (t ) , X tm1 (t )) , (10)

10
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 6. Sensitivity analysis of the 11-dimensional hyperchaotic system (3) under two initial conditions: IC 1 = (−1, 0.1, 1, 0.1, −1,
0.1, −1, 1, 0.1, −1, 1) (in blue solid lines) and IC 2 = (−1, 0.5, 1, 0.5, −1, 0.5, −1, 1, 0.5, −1, 1) (in black dashed lines).

X m2 (t ) = AX m2 (t ) + BX tm2 (t ) + F2 (t , X m2 (t ) , X tm2 (t )). (11)

Equations (10) and (11) represent first and second drive systems respectively as well as their corresponding first
and second response systems are as follows:

X s1 (t ) = AX s1 (t ) + BX ts1 (t ) + G1 (t , X s1 (t ) , X ts1 (t )) + U1 (t ) , (12)

X s2 (t ) = AX s2 (t ) + BX ts2 (t ) + G 2 (t , X s2 (t ) , X ts2 (t )) + U 2 (t ) , (13)

where X mi (t ) = (x1mi (t ), x 2mi (t ),¼,x11

mi
(t )) Î 11 are state vectors of system (10),(11), X si (t ) =
(x1 (t ), x 2 (t ),¼,x11 (t )) Î  are state vectors of system (12), (13), Fi (t , X mi (t ), X tmi (t )), Gi (t , X si (t ), X tsi (t ))
si si si 11

are the nonlinear terms of drive system (10), (11) and response system (12), (13) respectively and U i =
i  , i = 1, 2 are nonlinear controllers to be designed later. τ is the time delay. The matrix
⎣u1 u2 ¼ u11⎤
i i
⎡ ⎦
projective combination-combination synchronization scheme for the specified hyperchaotic delayed systems, is
defined as follows:

11
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Definition 3.1 The 11-dimensional hyperchaotic systems (10) and (11) as two drive system and two response
systems as the system (12),(13) with control function are said to be matrix projective combination-combination
synchronization if there exist a controller U i = ⎡ i  and given constant matrices [P ]
⎣u1 u2 ¼ u11⎤
i i
⎦ 1 11 ´ 11 and
[P2]11 ´ 11 such that the synchronization error will be

2 2
E= å Xs i - å Pi X m i (14)
i=1 i=1

and it satisfies limt ¥∣∣E∣∣ = lim ∣∣å i = 1 X si - å i= 1 Pi X m ∣∣ = 0, where P1 and P2 are projective matrices
2 2
i
t ¥
and E = [ E (1) E (2) ¼ 
E (11) ] .

Then, the derivative of synchronization error is expressed as follows:

2
E = å (AX si (t ) + BX tsi (t ) + Gi (t , X si (t ) , X tsi (t )) + U i (t ))
i=1
2
- å Pi (AX mi (t ) + BX tmi (t ) + Fi (t , X mi (t ) , X tmi (t ))). (15)
i=1

In order to achieve matrix projective combination-combination synchronization among two drive systems (10),
(11) and two response system (12), (13), our goal is to build a suitable controller U i, i = 1, 2 for the response
system (12), (13) such that the error system (15) is asymptotically stable.
Theorem 3.1 The matrix projective combination-combination synchronization is achieved between equations (10)
− (13) if there exists U i Î 11 such that
2 2
U i = - (K1 + A) E (t ) - (K2 + B) E t - Aå Pi X mi - Bå Pi X tmi
i=1 i=1
2 2
+ å Pi X  mi - å Gi (t , X s (t ) , X ts (t )) ,
i i
i = 1, 2 (16)
i=1 i=1

provided that − K1 + I must be negative definite and K2 must be positive definite where K1 and K2 are the control
gain matrices.
Proof. The derivative of E(t) is

E (t ) = X1 + X 2 - P1 X m1 - P2 X m2.
s s

Using the value of U i from (16) in (15) gives

E (t ) = - K1 E (t ) - K2 E t . (17)

Now, the Lyapunov function is defined as

1 2 t
V=
2
E (t ) + òt-t E 2 (s) ds.
The derivative of V is

V = E (t ) E (t ) + E 2 (t ) - Et2
= - K1 E 2 (t ) - K2 E (t ) E t + E 2 (t ) - Et2
= ( - K1 + I ) E 2 (t ) - K2 E (t ) E t - Et2
< 0,

The error system (17) is asymptotically stable in accordance with the Lyapunov stability theory [48]. The proof is
now complete. ,

12
Phys. Scr. 99 (2024) 125008 J K Bharti et al

4. Numerical simulation

In this section, we will apply our newly proposed scheme to four identical 11-dimensional hyperchaotic system
(3) to verify the effectiveness of the theoretical results. Systems (18) and (19) serve as the first and second drive
systems in the study of 11-dimensional hyperchaotic systems with time delay.

⎛ x1
m1
= a (x 3m1 - x1m1) + x 6m1 - a1 (x 3mt1 - x1mt 1) - x 6mt1,
⎜ x 2m1 = a (x4m1 - x 2m1) + x 7m1 - a1 (x 4mt1 - x 2mt1) - x 7mt 1,
⎜ x m1 = cx1m1 - x1m1 x 5m1 + x 8m1 - x 3m1 - c1 x1mt 1 + x1mt 1 x 5mt1 - x 8mt1 + x 3mt1,
3
⎜ m1
x
⎜ 4 = cx 2m1 - x 2m1 x 5m1 + x 9m1 - x4m1 - c1 x 2mt1 + x 2mt1 x 5mt1 - x 9mt 1 + x 4mt1,
m
⎜ x 5 1 = - bx 5m1 + x1m1 x 3m1 + x 2m1 x4m1 + b1 x 5mt1 - x1mt 1 x 3mt1 - x 2mt1 x 4mt1,
X m1 = ⎜ x 6m1 = dx 6m1 - x1m1 x 5m1 - d1 x 6mt1 + x1mt 1 x 5mt1, (18)
⎜ m1
x = dx 7m1 - x 2m1 x 5m1 - d1 x 7mt 1 + x 2mt1 x 5mt1,
⎜ 7
⎜ x 8
m1
= - hx 3 + x10
m1 m1
+ h1 x 3mt1 - x10m1
t
,
⎜ x 9m1 = - hx4m1 + x11 m1
+ h1 x4m1 - x11 m1
t
,
⎜ x m1 = - k1 x1m1 + k2 x 3m1 + l1 x1mt 1 - l2 x 3mt1,
⎜⎜ 10m
x 1 = - k1 x 2m1 + k2 x4m1 + l1 x 2mt1 - l2 x 4mt1.
⎝ 11

⎛ x1
m2
= a (x 3m2 - x1m2) + x 6m2 - a1 (x 3mt 2 - x1mt 2) - x 6mt 2,
⎜ x 2m2 = a (x4m2 - x 2m2) + x 7m2 - a1 (x 4mt 2 - x 2mt 2) - x 7mt 2,
⎜ x m2 = cx1m2 - x1m2 x 5m2 + x 8m2 - x 3m2 - c1 x1mt 2 + x1mt 2 x 5mt 2 - x 8mt 2 + x 3mt 2,
3
⎜ m2
⎜ x4 = cx 2m2 - x 2m2 x 5m2 + x 9m2 - x4m2 - c1 x 2mt 2 + x 2mt 2 x 5mt 2 - x 9mt 2 + x 4mt 2,
m
⎜ x 5 2 = - bx 5m2 + x1m2 x 3m2 + x 2m2 x4m2 + b1 x 5mt 2 - x1mt 2 x 3mt 2 - x 2mt 2 x 4mt 2,
X m2 = ⎜ x 6m2 = dx 6m2 - x1m2 x 5m2 - d1 x 6mt 2 + x1mt 2 x 5mt 2, (19)
⎜ m2
x = dx 7m2 - x 2m2 x 5m2 - d1 x 7mt 2 + x 2mt 2 x 5mt 2,
⎜ 7
⎜ x 8
m2
= - hx 3m2 + x10m2
+ h1 x 3mt 2 - x10
m2
t
,
⎜ x 9m2 = - hx4 + x11 + h1 x4 - x11t ,
m2 m2 m2 m2

⎜ x m2 = - k1 x1m2 + k2 x 3m2 + l1 x1mt 2 - l2 x 3mt 2,

⎜⎜ 10m
x 2 = - k1 x 2m2 + k2 x4m2 + l1 x 2mt 2 - l2 x 4mt 2.
⎝ 11

The corresponding response system is as follows:

⎛ x1 1
s
= a (x 3s1 - x1s1) + x 6s1 - a1 (x 3s1t - x1st1 ) - x 6s1t + u11,
⎜ x s1 = a (x4s1 - x 2s1) + x 7s1 - a1 (x 4s1t - x 2s1t ) - x 7st1 + u 21,
2
⎜ s1
⎜ x 3 = cx1s1 - x1s1 x 5s1 + x 8s1 - x 3s1 - c1 x1st1 + x1st1 x 5s1t - x 8s1t + x 3s1t + u31,
⎜ x4s1 = cx 2s1 - x 2s1 x 5s1 + x 9s1 - x4s1 - c1 x 2s1t + x 2s1t x 5s1t - x 9st1 + x 4s1t + u41,
⎜ x s1 = - bx 5s1 + x1s1 x 3s1 + x 2s1 x4s1 + b1 x 5s1t - x1st1 x 3s1t - x 2s1t x 4s1t + u51,
⎜ 5
X s1 = ⎜ x 6s1 = dx 6s1 - x1s1 x 5s1 - d1 x 6s1t + x1st1 x 5s1t + u 61, (20)
⎜ x 7s1 = dx 7s1- x 2s1 x 5s1 - d1 x 7st1 + x 2s1t x 5s1t + u 71,
⎜ s1
x = - hx 3s1 + x10s1 + h1 x 3s1t - x10s1 t + u81,
⎜ 8
⎜ x 9s1 = - hx4s1 + x11s1 + h1 x 4s1t - x11s1t + u 91,
⎜ x s1 = - k1 x1s1 + k2 x 3s1 + l1 x1st1 - l2 x 3s1t + u10 1
,
10
⎜ s
⎜ x 1 = - k1 x 2 + k2 x4 + l1 x 2t - l2 x 4t + u11.
s1 s1 s1 s1 1
⎝ 11

13
Phys. Scr. 99 (2024) 125008 J K Bharti et al

s
⎛ x1 2 = a (x 3s2 - x1s2) + x 6s2 - a1 (x 3st2 - x1st2 ) - x 6st2 + u12,
⎜ x s2 = a (x4s2 - x 2s2) + x 7s2 - a1 (x 4s2t - x 2s2t ) - x 7st2 + u 22,
2
⎜ s2
⎜ x 3 = cx1s2 - x1s2 x 5s2 + x 8s2 - x 3s2 - c1 x1st2 + x1st2 x 5st2 - x 8st2 + x 3st2 + u32,
⎜ x4s2 = cx 2s2 - x 2s2 x 5s2 + x 9s2 - x4s2 - c1 x 2s2t + x 2s2t x 5st2 - x 9st2 + x 4s2t + u42,
⎜ x s2 = - bx 5s2 + x1s2 x 3s2 + x 2s2 x4s2 + b1 x 5st2 - x1st2 x 3st2 - x 2s2t x 4s2t + u52,
⎜ 5
X s2 = ⎜ x 6s2 = dx 6s2 - x1s2 x 5s2 - d1 x 6st2 + x1st2 x 5st2 + u 62, (21)
⎜ x 7s2 = dx 7s2- x 2s2 x 5s2 - d1 x 7st2 + x 2s2t x 5st2 + u 72,
⎜ s2
x = - hx 3s2 + x10s2 + h1 x 3st2 - x10s2t + u82,
⎜ 8
⎜ x 9s2 = - hx4s2 + x11s2 + h1 x 4s2t - x11s2t + u 92,
⎜ x s2 = - k1 x1s2 + k2 x 3s2 + l1 x1st2 - l2 x 3st2 + u102,
10
⎜ s
⎜ x 2 = - k1 x 2s2 + k2 x4s2 + l1 x 2s2t - l2 x 4s2t + u112.
⎝ 11

On comparing systems (18)−(21) to equations (10)−(13), we get

⎡ -a 0 a 0 0 1 0 0 0 0 0⎤
⎢ 0 - a 0 a 0 0 1 0 0 0 0⎥
⎢ c 0 -1 0 0 0 0 1 0 0 0⎥
⎢ 0 c 0 -1 0 0 0 0 1 0 0⎥
⎢ 0 0 0 0 -b 0 0 0 0 0 0⎥
⎢ 0⎥
A =⎢ 0 0 0 0 0 d 0 0 0 0 ⎥,
⎢ 0 0 0 0 0 0 d 0 0 0 0⎥
⎢ 0 0 -h 0 0 0 0 0 0 1 0⎥
⎢ 0 0 0 -h 0 0 0 0 0 0 1⎥
⎢- k 0 k 0 0 0 0 0 0 0 0⎥
⎢ 1 2
⎥
⎣ 0 - k1 0 k2 0 0 0 0 0 0 0⎦

⎡ a1 0 - a1 0 0 -1 0 0 0 0 0⎤
⎢ 0 a1 0 - a1 0 0 -1 0 0 0 0 ⎥
⎢- c1 0 1 0 0 0 0 -1 0 0 0 ⎥
⎢ 0 -c 0 1 0 0 0 0 -1 0 0 ⎥
1
⎢ ⎥
⎢ 0 0 0 0 b1 0 0 0 0 0 0⎥
⎢
B= 0 0 0 0 0 - d1 0 0 0 0 0 ⎥,
⎢ ⎥
⎢ 0 0 0 0 0 0 - d1 0 0 0 0 ⎥
⎢ 0 0 h1 0 0 0 0 0 0 -1 0 ⎥
⎢ 0 0 0 h1 0 0 0 0 0 0 - 1⎥
⎢ ⎥
⎢ l1 0 - l2 0 0 0 0 0 0 0 0⎥
⎢
⎣ 0 l1 0 - l2 0 0 0 0 0 0 0⎥ ⎦

⎡ 0 ⎤ ⎡ 0 ⎤
⎢ 0 ⎥ ⎢ 0 ⎥
⎢ - x1mi x 5mi + x1mt i x 5mt i ⎥ ⎢ - x1si x 5si + x1sti x 5sti ⎥
⎢ ⎥ ⎢ ⎥
⎢ - x 2 x 5 + x 2t x 5t
mi mi mi mi
⎥ ⎢ - x 2 x 5 + x 2t x 5t
si si si si
⎥
⎢ x1mi x 3mi + x 2mi x4mi - x1mt i x 3mt i - x 2mt i x 4mt i ⎥ ⎢ x1si x 3si + x 2si x4si - x1si x 3si - x 2si x 4si ⎥
t t t t
Fi = ⎢ - x1mi x 5mi + x1mt i x 5mt i
⎥ &G = ⎢
i - x1si x 5si + x1sti x 5sti
⎥.
⎢ ⎥ ⎢ ⎥
⎢ - x 2 x 5 + x 2t x 5t
mi mi mi mi ⎥ ⎢ - x 2 x 5 + x 2t x 5t
si si si si ⎥
⎢ ⎥ ⎢ ⎥
⎢ 0 ⎥ ⎢ 0 ⎥
⎢ 0 ⎥ ⎢ 0 ⎥
⎢ 0 ⎥ ⎢ 0 ⎥
⎢ 0 ⎥ ⎢ 0 ⎥
⎣ ⎦ ⎣ ⎦

14
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Here, uji (1 „i „2 and 1 „j „11) are designed controller. As stated in theorem 3.1, the synchronization
condition emphasizes the importance of the projective matrix P1 & P2 Î 11 ´ 11 in achieving synchronization
between hyperchaotic systems with time delay.
The projective matrices are chosen as:

⎡0 1 0 0 1 1 1 0 0 1 1⎤ ⎡1 0 1 0 2 1 2 0 0 0 1⎤
⎢1 0 0 0 1 1 0 0 1 1 1⎥ ⎢0 0 2 2 1 2 0 1 2 0 2⎥
⎢0 0 1 0 0 0 0 0 0 1 1⎥ ⎢0 0 0 1 0 0 2 1 1 0 0⎥
⎢0 0 0 0 0 1 0 1 1 1 0⎥ ⎢0 0 0 1 1 2 1 2 2 0 0⎥
⎢1 1 0 0 1 0 1 1 0 0 0⎥ ⎢0 0 0 0 2 0 0 2 2 2 1⎥
P1 = ⎢1 1 0 1 0 1 0 1 0 0 0⎥ and P2 = ⎢0 0 0 0 0 1 0 1 2 0 1 ⎥.
⎢1 0 0 0 1 0 0 0 1 0 1⎥ ⎢0 0 0 0 0 0 0 1 1 1 2⎥
⎢ ⎥ ⎢ ⎥
⎢0 0 0 1 1 1 0 0 1 1 0⎥ ⎢0 0 0 0 0 0 0 1 0 2 2⎥
⎢0 1 0 1 0 0 1 1 0 1 0⎥ ⎢0 0 0 0 0 0 0 0 0 0 2⎥
⎢1 1 1 1 0 0 0 1 1 1 0⎥ ⎢0 0 0 0 0 0 0 0 0 0 1⎥
⎢
⎣1 1 1 0 0 0 1 0 0 0 1⎥
⎦ ⎢
⎣0 0 0 0 0 0 0 0 0 0 2⎥⎦

By definition 3.1, the error function is

E = X s1 + X s2 - P1 X m1 - P2 X m2.

Using the value of the projective matrix P1 and P2, the error function is obtained as follows:

s1 s2 m1
(1)
⎡ x1 ⎤ ⎡ x1 ⎤ ⎡ x1 ⎤
E s s
⎤ ⎢ x 21 ⎥ ⎢ x 22 ⎥ 0 x 2m1 ⎥
⎡ 1 0 0 1 1 1 0 0 1 1⎤ ⎢
⎢ E (2) ⎥ ⎢ s1 ⎥ ⎢ s2 ⎥ ⎡ ⎢ m1 ⎥
⎢ E (3) ⎥ ⎢ x 3 ⎥ ⎢ x 3 ⎥ ⎢1 0 0 0 1 1 0 0 1 1 1 ⎥ ⎢ x3 ⎥
⎢ (4) ⎥ ⎢ x4s1 ⎥ ⎢ x4s2 ⎥ ⎢ 0 0 1 0 0 0 0 0 0 1 1 ⎥ ⎢ x m1 ⎥
⎢ E ⎥ ⎢ s1 ⎥ ⎢ s2 ⎥ ⎢ 0 0 0 0 0 1 0 1 1 1 0⎥ ⎢ 4m1 ⎥
⎢ E (5) ⎥ ⎢ x 5 ⎥ ⎢ x 5 ⎥ ⎢1 1 0 0 1 0 1 1 0 0 0⎥ ⎢ x5 ⎥
⎢ E (6) ⎥ = ⎢ x 6s1 ⎥ + ⎢ x 6s2 ⎥ - ⎢1 1 0 1 0 1 0 1 0 0 0⎥ ⎢ x 6m1 ⎥
⎢ (7) ⎥ ⎢ s1 ⎥ ⎢ s2 ⎥ ⎢1 0 0 0 1 0 0 0 1 0 1⎥ ⎢ m⎥
⎢ E ⎥ ⎢ x 7 ⎥ ⎢ x 7 ⎥ ⎢0 0 0 1 1 1 0 0 1 1
⎥ ⎢ x 7 1⎥
0⎥ ⎢ m1 ⎥
⎢ E (8) ⎥ ⎢ x 8s1 ⎥ ⎢ x 8s2 ⎥ ⎢
⎢ E (9) ⎥ ⎢ s1 ⎥ ⎢ s2 ⎥ ⎢ 0 1 0 1 0 0 1 1 0 1 0⎥ ⎢ x8 ⎥
m
⎢ (10) ⎥ ⎢ x 9 ⎥ ⎢ x 9 ⎥ ⎢1 1 1 1 0 0 0 1 1 1 0⎥ ⎢ x 9 1 ⎥
⎢ E (11) ⎥ ⎢ x10s1 ⎥ ⎢ x10s2 ⎥ ⎢ ⎣1 1 1 0 0 0 1 0 0 0 ⎥
1 ⎦ ⎢ x10m1 ⎥
⎣ E ⎦ ⎢ s1 ⎥ ⎢ s2 ⎥ ⎢ m1 ⎥
⎣ x11⎦ ⎣ x11 ⎦ ⎣ x11 ⎦
m2
⎡ x1 ⎤
1⎤ ⎢ x 2m2 ⎥
⎡1 0 1 0 2 1 2 0 0 0 ⎢ m2 ⎥
⎢0 0 2 2 1 2 0 1 2 0 2⎥ ⎢ x3 ⎥
⎢0 0 0 1 0 0 2 1 1 0 0 ⎥ ⎢ x m2 ⎥
⎢0 0 0 1 1 2 1 2 2 0 0⎥ ⎢ 4m2 ⎥
⎢0 0 0 0 2 0 0 2 2 2 1 ⎥ ⎢ x5 ⎥
- ⎢0 0 0 0 0 1 0 1 2 0 1 ⎥ ⎢ x 6m2 ⎥. (22)
⎢0 0 0 0 0 0 0 1 1 1 2⎥ ⎢ m⎥
⎢ ⎥ ⎢ x7 2⎥
⎢0 0 0 0 0 0 0 1 0 2 2 ⎥ ⎢ m2 ⎥
⎢0 0 0 0 0 0 0 0 0 0 2⎥ ⎢ x8 ⎥
m
⎢0 0 0 0 0 0 0 0 0 0 1⎥ ⎢ x 9 2⎥
⎢
⎣0 0 0 0 0 0 0 0 0 0 2⎥⎦ ⎢ x10m2 ⎥
⎢ m2 ⎥
⎣ x11 ⎦

15
Phys. Scr. 99 (2024) 125008 J K Bharti et al

E (1) = x1s1 + x1s2 - (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10

m1
+ x11
m1
)
- (x1 + x 3 + 2x 5 + x 6 + 2x 7 + x11 ),
m2 m2 m2 m2 m2 m2

E (2) = x 2s1 + x 2s2 - (x1m1 + x 5m1 + x 6m1 + x 9m1 + x10

m1
+ x11
m1
)
- (2x 3 + 2x 4 + x 5 + 2x 6 + x 8 + 2x 9 + 2x11
m2 m2 m2 m2 m2 m2 m2
),
E (3) = x 3s1 + x 3s2 - (x 3m1 + x10
m1
+ x11
m1
) - (x 4m2 + 2x 7m2 + x 8m2 + x 9m2),
E (4) = x 4s1 + x 4s2 - (x 6m1 + x 8m1 + x 9m1 + x10
m1
) - (x 4m2 + x 5m2 + 2x 6m2
+ x 7 + 2x 8 + 2x 9 ),
m2 m2 m2

E (5) = x 5s1 + x 5s2 - (x1m1 + x 2m1 + x 5m1 + x 7m1 + x 8m1) - (2x 5m2 + 2x 8m2
+ 2x 9m2 + 2x10 m2
+ x11
m2
),
E (6) = x 6s1 + x 6s2 - (x1m1 + x 2m1 + x 4m1 + x 6m1 + x 8m1) - (x 6m2 + x 8m2
+ 2x 9m2 + x11 m2
),
E (7) = x 7s1 + x 7s2 - (x1m1 + x 5m1 + x 9m1 + x11
m1
) - (x 8m2 + x 9m2 + x10
m2

+ 2x11 ),
m2

E (8) = x 8s1 + x 8s2 - (x 4m1 + x 5m1 + x 6m1 + x 9m1 + x10

m1
) - (x 8m2 + 2x10
m2

+ 2x11 ),
m2

E (9) = x 9s1 + x 9s2 - (x 2m1 + x 4m1 + x 7m1 + x 8m1 + x10

m1
) - 2x11
m2
,
E (10) = x10s1 + x10s2 - (x1m1 + x 2m1 + x 3m1 + x 4m1 + x 8m1 + x 9m1 + x10
m1
) - x11
m2
,
E (11) = x11s1 + x11s2 - (x1m1 + x 2m1 + x 3m1 + x 7m1 + x11
m1
) - 2x11
m2
. (23)

The error dynamics is derived from the error function (23) which is expressed as follows:

E (1) = a (x 3s1 - x1s1) + x 6s1 - a1 (x 3s1t - x1st1 ) - x 6s1t + u11 + a (x 3s2 - x1s2) + x 6s2
- a1 (x 3st2 - x1st2 ) - x 6st2 + u12 - (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10m1 + x11m1)
- (x1m2 + x 3m2 + 2x 5m2 + x 6m2 + 2x 7m2 + x11m2) ,
E (2) = a (x4s1 - x 2s1) + x 7s1 - a1 (x 4s1t - x 2s1t ) - x 7st1 + u 21 + a (x4s2 - x 2s2) + x 7s2
- a1 (x 4s2t - x 2s2t ) - x 7st2 + u 22 - (x1m1 + x 5m1 + x 6m1 + x 9m1 + x10m1 + x11m1)
- (2x 3m2 + 2x4m2 + x 5m2 + 2x 6m2 + x 8m2 + 2x 9m2 + 2x11m2) ,
E (3) = cx1s1 - x1s1 x 5s1 + x 8s1 - x 3s1 - c1 x1st1 + x1st1 x 5s1t - x 8s1t + x 3s1t + u31 + cx1s2
- x1s2 x 5s2 + x 8s2 - x 3s2 - c1 x1st2 + x1st2 x 5s2t - x 8s2t + x 3s2t + u32 - (x 3m1
+ x10m1 + x11m1) - (x4m2 + 2x 7m2 + x 8m2 + x 9m2) ,
E (4) = cx 2s1 - x 2s1 x 5s1 + x 9s1 - x4s1 - c1 x 2s1t + x 2s1t x 5s1t - x 9s1t + x 4s1t + u41 + cx 2s2
- x 2s2 x 5s2 + x 9s2 - x4s2 - c1 x 2s2t + x 2s2t x 5s2t - x 9st2 + x 4s2t + u42 - (x 6m1 + x 8m1
+ x 9m1 + x10m1) - (x4m2 + x 5m2 + 2x 6m2 + x 7m2 + 2x 8m2 + 2x 9m2) ,
E (5) = - bx 5s1 + x1s1 x 3s1 + x 2s1 x4s1 + b1 x 5s1t - x1st1 x 3s1t - x 2s1t x 4s1t + u51 - bx 5s2
+ x1s2 x 3s2 + x 2s2 x4s2 + b1 x 5s2t - x1st2 x 3s2t - x 2s2t x 4s2t + u52 - (x1m1 + x 2m1 + x 5m1
+ x 7m1 + x 8m1) - (2x 5m2 + 2x 8m2 + 2x 9m2 + 2x10m2 + x11m2) ,
E (6) = dx 6s1 - x1s1 x 5s1 - d1 x 6s1t + x1st1 x 5s1t + u 61 + dx 6s2 - x1s2 x 5s2 - d1 x 6s2t + x1st2 x 5s2t
+ u 62 - (x1m1 + x 2m1 + x4m1 + x 6m1 + x 8m1) - (x 6m2 + x 8m2 + 2x 9m2 + x11m2) ,
E (7) = dx 7s1 - x 2s1 x 5s1 - d1 x 7st1 + x 2s1t x 5s1t + u 71 + dx 7s2 - x 2s2 x 5s2 - d1 x 7st2 + x 2s2t x 5s2t
+ u 72 - (x1m1 + x 5m1 + x 9m1 + x11m1) - (x 8m2 + x 9m2 + x10m2 + 2x11m2) ,
E (8) = - hx 3s1 + x10s1 + h1 x 3s1t - x10s1 t + u81 - hx 3s2 + x10s2 + h1 x 3s2t - x10s2t + u82
- (x4m1 + x 5m1 + x 6m1 + x 9m1 + x10m1) - (x 8m2 + 2x10m2 + 2x11m2) ,
 (9)
E = - hx4s1 + x11s1 + h1 x 4s1t - x11s1t + u 91 - hx4s2 + x11s2 + h1 x 4s2t - x11s2t + u 92
- (x 2m1 + x4m1 + x 7m1 + x 8m1 + x10m1) - 2x11m2,
E (10) = - k1 x1s1 + k2 x 3s1 + l1 x1st1 - l2 x 3s1t + u10
1
- k1 x1s2 + k2 x 3s2 + l1 x1st2
- l2 x 3s2t + u102 - (x1m1 + x 2m1 + x 3m1 + x4m1 + x 8m1 + x 9m1 + x10m1) - x11m2,
E (11) = - k1 x 2s1 + k2 x4s1 + l1 x 2s1t - l2 x 4s1t + u11
1
- k1 x 2s2 + k2 x4s2 + l1 x 2s2t
- l2 x 4s2t + u112 - (x1m1 + x 2m1 + x 3m1 + x 7m1 + x11m1) - 2x11m2. (24)

For simplicity, let U = [u1 u2 u3 u4 u5 u6 u7 u8 u9 u10 u11] , where uk = u1j + uj2, 1  j , k  11.

16
Phys. Scr. 99 (2024) 125008 J K Bharti et al

The appropriate control gain matrices are specified as:

K1 = diag (2, 2, 7, 8, 3, 5, 9, 2, 6, 8, 6) and K2 = diag (0.04, 0.07, 0.02, 0.05, 0.08, 0.07, 0.01, 0.05, 0.07, 0.09, 0.01).
Using this control gain matrices K1 and K2, the control functions are designed according to theorem 3.1 as
follows:

u1 = - ( - a + 2) E (1) - (a1 + 0.4) Et(1) - [ - a (x 2m1 + x 5m1 + x 6m1 + x 7m1

+ x10m1
+ x11
m1
+ x1m2 + x 3m2 + 2x 5m2 + x 6m2 + 2x 7m2 + x11 m2
) + a (x 3m1
+ x10m1
+ x11
m1
+ x4m2 + 2x 7m2 + x 8m2 + x 9m2) + (x1m1 + x 2m1 + x4m1
+ x 6 + x 8 + x 6m2 + x 8m2 + 2x 9m2 + x11
m1 m1 m2
)] - [a1 (x 2mt1 + x 5mt1
+ x 6mt1 + x 7mt 1 + x10
m1
t
+ x11
m1
t
+ x1mt 2 + x 3mt 2 + 2x 5mt 2 + x 6mt 2 + 2x 7mt 2
+ x11
m2
t
) - a1 (x 3mt1 + x10
m1
t
+ x11
m1
t
+ x 4mt 2 + 2x 7mt 2 + x 8mt 2 + x 9mt 2) - (x1mt 1
+ x 2mt1 + x 4mt1 + x 6mt1 + x 8mt1 + x 6mt 2 + x 8mt 2 + 2x 9mt 2 + x11
m2
t
)] + x 2m1
+ x 5m1 + x 6m1 + x 7m1 + x10m1 + x11m1 + x1m2 + x 3m2 + 2x 5m2 + x 6m2 + 2x 7m2
+ x11m2,
u2 = - ( - a + 2) E (2) - (a1 + 0.7) Et(2) - [ - a (x1m1 + x 5m1 + x 6m1 + x 9m1
+ x10m1
+ x11
m1
+ 2x 3m2 + 2x4m2 + x 5m2 + 2x 6m2 + x 8m2 + 2x 9m2 + 2x11 m2
)
+ a (x 6 + x 8 + x 9 + x10 + x4 + x 5 + 2x 6 + x 7 + 2x 8
m1 m1 m1 m1 m2 m2 m2 m2 m2

+ 2x 9m2) + (x1m1 + x 5m1 + x 9m1 + x11m1

+ x 8m2 + x 9m2 + x10
m2
+ 2x11 m2
)]
- [a1 (x1t + x 5t + x 6t + x 9t + x10t + x11t + 2x 3t + 2x 4t + x 5mt 2
m1 m1 m1 m1 m1 m1 m2 m2

+ 2x 6mt 2 + x 8mt 2 + 2x 9mt 2 + 2x11

m2
t
) - a1 (x 6mt1 + x 8mt1 + x 9mt 1 + x10
m1
t
+ x 4mt 2
+ x 5mt 2 + 2x 6mt 2 + x 7mt 2 + 2x 8mt 2 + 2x 9mt 2) - (x1mt 1 + x 5mt1 + x 9mt 1 + x11
m1
t

+ x 8mt 2 + x 9mt 2 + x10

m2
t
+ 2x11
m2
t
)] + x1m1 + x 5m1 + x 6m1 + x 9m1 + x10m1 + x11m1
+ 2x 3m2 + 2x4m2 + x 5m2 + 2x 6m2 + x 8m2 + 2x 9m2 + 2x11m2,

u3 = - 6E (3) - 1.02Et(3) - [c (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10

m1
+ x11
m1
+ x1m2
+ x 3m2 + 2x 5m2 + x 6m2 + 2x 7m2 + x11m2
) - (x 3m1 + x10
m1
+ x11
m1
+ x4m2
+ 2x 7m2 + x 8m2 + x 9m2) + (x4m1 + x 5m1 + x 6m1 + x 9m1 + x10m1
+ x 8m2 + 2x10m2

+ 2x11 )] - [ - c1 (x 2t + x 5t + x 6t + x 7t + x10t + x11t + x1t + x 3t

m2 m1 m1 m1 m1 m1 m1 m2 m2

+ 2x 5mt 2 + x 6mt 2 + 2x 7mt 2 + x11

m2
t
) + (x 3mt1 + x10
m1
t
+ x11
m1
t
+ x 4mt 2 + 2x 7mt 2
+ x 8mt 2 + x 9mt 2) - (x 4mt1 + x 5mt1 + x 6mt1 + x 9mt 1 + x10
m1
t
+ x 8mt 2 + 2x10
m2
t
+ 2x11
m2
t
)]
+ x 3m1 + x10m1 + x11m1 + x4m2 + 2x 7m2 + x 8m2 + x 9m2 - ( - x1s1 x 5s1 + x1st1 x 5s1t
- x1s2 x 5s2 + x1st2 x 5s2t ) ,
u 4 = - 7E (4) - 1.05Et(4) - [c (x1m1 + x 5m1 + x 6m1 + x 9m1 + x10 m1
+ x11
m1
+ 2x 3m2
+ 2x4m2 + x 5m2 + 2x 6m2 + x 8m2 + 2x 9m2 + 2x11 m2
) - (x 6m1 + x 8m1 + x 9m1
+ x10 + x4 + x 5 + 2x 6 + x 7 + 2x 8 + 2x 9m2) + (x 2m1 + x4m1 + x 7m1
m1 m2 m2 m2 m2 m2

+ x 8m1 + x10
m1
+ 2x11m2
)] - [ - c1 (x1mt 1 + x 5mt1 + x 6mt1 + x 9mt 1 + x10
m1
t
+ x11
m1
t

+ 2x 3mt 2 + 2x 4mt 2 + x 5mt 2 + 2x 6mt 2 + x 8mt 2 + 2x 9mt 2 + 2x11

m2
t
) + (x 6mt1 + x 8mt1
+ x 9mt 1 + x10
m1
t
+ x 4mt 2 + x 5mt 2 + 2x 6mt 2 + x 7mt 2 + 2x 8mt 2 + 2x 9mt 2) - (x 2mt1 + x 4mt1
+ x 7mt 1 + x 8mt1 + x10
m1
t
+ 2x11
m2
t
)] + x 6m1 + x 8m1 + x 9m1 + x10m1 + x4m2 + x 5m2
+ 2x 6m2 + x 7m2 + 2x 8m2 + 2x 9m2 - ( - x 2s1 x 5s1 + x 2s1t x 5s1t - x 2s2 x 5s2 + x 2s2t x 5s2t ) ,

u5 = - ( - b + 3) E (5) - (b1 + 0.08) Et(5) - [ - b (x1m1 + x 2m1 + x 5m1 + x 7m1 + x 8m1

+ 2x 5m2 + 2x 8m2 + 2x 9m2 + 2x10
m2
+ x11
m2
)] - [b1 (x1mt 1 + x 2mt1 + x 5mt1 + x 7mt 1
+ x 8mt1 + 2x 5mt 2 + 2x 8mt 2 + 2x 9mt 2 + 2x10
m2
t
+ x11
m2
t
)] + x1m1 + x 2m1 + x 5m1 + x 7m1
+ x 8m1 + 2x 5m2 + 2x 8m2 + 2x 9m2 + 2x10m2 + x11m2 - (x1s1 x 3s1 + x 2s1 x4s1 - x1st1 x 3s1t
- x 2s1t x 4s1t + x1s2 x 3s2 + x 2s2 x4s2 - x1st2 x 3st2 - x 2s2t x 4s2t ) ,

17
Phys. Scr. 99 (2024) 125008 J K Bharti et al

u 6 = - (d + 5) E (6) - ( - d1 + 0.07) Et(6) - [d (x1m1 + x 2m1 + x4m1 + x 6m1 + x 8m1

+ x 6m2 + x 8m2 + 2x 9m2 + x11
m2
)] - [ - d1 (x1mt 1 + x 2mt1 + x 4mt1 + x 6mt1
+ x 8mt1 + x 6mt 2 + x 8mt 2 + 2x 9mt 2 + x11
m2
t
)] + x1m1 + x 2m1 + x4m1 + x 6m1 + x 8m1
+ x 6m2 + x 8m2 + 2x 9m2 + x11m2 - ( - x1s1 x 5s1 + x1st1 x 5s1t - x1s2 x 5s2 + x1st2 x 5s2t ) ,
u7 = - (d + 9) E (7) - ( - d1 + 0.01) Et(7) - [d (x1m1 + x 5m1 + x 9m1 + x11
m1
+ x 8m2
+ x 9 + x10 + 2x11 )] - [ - d1 (x1t + x 5t + x 9t + x11t + x 8t + x 9mt 2
m2 m2 m2 m1 m1 m1 m1 m2

+ x10
m2
t
+ 2x11
m2
t
)] + x1m1 + x 5m1 + x 9m1 + x11m1 + x 8m2 + x 9m2 + x10m2 + 2x11m2
- ( - x 2s1 x 5s1 + x 2s1t x 5s1t - x 2s2 x 5s2 + x 2s2t x 5s2t ) ,

u8 = - 2E (8) - 0.05Et(8) - [ - h (x 3m1 + x10

m1
+ x11m1
+ x4m2 + 2x 7m2 + x 8m2 + x 9m2)
+ (x1m1 + x 2m1 + x 3m1 + x4m1 + x 8m1 + x 9m1 + x10
m1
+ x11
m2
)] - [h1 (x 3mt1 + x10
m1
t

+ x11
m1
t
+ x 4mt 2 + 2x 7mt 2 + x 8mt 2 + x 9mt 2) - (x1mt 1 + x 2mt1 + x 3mt1 + x 4mt1 + x 8mt1
+ x 9mt 1 + x10
m1
t
+ x11
m2
t
)] + x4m1 + x 5m1 + x 6m1 + x 9m1 + x10m1 + x 8m2 + 2x10m2 + 2x11m2,
u 9 = - 6E (9) - 0.07Et(9) - [ - h (x 6m1 + x 8m1 + x 9m1 + x10
m1
+ x4m2 + x 5m2 + 2x 6m2
+ x 7 + 2x 8 + 2x 9 ) + (x1 + x 2 + x 3 + x 7 + x11
m2 m2 m2 m1 m1 m1 m1 m1
+ 2x11m2
)]
- [h1 (x 6t + x 8t + x 9t + x10t + x 4t + x 5t + 2x 6t + x 7t + 2x 8t + 2x 9mt 2)
m1 m1 m1 m1 m2 m2 m2 m2 m2

- (x1mt 1 + x 2mt1 + x 3mt1 + x 7mt 1 + x11

m1
t
+ 2x11
m2
t
)] + x 2m1 + x4m1 + x 7m1 + x 8m1
+ x10m1 + 2x11m2,

u10 = - 8E (10) - 0.09Et(10) - [ - k1 (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10 m1

+ x11
m1
+ x1m2
+ x 3m2 + 2x 5m2 + x 6m2 + 2x 7m2 + x11 m2
) + k2 (x 3m1 + x10 m1
+ x11
m1
+ x4m2 + 2x 7m2
+ x 8m2 + x 9m2)] - [l1 (x 2mt1 + x 5mt1 + x 6mt1 + x 7mt 1 + x10
m1
t
+ x11
m1
t
+ x1mt 2 + x 3mt 2
+ 2x 5mt 2 + x 6mt 2 + 2x 7mt 2 + x11
m2
t
) - l2 (x 3mt1 + x10
m1
t
+ x11
m1
t
+ x 4mt 2 + 2x 7mt 2 + x 8mt 2
+ x 9mt 2)] + x1m1 + x 2m1 + x 3m1 + x4m1 + x 8m1 + x 9m1 + x10m1 + x11m2,

u11 = - 6E (11) - 0.01Et(11) - [ - k1 (x1m1 + x 5m1 + x 6m1 + x 9m1 + x10

m1
+ x11 m1
+ 2x 3m2
+ 2x4m2 + x 5m2 + 2x 6m2 + x 8m2 + 2x 9m2 + 2x11 m2
) + k2 (x 6m1 + x 8m1 + x 9m1 + x10m1

+ x4 + x 5 + 2x 6 + x 7 + 2x 8 + 2x 9 )] - [l1 (x1t + x 5t + x 6t + x 9t
m2 m2 m2 m2 m2 m2 m1 m1 m1 m1

+ x10
m1
t
+ x11
m1
t
+ 2x 3mt 2 + 2x 4mt 2 + x 5mt 2 + 2x 6mt 2 + x 8mt 2 + 2x 9mt 2 + 2x11
m2
t
) - l2 (x 6mt1
+ x 8mt1 + x 9mt 1 + x10
m1
t
+ x 4mt 2 + x 5mt 2 + 2x 6mt 2 + x 7mt 2 + 2x 8mt 2 + 2x 9mt 2)] + x1m1 + x 2m1
+ x 3m1 + x 7m1 + x11m1 + 2x11m2. (25)

Using the values of controllers u1, u2, u3, K, u11 in the error dynamics, the error system (24) can now be
obtained as

E (1) = - 2E (1) - 0.04Et(1),

E (2) = - 2E (2) - 0.07Et(2),
E (3) = - 7E (3) - 0.02Et(3),
E (4) = - 8E (4) - 0.05Et(4),
E (5) = - 3E (5) - 0.08Et(5),
E (6) = - 5E (6) - 0.07Et(6),
E (7) = - 9E (7) - 0.01Et(7),
E (8) = - 2E (8) - 0.05Et(8),
E (9) = - 6E (9) - 0.07Et(9),
E (10) = - 8E (10) - 0.09Et(10),
E (11) = - 6E (11) - 0.01Et(11). (26)

From system (26), it is clear that all the eigenvalues of the error system (26) are negative. Therefore, the linear
system (26) is globally asymptotically stable. According to definition 3.1, the error functions E(1), E(2), E(3), K,

18
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 7. Phase portrait of first (a)−(e) and second (f)−(j) drive system (18), (19) with time delay.

E(11) converge to zero and the matrix projective combination-combination synchronization between the drive
systems (18), (19) and the response system (20), (21) are achieved.

5. Numerical results and discussion

The following section uses numerical data to demonstrate the validity of the proposed synchronization
scheme. Throughout this paper, numerical simulation of 11-dimensional hyperchaotic system are
conducted using MATLAB solver ode45 for system of ordinary differential equations while dde23 for
system of delay differential equations, as applicable. The time response, phase portraits and Lyapunov
exponents are calculated with an iteration step size of 0.001 and an iteration length of 1000. The
hardware used for the simulation is an intel i7 processor with 16 GB RAM, running MATLAB R2022b
8
on a 64 bit operating system. The parameters values of system (18)−(21) are as follows: a = 10, b = 3 ,
c = 28, d = − 2, h = 8.8, k 1 = 1, k 2 = 0.9, a 1 = 8, b 1 = 3, c 1 = 26, d 1 = 1, h1 = 9, l1 = 0.8 and l 2 = 1

19
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 8. Chaotic attractors of first (a)−(e) and second (f)−(j) response system (20), (21) with time delay.

and initial conditions of first (18) and second drive system (19) are ⎡ ⎣ x1 (0) x 2 (0) x3 (0) ¼ x11 (0)⎤
m1 m1 m1 m1
⎦
= [-1 0 1 0 -1 0 -1 1 0 -1 1] and ⎡ x m2
(0) x m2
(0) x m2
(0 ) ¼ x m2
(0) =
⎣ 1 2 3 11 ⎤
⎦
[1 1 1 -1 -1 1 -1 1 -1 1 -1] respectively. Similarly, initial conditions of response systems (20)
and (21) are ⎡
⎣ x1 (0) x 2 (0) x3 (0) ¼ x11 (0)⎤
s1 s1 s1 s1
⎦ = [-1 1 1 1 -1 1 -1 1 1 -1 1] and
⎣ x1 (0) x 2 (0) ¼ x11 (0)⎤ ⎦ = [-1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1] respectively. The time
s2 s2 s2
⎡
delay is taken as τ = 2. Equations (18) − (21) exhibits hyperchaotic behavior under the specified
parameter values and initial conditions, as shown in figures 7 and 8. Also, time series for drive systems
(18), (19) and response systems (20), (21) is depicted in figures 9–12 respectively. From the phase portrait
of 11-dimensional hyperchaotic system (3) which is without time delay shown in figure 1 and phase portrait of drive
system (18), (19) and response system (20), (21) which include time delay as shown in figures 7 and 8 respectively, it
can be clearly seen that the effect of delay is visible. When the delay component is incorporated into the hyperchaotic
system, the phase portrait and time series of the proposed systems exhibit noticeable differences. Hence, based

20
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 9. Trajectory of time series for first drive system (18).

on the description of the error function (23), the initial condition of the error system (26) is defined as follows:
⎣ E (0) E (0) E (0) ¼ E (0)⎤ ⎦ = [ 0.1 0.1 0.1 -0.1 -0.1 0.1 -0.1 0.1 -0.1 0.1 -0.1].
(1) (2) (3) (11)
⎡
Synchronization errors with time delay are represented by the variation in the error state curves. The matrix
projective combination-combination synchronization error is synchronized at t > 4 as illustrated in figure 13.
This figure demonstrates that as time increases, the error vectors asymptotically converge to zero indicating that
matrix projective combination-combination synchronization between the considered systems has been
successfully achieved. Therefore, it is evident that incorporating time delay into chaotic systems (18), (19), (20)
and (21) results in faster synchronization of error trajectories than already published literature, as listed in
table 1.

6. Application in image encryption

Cryptography [49] is the science of securing communication and data through the use of codes and ciphers. In
the context of images, encryption transforms an image into an encrypted form that conceals its content from

21
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 10. Time series of second drive system (19).

unauthorized viewers. Effective cryptography is essential for protecting sensitive data in various applications,
from personal privacy to national security . In this section, cryptography is used for the secret sharing of images
with the suggested synchronized complex hyperchaotic delayed 11-dimensional system. We propose an affine
cipher based on synchronized connections of chaotic systems, using the secret message as the key. The
description of an affine cipher and secret keys are as follows:
An affine cipher is formed by combining the shift cipher and the multiplication cipher. And the affine cipher
key is made up of two integers r = (u, v). The encryption E(I) and decryption D(I) processes of an affine cipher is
given by
E (I ) = (u * I ) + v (mod q) ,
D (I ) = u-1 * (E (I ) - v )(mod q) ,

where the greatest common divisor of u and q is gcd(u, q) = 1, q is a positive integer and u−1 is the inverse of u
modulo q.

22
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 11. Time series of first response system (20).

Remark 3. The E(I) and D(I) operation is known as a shift cipher if u = 1.

Remark 4. The E(I) and D(I) operation is called a multiplicative cipher if v = 0.

6.1. Proposed scheme for image encryption

A higher security image encryption scheme can be developed based on the matrix projective combination-
combination synchronization between the combination of 11-dimensional time delay system. The security of
the key system described in [50] is enhanced by proposing a new key system. The key in the form of a row matrix
contains the private message shared between the sender and the receiver. Let M = (M1, M2) represent the set of
private messages of sender and receiver. The M is divided into two parts: first part M1 is secret message of sender
and second part of message M is M2 which is the secret message of receiver. It can be written as follows:
M1 = ⎡ ⎣ M11 M12 M13 ¼ M111 ⎤ ⎦ and M2 = ⎡ ⎣ M21 M22 M23 ¼ M211 ⎤ ⎦. The message M is only shared by
sender and receiver. The secret key for the proposed affine cipher is created by combining the secret message
M1 and M2 of the communicator with the results of synchronized complex hyperchaotic systems. An affine will

23
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 12. Trajectory of time series for second response system (21).

be used in image encryption and decryption techniques for the first time using matrix projective combination-
combination synchronized chaotic systems.
The private message may contain alphabets, decimals, numbers, space between two words or sentences or
any punctuation marks. Alphabet (26 letters) from A to Z are assigned by the numbers 0 to 25. The blank space
(gap) between two words is assigned by 26 and the decimal (full stop or any punctuation mark) is assigned by 27.
Numeric numbers 0 to 9 are assigned by 28, 29, K, 37. The complete assignment is shown in table 2.
The complete method of generating matrices M1 and M2 from the private message M is explained through
the following steps:

Step 1: First, start writing the message as a sequence of words.

Step 2: Convert each word in the message to a numerical value. For example if message is ‘SHE IS
BEAUTIFUL.’. Then, SHE represents 18 + 7 + 4 = 29 (mod 38) = 29, IS represent 8 + 18 =

24
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 13. Matrix projective combination-combination synchronization error states between equations (18) to (21).

26 (mod 38) = 26, BEAUTIFUL represents 1 + 4 + 0 + 20 + 19 + 8 + 5 + 20 + 11 =

88 (mod 38) = 12, blank space (-) means 26 and full stop (.) means 27 as shown in table 2.
Step 3: Arrange the numerical values into a row matrix, with each element representing a block of the message.
For example the sequence [ SHE - IS - BEAUTIFUL.] can be grouped as [ 29 26 26 26 12 27 ],
where each block can represent parts of the message, including words, spaces between words and
punctuation marks such as decimals.
Step 4: To satisfy the requirements for elements in M1 and M2, blank spaces will be included and assigned its
numerical value.

Before we implement an affine cipher, we need to establish some basic assumptions.

Assume that a sender S has an image I, that they wish to transmit securely to a receiver R. The chaotic drive
systems represented by equations (18) and (19) are denoted as the system of S, while the response system
described by equations (20) and (21) is referred to as the system of R. The involvement of projective matrices
P1 and P2 leads to the matrix projective combination-combination synchronization error of systems (18), (19),
(20) and (21) approaching zero as time increases beyond t0. Since, projective matrices are
⎡0 1 0 0 1 1 1 0 0 1 1 ⎤ ⎡1 0 1 0 2 1 2 0 0 0 1 ⎤
⎢1 0 0 0 1 1 0 0 1 1 1 ⎥ ⎢0 0 2 2 1 2 0 1 2 0 2⎥
⎢ 0 0 1 0 0 0 0 0 0 1 1 ⎥ ⎢0 0 0 1 0 0 2 1 1 0 0⎥
⎢0 0 0 0 0 1 0 1 1 1 0⎥ ⎢0 0 0 1 1 2 1 2 2 0 0⎥
⎢1 1 0 0 1 0 1 1 0 0 0 ⎥ ⎢0 0 0 0 2 0 0 2 2 2 1 ⎥
P1 = ⎢1 1 0 1 0 1 0 1 0 0 0 ⎥ and P2 = ⎢ 0 0 0 0 0 1 0 1 2 0 1 ⎥ and time delay τ = 2.
⎢1 0 0 0 1 0 0 0 1 0 1 ⎥ ⎢0 0 0 0 0 0 0 1 1 1 2⎥
⎢ ⎥ ⎢ ⎥
⎢0 0 0 1 1 1 0 0 1 1 0⎥ ⎢0 0 0 0 0 0 0 1 0 2 2⎥
⎢0 1 0 1 0 0 1 1 0 1 0⎥ ⎢0 0 0 0 0 0 0 0 0 0 2⎥
⎢1 1 1 1 0 0 0 1 1 1 0 ⎥ ⎢0 0 0 0 0 0 0 0 0 0 1 ⎥
⎢
⎣ 1 1 1 0 0 0 1 0 0 0 1 ⎥
⎦ ⎢
⎣0 0 0 0 0 0 0 0 0 0 2⎥ ⎦
Therefore, the S and R can choose the values of the state variables after t0 = 4.0 to obtain the secret keys.

6.1.1. Key generation process

• The S selects an integer t1 t0 and solves the combined chaotic drive systems described in equations (18) and
(19) at time t1.
• Then, S computes the key

X1 = M1 (P1 X m1 + P2 X m2)(mod q) ,
X2 = M2 (P1 X m1 + P2 X m2)(mod q).

25
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Table 1. Comparison of various synchronization scheme and corresponding time.

Synchronization scheme Synchronization time (approx. in secs)

Dual combination synchronization [28] 7

Fuzzy synchronization using adaptive control [22] 13
Combination-combination synchronization [30] 10
Triple compound combination synchronization [31] 7
Triple compound synchronization [27] 6.5
Fixed time projective synchronization [34] 5.0141
Projective synchronization [36] 5
Matrix projective synchronization [21] 6
Hybrid projective combination difference synchronization [39] 5
Matrix projective combination-combination synchronization (Current approach) 4

Table 2. Determining numbers with unit messages.

Unit message A B C — Z — . 0 1 — 9
Number assigned 0 1 2 — 25 26 27 28 29 — 37

 m1
⎡ M11 ⎤
⎧ ⎡ x1 ⎤
⎪ m1
⎢ M12 ⎥ ⎪⎡ 0 1 0 0 1 1 1 0 0 1 1⎤ ⎢ x2 ⎥
⎢M ⎥ ⎢ m1 ⎥
⎪⎢ 1 0 0 0 1 1 0 0 1 1 1⎥ ⎢ x3 ⎥
⎢ 13 ⎥ ⎪⎢ 0 0 1 0 0 0 0 0 0 1 1 ⎥ ⎢ x m1 ⎥
⎢ M14 ⎥ ⎪⎢ 0 0 0 0 0 1 0 1 1 1 0 ⎥ ⎢ 4m1 ⎥
⎢ M1 ⎥
⎢ 5⎥
⎪⎢ 1
⎪⎢ 1 0 0 1 0 1 1 0 0 0⎥ ⎢ x5 ⎥
X1 = ⎢ M16 ⎥ 1 1 0 1 0 1 0 1 0 0 0 ⎥ ⎢ x 6m1 ⎥
⎢ M17 ⎥ ⎨⎢ 1 0 0 0 1 0 0 0 1 0
⎥ m⎥
1⎥ ⎢
⎢ ⎥ ⎪ ⎢ ⎢ x 7 1⎥
⎢ M18 ⎥ ⎪⎢ 0 0 0 1 1 1 0 0 1 1 0 ⎥ ⎢ m1 ⎥
x
⎪⎢ 0 1 0 1 0 0 1 1 0 1 0⎥ ⎢ 8 ⎥
⎢ M19 ⎥ m
⎢M ⎥ ⎪⎢ 1 1 1 1 0 0 0 1 1 1 0⎥ ⎢ x 9 1⎥
⎢ 110 ⎥ ⎪⎢⎣1 1 1 0 0 0 1 0 0 0 ⎥ ⎢
1 ⎦ x10 m1 ⎥
⎪ ⎢ m⎥
⎣ M111 ⎥
⎢ ⎦ ⎪
⎣ x11 ⎥
⎢ 1
⎦
⎩
m2
⎡ x1 ⎤ ⎫
⎢ x m2 ⎥ ⎪
⎡1 0 1 0 2 1 2 0 0 0 1 ⎤ ⎢ 2m2 ⎥ ⎪
⎢0 0 2 2 1 2 0 1 2 0 2⎥ ⎢ x 3 ⎥ ⎪
⎢ 0 0 0 1 0 0 2 1 1 0 0 ⎥ ⎢ x m2 ⎥ ⎪
⎢ 0 0 0 1 1 2 1 2 2 0 0 ⎥ ⎢ 4m ⎥ ⎪
⎢0 0 0 0 2 0 0 2 2 2 1 ⎥ ⎢ x5 2 ⎥ ⎪ ⎪
+ ⎢ 0 0 0 0 0 1 0 1 2 0 1 ⎥ ⎢ x 6m2 ⎥ (mod q)
⎢ ⎥⎢ m ⎥⎬
⎢0 0 0 0 0 0 0 1 1 1 2⎥ ⎢ x 7 2 ⎥ ⎪
⎢ 0 0 0 0 0 0 0 1 0 2 2 ⎥ ⎢ x m2 ⎥ ⎪
⎢ 0 0 0 0 0 0 0 0 0 0 2 ⎥ ⎢ 8m ⎥ ⎪
⎢0 0 0 0 0 0 0 0 0 0 1 ⎥ ⎢ x 9 2 ⎥ ⎪
⎣0 0 0 0 0 0 0 0 0 0 2⎥
⎢ m2 ⎥ ⎪
⎦ ⎢ x10
⎢ m ⎥⎪
2 ⎪
⎣ x11 ⎥
⎢ ⎦⎭
X1 = M11 * (x 2 + x 5 + x 6 + x 7 + x10 + x11 + x1m2 + x 3m2
m1 m1 m1 m1 m1 m1

+ 2x 5m2 + x 6m2 + 2x 7m2 + x11 m2

) + M12 (x1m1 + x 5m1 + x 6m1
+ x 9m1 + x10m1
+ x11 m1
+ 2x 3m2 + 2x 4m2 + x 5m2 + 2x 6m2 + x 8m2
+ 2x 9 + 2x11 ) + M13 * (x 3m1 + x10
m2 m2 m1
+ x11m1
+ x 4m2 + 2x 7m2
+ x 8 + x 9 ) + M14 * (x 6 + x 8 + x 9 + x10
m2 m2 m1 m1 m1 m1
+ x 4m2
+ x 5 + 2x 6 + x 7 + 2x 8 + 2x 9 ) + M15 * (x1m1 + x 2m1
m2 m2 m2 m2 m2

+ x 5m1 + x 7m1 + x 8m1 + 2x 5m2 + 2x 8m2 + 2x 9m2 + 2x10 m2

+ x11m2
)
+ M16 * (x1 + x 2 + x 4 + x 6 + x 8 + x 6 + x 8
m1 m1 m1 m1 m1 m2 m2

+ 2x 9m2 + x11 m2
) + M17 * (x1m1 + x 5m1 + x 9m1 + x11 m1
+ x 8m2
+ x 9m2 + x10m2
+ 2x11 m2
) + M18 * (x 4m1 + x 5m1 + x 6m1 + x 9m1
+ x10 + x 8 + 2x10 + 2x11
m1 m2 m2 m2
) + M19 * (x 2m1 + x 4m1 + x 7m1
+ x 8 + x10 + 2x11 ) + M110 * (x1m1 + x 2m1 + x 3m1 + x 4m1
m1 m1 m2

+ x 8m1 + x 9m1 + x10 m1

+ x11m2
) + M111 * (x1m1 + x 2m1 + x 3m1
+ x 7 + x11 + 2x11 )(mod q).
m1 m1 m2
(27)

26
Phys. Scr. 99 (2024) 125008 J K Bharti et al

m1
⎧ ⎡ x1 ⎤
⎡ M21 ⎤ ⎪
⎢ M22 ⎥ ⎪⎡ 0 1 0 0 1 1 1 0 0 1 1⎤ ⎢ x 2m1 ⎥
⎢ m1 ⎥
⎢ M 2 ⎥ ⎪⎢ 1 0 0 0 1 1 0 0 1 1 1 ⎥ ⎢ x3 ⎥
⎢ 3 ⎥ ⎪⎢ 0 0 1 0 0 0 0 0 0 1 1 ⎥ ⎢ x m1 ⎥
⎢ M 2 4 ⎥ ⎪⎢ 0 0 0 0 0 1 0 1 1 1 0⎥ ⎢ 4m1 ⎥
⎢ M 2 5 ⎥ ⎪⎢ 1 1 0 0 1 0 1 1 0 0 0⎥ ⎢ x5 ⎥
⎢ ⎥ ⎪
X2 = ⎢ M2 6 ⎥ ⎢ 1 1 0 1 0 1 0 1 0 0 0⎥ ⎢ x 6m1 ⎥
⎨⎢ 1⎥ ⎢ m⎥
⎢ M27 ⎥ ⎪⎢1 0 0 0 1 0 0 0 1 0 ⎥ ⎢ x 7 1⎥
⎢ M 2 ⎥ ⎪⎢ 0 0 0 1 1 1 0 0 1 1 0⎥ ⎢ m1 ⎥
⎢ 8 ⎥ ⎪⎢ 0 1 0 1 0 0 1 1 0 1 0⎥ ⎢ x8 ⎥
⎢ M 2 9 ⎥ ⎪⎢ 1 1 1 1 0 0 0 1 1 1
m
0⎥ ⎢ x 9 1 ⎥
⎢ M210 ⎥ ⎪⎢1 1⎥
⎢ M ⎥ ⎪⎣
1 1 0 0 0 1 0 0 0 ⎦ ⎢ x10m1 ⎥
⎢ m1 ⎥
⎣ 211 ⎦ ⎪
⎩ ⎣ x11 ⎦
m2
⎡ x1 ⎤ ⎫
1⎤ ⎢ x 2m2 ⎥ ⎪
⎡1 0 1 0 2 1 2 0 0 0 ⎢ m2 ⎥ ⎪
⎢0 0 2 2 1 2 0 1 2 0 2⎥ ⎢ x3 ⎥ ⎪
⎢0 0 0 1 0 0 2 1 1 0 0 ⎥ ⎢ x m2 ⎥ ⎪
⎢0 0 0 1 1 2 1 2 2 0 0⎥ ⎢ 4m2 ⎥ ⎪
⎢0 0 0 0 2 0 0 2 2 2 1 ⎥ ⎢ x5 ⎥ ⎪ ⎪
+⎢0 0 0 0 0 1 0 1 2 0 1 ⎥ ⎢ x 6m2 ⎥ (mod q)
⎢0 0 0 0 0 0 0 1 1 1 2⎥ ⎢ m ⎥⎬
⎢ ⎥ ⎢ x7 2⎥ ⎪
⎢0 0 0 0 0 0 0 1 0 2 2 ⎥ ⎢ m2 ⎥ ⎪
⎢0 0 0 0 0 0 0 0 0 0 2⎥ ⎢ x8 ⎥ ⎪
m
⎢0 0 0 0 0 0 0 0 0 0 1⎥ ⎢ x 9 2⎥ ⎪
⎢
⎣0 0 0 0 0 0 0 0 0 0 2⎥⎦ ⎢ x10m2 ⎥ ⎪
⎢ m2 ⎥ ⎪
⎪
⎣ x11 ⎦ ⎭

X2 = M21 * (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10 m1

+ x11
m1
+ x1m2
+ x 3 + 2x 5 + x 6 + 2x 7 + x11 ) + M22 * (x1m1 + x 5m1
m2 m2 m2 m2 m2

+ x 6m1 + x 9m1 + x10m1

+ x11m1
+ 2x 3m2 + 2x4m2 + x 5m2 + 2x 6m2
+ x 8 + 2x 9 + 2x11 ) + M23 * (x 3m1 + x10
m2 m2 m2 m1
+ x11
m1
+ x4m2
+ 2x 7m2 + x 8m2 + x 9m2) + M24 * (x 6m1 + x 8m1 + x 9m1 + x10 m1

+ x4m2 + x 5m2 + 2x 6m2 + x 7m2 + 2x 8m2 + 2x 9m2) + M25 * (x1m1

+ x 2m1 + x 5m1 + x 7m1 + x 8m1 + 2x 5m2 + 2x 8m2 + 2x 9m2 + 2x10 m2

+ x11m2
) + M26 * (x1m1 + x 2m1 + x4m1 + x 6m1 + x 8m1 + x 6m2
+ x 8 + 2x 9m2 + x11
m2 m2
) + M27 * (x1m1 + x 5m1 + x 9m1 + x11 m1

+ x 8m2 + x 9m2 + x10m2

+ 2x11 m2
) + M28 * (x4m1 + x 5m1 + x 6m1
+ x 9m1 + x10m1
+ x 8m2 + 2x10m2
+ 2x11m2
) + M29 * (x 2m1 + x4m1
+ x 7 + x 8 + x10 + 2x11 ) + M210 * (x1m1 + x 2m1 + x 3m1
m1 m1 m1 m2

+ x4m1 + x 8m1 + x 9m1 + x10 m1

+ x11m2
) + M211 * (x1m1 + x 2m1
+ x 3 + x 7 + x11 + 2x11 )(mod q).
m1 m1 m1 m2
(28)

• The S kept their key pair (Ξ1, Ξ2) confidential.

• At time t1, R solves the chaotic response systems (20) and (21).
• Then, R calculates the key
u1 = M1 (X s1 + X s2)(mod q) , u2 = M2 (X s1 + X s2)(mod q).

u1 = M11 * (x1s1 + x1s2) + M12 * (x 2s1 + x 2s2) + M13 * (x 3s1 + x 3s2)

+ M14 * (x4s1 + x4s2) + M15 * (x 5s1 + x 5s2) + M16 * (x 6s1 + x 6s2)
+ M17 * (x 7s1 + x 7s2) + M18 * (x 8s1 + x 8s2) + M19 * (x 9s1 + x 9s2)
+ M110 * (x10s1 + x10s2) + M111 * (x11s1 + x11s2)(mod q) , (29)

27
Phys. Scr. 99 (2024) 125008 J K Bharti et al

u2 = M21 * (x1s1 + x1s2) + M22 * (x 2s1 + x 2s2) + M23 * (x 3s1 + x 3s2)

+ M24 * (x4s1 + x4s2) + M25 * (x 5s1 + x 5s2) + M26 * (x 6s1 + x 6s2)
+ M27 * (x 7s1 + x 7s2) + M28 * (x 8s1 + x 8s2) + M29 * (x 9s1 + x 9s2)
+ M210 * (x10s1 + x10s2) + M211 * (x11s1 + x11s2)(mod q). (30)

• The R kept their key pair (υ1, υ2) secret.

6.1.2. Procedure for encryption

• The S intends to securely transmit a digital image I of dimension K × L over an insecure channel.
• The process by S generates an encrypted version of image I through the formula

E (I ) = (I * X1) + X2 (mod q). (31)

• Further, S sends the encrypted image E(I) to R.

6.1.3. Procedure for decryption

• The receiver R receives the encrypted image, E(I) from the sender S.
• For decryption, R calculates

D (I ) = (E (I ) - u2) * u1-1 (mod q). (32)

• Eventually, R successfully recovers the original image I.

For, D (I ) = (E (I ) - u2) * u1-1 (mod q)

= ((I * X1) + X2 (mod q) - u2) * u1-1 (mod q)
= (I * X1 + X2 - u2) * u1-1 (mod q)
= (I * X1 + M21 * (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10 m1
+ x11 m1

+ x1 + x 3 + 2x 5 + x 6 + 2x 7 + x11 ) + M22 * (x1m1

m2 m2 m2 m2 m2 m2

+ x 5m1 + x 6m1 + x 9m1 + x10 m1

+ x11 m1
+ 2x 3m2 + 2x4m2 + x 5m2
+ 2x 6m2 + x 8m2 + 2x 9m2 + 2x11 m2
) + M23 * (x 3m1 + x10 m1
+ x11 m1

+ x4 + 2x 7 + x 8 + x 9 ) + M24 * (x 6 + x 8 + x 9
m2 m2 m2 m2 m1 m1 m1

+ x10m1
+ x4m2 + x 5m2 + 2x 6m2 + x 7m2 + 2x 8m2 + 2x 9m2)
+ M25 * (x1m1 + x 2m1 + x 5m1 + x 7m1 + x 8m1 + 2x 5m2 + 2x 8m2
+ 2x 9m2 + 2x10 m2
+ x11 m2
) + M26 * (x1m1 + x 2m1 + x4m1 + x 6m1
+ x 8m1 + x 6m2 + x 8m2 + 2x 9m2 + x11 m2
) + M27 * (x1m1 + x 5m1
+ x 9 + x11 + x 8 + x 9 + x10 + 2x11
m1 m1 m2 m2 m2 m2
) + M28 * (x4m1
+ x 5m1 + x 6m1 + x 9m1 + x10 m1
+ x 8m2 + 2x10 m2
+ 2x11
m2
) + M 2 9*
(x 2 + x4 + x 7 + x 8 + x10 + 2x11 ) + M210 * (x1m1 + x 2m1
m1 m1 m1 m1 m1 m2

+ x 3m1 + x4m1 + x 8m1 + x 9m1 + x10 m1

+ x11m2
) + M211*(x1m1 + x 2m1
+ x 3m1 + x 7m1 + x11 m1
+ 2x11 m2
) - (M21 * (x1s1 + x1s2) + M22 * (x 2s1
+ x 2 ) + M23 * (x 3 + x 3 ) + M24*(x4s1 + x4s2) + M25 * (x 5s1
s2 s1 s2

+ x 5s2) + M26 * (x 6s1 + x 6s2) + M27 * (x 7s1 + x 7s2) + M28 * (x 8s1

+ x 8s2) + M29 * (x 9s1 + x 9s2) + M210 * (x10s1 + x10s2) + M211 * (x11s1
+ x11s2 )) * u1-1 (mod q)

28
Phys. Scr. 99 (2024) 125008 J K Bharti et al

= (I * X1 + M21 ((x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10 m1

+ x11m1
+ x1m2
+ x 3 + 2x 5 + x 6 + 2x 7 + x11 ) - (x1 + x1 )) + M22 ((x1m1
m2 m2 m2 m2 m2 s1 s2

+ x 5m1 + x 6m1 + x 9m1 + x10 m1

+ x11m1
+ 2x 3m2 + 2x 4m2 + x 5m2 + 2x 6m2
+ x 8m2 + 2x 9m2 + 2x11 m2
) - (x 2s1 + x 2s2)) + M23 ((x 3m1 + x10 m1
+ x11 m1

+ x 4 + 2x 7 + x 8 + x 9 ) - (x 3 + x 3 )) + M2 4 ((x 6 + x 8
m2 m2 m2 m2 s1 s2 m1 m1

+ x 9m1 + x10m1
+ x 4m2 + x 5m2 + 2x 6m2 + x 7m2 + 2x 8m2 + 2x 9m2) - (x 4s1
+ x 4s2)) + M2 5 ((x1m1 + x 2m1 + x 5m1 + x 7m1 + x 8m1 + 2x 5m2 + 2x 8m2
+ 2x 9m2 + 2x10 m2
+ x11 m2
) - (x 5s1 + x 5s2)) + M2 6 ((x1m1 + x 2m1 + x 4m1
+ x 6 + x 8 + x 6 + x 8m2 + 2x 9m2 + x11
m1 m1 m2 m2
) - (x 6s1 + x 6s2))
+ M27 ((x1 + x 5 + x 9 + x11 + x 8 + x 9m2 + x10
m1 m1 m1 m1 m2 m2
+ 2x11m2
)
- (x 7 + x 7 )) + M2 8 ((x 4 + x 5 + x 6 + x 9 + x10 + x 8 +
s1 s2 m1 m1 m1 m1 m1 m2

m2
2x10 + 2x11m2
) - (x 8s1 + x 8s2)) + M2 9 ((x 2m1 + x 4m1 + x 7m1 + x 8m1
+ x10 + 2x11 ) - (x 9s1 + x 9s2)) + M210 ((x1m1 + x 2m1 + x 3m1 + x 4m1
m1 m2

+ x 8m1 + x 9m1 + x10 m1

+ x11m2
) - (x10s1 + x10s2)) + M211 ((x1m1 + x 2m1
+ x 3m1 + x 7m1 + x11
m1
+ 2x11
m2
) - (x11s1 + x11s2))) * u1-1 (mod q)
= (I * X1) * u1-1 (mod q)
= I * (M11 * (x 2m1 + x 5m1 + x 6m1 + x 7m1 + x10 m1
+ x11m1
+ x1m2 + x 3m2
+ 2x 5m2 + x 6m2 + 2x 7m2 + x11 m2
) + M12 * (x1m1 + x 5m1 + x 6m1 + x 9m1
+ x10 + x11 + 2x 3 + 2x 4 + x 5m2 + 2x 6m2 + x 8m2 + 2x 9m2 + 2x11
m1 m1 m2 m2 m2
)
+ M13 * (x 3 + x10 + x11 + x 4 + 2x 7 + x 8 + x 9 ) + M14*
m1 m1 m1 m2 m2 m2 m2

(x 6m1 + x 8m1 + x 9m1 + x10 m1

+ x 4m2 + x 5m2 + 2x 6m2 + x 7m2 + 2x 8m2
+ 2x 9m2) + M15 * (x1m1 + x 2m1 + x 5m1 + x 7m1 + x 8m1 + 2x 5m2 + 2x 8m2
+ 2x 9m2 + 2x10 m2
+ x11m2
) + M16 * (x1m1 + x 2m1 + x 4m1 + x 6m1 + x 8m1
+ x 6 + x 8 + 2x 9 + x11
m2 m2 m2 m2
) + M17 * (x1m1 + x 5m1 + x 9m1 + x11 m1

+ x 8m2 + x 9m2 + x10 m2

+ 2x11 m2
) + M18 * (x 4m1 + x 5m1 + x 6m1 + x 9m1
+ x10 + x 8 + 2x10 + 2x11 ) + M19 * (x 2m1 + x 4m1 + x 7m1 + x 8m1
m1 m2 m2 m2

+ x10 m1
+ 2x11 m2
) + M110 * (x1m1 + x 2m1 + x 3m1 + x 4m1 + x 8m1 + x 9m1
+ x10 m1
+ x11 m2
) + M111 * (x1m1 + x 2m1 + x 3m1 + x 7m1 + x11 m1
+ 2x11m2
))*
(M11 * (x1 + x1 ) + M12 * (x 2 + x 2 ) + M13 * (x 3 + x 3 )
s1 s2 s1 s2 s1 s2

+ M14 * (x 4s1 + x 4s2) + M15 * (x 5s1 + x 5s2) + M16 * (x 6s1 + x 6s2)

+ M17 * (x 7s1 + x 7s2) + M18 * (x 8s1 + x 8s2) + M19 * (x 9s1 + x 9s2)
+ M110 * (x10s1 + x10s2) + M111 * (x11s1 + x11s2))-1 (mod q)
D (I ) = I (mod q).
[Q equations (18)−(21) are synchronized, for every t1 t0 under suitable projective matrices P1 and P2].

6.2. Practical demonstration of proposed image cryptosystem

This subsection will provide an experimental demonstration of the performance of the proposed algorithm.
Assume that t1 = 10.5 and set q = 256. At t1 = 10.5, x1m1 (t1) = 18.778, x 2m1 (t1) = 1.637, x3m1 (t1) = 28.010,
x4 (t1) = 1.837, x5m1 (t1) = 35.424, x6m1 (t1) = -24.150, x 7m1 (t1) = -2.243, x8m1 (t1) = -33.713,
m1

x 9m1 (t1) = -2.807, x10 m1

(t1) = 0.454, x11
m1
(t1) = 1.110, x1m2 (t1) = 8.779, x 2m2 (t1) = -15.661, x3m2 (t1) = 10.370,
x4 (t1) = -23.753, x5 (t1) = 35.82, x6 (t1) = -14.191, x 7m2 (t1) = 22.673, x8m2 (t1) = -14.955,
m2 m2 m2

x 9m2 (t1) = 29.655, x10m2

(t1) = 1.526, x11
m2
(t1) = -2.267, x1s1 (t1) = 12.232, x 2s1 (t1) = 28.809, x3s1 (t1) = 29.574,
x4 (t1) = -60.216, x5 (t1) = 19.883, x6 (t1) = -35.611, x 7s1 (t1) = 52.505, x8s1 (t1) = 10.758,
s1 s1 s1

x 9s1 (t1) = -32.028, x10s1 (t1) = 14.196, x11s1 (t1) = 47.292, x1s2 (t1) = 119.677, x 2s2 (t1) = 20.493, x3s2 (t1) = 36.293,
x4s2 (t1) = 35.758, x5s2 (t1) = 101.825, x6s2 (t1) = 27.897, x 7s2 (t1) = 11.692, x8s2 (t1) = -16.437, x 9s2 (t1) = -4.534,
x10s2 (t1) = -2.267 and x11s2 (t1) = -4.534.
Four cases arise depending on the number of words, spaces between words and punctuation marks in the
message M. Let n denote the total number of words, spaces and punctuation marks in the message M.

Case 1: If n „ 11.
Case 2: If 11 < n < 22.
Case 3: If n = 22.
Case 4: If n > 22.

To validate the cases, examples are provided for each as follows:

29
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Example 1. This situation corresponds to case 1. Suppose S and R have the message M as ‘Hello, how are you
today?’. This message contains 11 elements (including spaces between words and punctuation mark). Since, this
length is not a multiple of 22, we can add blank spaces to complete the sequence and assign the numerical value
of each element of the message M. The message M is then divided into two parts, M1 and M2. M1 is the private
message for S and M2 is the private message for R, as shown in table 3. Using M1 andM2, S and R calculate their
respective keys (Ξ1, Ξ2) and (υ1, υ2) as follows:

X1 = 9 * (1.637 + 35.424 + ( - 24.150) + ( - 2.243) + 0.454 + 1.110 + 8.779

+ 10.370 + 2(35.82) + ( - 14.191) + 2(22.673) + ( - 2.267)) + 27 * (18.778
+ 35.424 + ( - 24.150) + ( - 2.807) + 0.454 + 1.110 + 2(10.370)
+ 2( - 23.753) + 35.82 + 2( - 14.191) + ( - 14.955) + 2(29.655) + 2( - 2.267))
+ 26 * (28.010 + 0.454 + 1.110 + ( - 23.753) + 2(22.673) + ( - 14.955)
+ 29.655) + 5 * (( - 24.150) + ( - 33.713) + ( - 2.807) + 0.454 + ( - 23.753)
+ 35.82 + 2( - 14.191) + 22.673 + 2( - 14.955) + 2(29.655)) + 26 * (18.778
+ 1.637 + 35.424 + ( - 2.243) + ( - 33.713) + 2(35.82) + 2( - 14.955)
+ 2(29.655) + 2(1.526) + ( - 2.267)) + 21 * (18.778 + 1.637 + 1.837
+ ( - 24.150) + ( - 33.713) + ( - 14.191) + ( - 14.955) + 2(29.655) + ( - 2.267))
+ 26 * (18.778 + 35.424 + ( - 2.807) + 1.110 + ( - 14.955) + 29.655 + 1.526
+ 2( - 2.267)) + 20 * (1.837 + 35.424 + ( - 24.150) + ( - 2.807) + 0.454
+ ( - 14.955) + 2(1.526) + 2( - 2.267)) + 26 * (1.637 + 1.837 + ( - 2.243)
+ ( - 33.713) + 0.454 + 2( - 2.267)) + 22 * (18.778 + 1.637 + 28.010 + 1.837
+ ( - 33.713) + ( - 2.807) + 0.454 + ( - 2.267)) + 27 * (18.778 + 1.637
+ 28.010 + ( - 2.243) + 1.110 + 2( - 2.267))(mod 256) ,
X1 = 168.051(mod 256).
X2 = 26 * (1.637 + 35.424 + ( - 24.150) + ( - 2.243) + 0.454 + 1.110 + 8.779
+ 10.370 + 2(35.82) + ( - 14.191) + 2(22.673) + ( - 2.267)) + 26 * (18.778
+ 35.424 + ( - 24.150) + ( - 2.807) + 0.454 + 1.110 + 2(10.370)
+ 2( - 23.753) + 35.82 + 2( - 14.191) + ( - 14.955) + 2(29.655) + 2( - 2.267))
+ 26 * (28.010 + 0.454 + 1.110 + ( - 23.753) + 2(22.673) + ( - 14.955)
+ 29.655) + 26 * (( - 24.150) + ( - 33.713) + ( - 2.807) + 0.454 + ( - 23.753)
+ 35.82 + 2( - 14.191) + 22.673 + 2( - 14.955) + 2(29.655)) + 26 * (18.778
+ 1.637 + 35.424 + ( - 2.243) + ( - 33.713) + 2(35.82) + 2( - 14.955)
+ 2(29.655) + 2(1.526) + ( - 2.267)) + 26 * (18.778 + 1.637 + 1.837
+ ( - 24.150) + ( - 33.713) + ( - 14.191) + ( - 14.955) + 2(29.655) + ( - 2.267))
+ 26 * (18.778 + 35.424 + ( - 2.807) + 1.110 + ( - 14.955) + 29.655 + 1.526
+ 2( - 2.267)) + 26 * (1.837 + 35.424 + ( - 24.150) + ( - 2.807) + 0.454
+ ( - 14.955) + 2(1.526) + 2( - 2.267)) + 26 * (1.637 + 1.837 + ( - 2.243)
+ ( - 33.713) + 0.454 + 2( - 2.267)) + 26 * (18.778 + 1.637 + 28.010
+ 1.837 + ( - 33.713) + ( - 2.807) + 0.454 + ( - 2.267)) + 26 * (18.778
+ 1.637 + 28.010 + ( - 2.243) + 1.110 + 2( - 2.267))(mod 256) ,
X2 = 243.898(mod 256).
u1 = 9*(12.232 + 119.677) + 27*(28.809 + 20.493) + 26 * (29.574
+ 36.293) + 5*( - 60.216 + 35.758) + 26*(19.883 + 101.825)
+ 21*( - 35.611 + 27.897) + 26*(52.505 + 11.692) + 20 * (10.758
+ ( - 16.437)) + 26*( - 32.028 + ( - 4.534)) + 22*(14.196 + ( - 2.267))
+ 27*(47.292 + ( - 4.534))(mod 256) ,
u1 = 168.051(mod 256).
u2 = 26*(12.232 + 119.677) + 26*(28.809 + 20.493) + 26 * (29.574
+ 36.293) + 26*( - 60.216 + 35.758) + 26*(19.883 + 101.825)
+ 26*( - 35.611 + 27.897) + 26*(52.505 + 11.692) + 26 * (10.758
+ ( - 16.437) + 26*( - 32.028 + ( - 4.534)) + 26*(14.196 + ( - 2.267))
+ 26*(47.292 + ( - 4.534))(mod 256) ,
u2 = 243.898(mod 256).

30
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Table 3. Description of the private message of S and R in Example 1.

M1 Private message (S) Value M2 Private message (R) Value

M11 Hello 9 M21 — 26

M12 , 27 M22 — 26
M13 — 26 M23 — 26
M14 how 5 M24 — 26
M15 — 26 M25 — 26
M16 are 21 M 26 — 26
M17 — 26 M27 — 26
M18 you 20 M28 — 26
M19 — 26 M29 — 26
M110 today 22 M210 — 26
M111 ? 27 M211 — 26

Table 4. Details related to the message used in Example 2.

M1 Private message (S) Value M2 Private message (R) Value

M11 The 30 M21 — 26

M12 — 26 M22 13 22
M13 quick 18 M23 — 26
M14 — 26 M24 lazy 22
M15 brown 29 M25 — 26
M16 — 26 M 26 dogs 3
M17 fox 4 M27 · 27
M18 — 26 M28 — 26
M19 jumps 36 M29 — 26
M110 — 26 M210 — 26
M111 over 18 M211 — 26

Example 2. In this example, we are dealing with case 2. Here, S and R consider a message M stating ‘The quick
brown fox jumps over 13 lazy dogs.’, where n = 18. Since 18 is not a multiple of 22, we add additional blank
spaces to complete the sequence. Each entities in the message M is then assigned a numerical value as shown in
table 4. In a similar way to Example 1, S and R compute their keys using equations (27)−(30) as follows:

X1 = 29.752(mod 256) , u1 = 29.752(mod 256) ,

X2 = 129.957(mod 256) , u2 = 129.957(mod 256).

Example 3. This example relates to case 3. Consider the message M shared between S and R is ‘Education is one
thing no one can take away from you.’. This message contains exactly 22 elements, which satisfies the
requirement of M1 and M2. Each element of the message M is assigned a numerical value as shown in table 5.
Using these values, S and R calculates their respective keys.

X1 = 123.252(mod 256) , u1 = 123.252(mod 256) ,

X2 = 243.27(mod 256) , u2 = 243.27(mod 256).

Example 4. In the example, a message M corresponding to case 4 by S and R says ‘Please buy the following: 3
apples, 2.5 lbs of bananas, 1 gallon of milk, and a dozen eggs.’ This message contains 40 units (words including
spaces between words and punctuation mark), which is not align with a multiple of 22. Therefore, to align with
the required format, additional blank spaces may be added, as shown in table 6. Since, the message contains
more than 22 units, the determination of values for M1 and M2 will be discussed further.

31
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Table 5. Information about the keys through Example 3.

M1 Private message (S) Value M1 Private message (R) Value

M11 Education 7 M21 — 26

M12 — 26 M22 can 15
M13 is 26 M23 — 26
M14 — 26 M24 take 33
M15 one 31 M25 — 26
M16 — 26 M 26 away 8
M17 thing 15 M27 — 26
M18 — 26 M28 from 10
M19 no 27 M29 — 26
M110 — 26 M210 you 20
M111 one 31 M211 · 27

Table 6. Key components of Example 4.

n (units) Private message (S) Value n (units) Private message (R) Value

1 Please 14 12 apples 25
2 — 26 13 , 27
3 buy 7 14 — 26
4 — 26 15 2.5 14
5 the 30 16 — 26
6 — 26 17 lbs 30
7 following 28 18 — 26
8 : 27 19 of 19
9 — 26 20 — 26
10 3 31 21 bananas 7
11 — 26 22 , 27
23 — 26 34 — 26
24 1 29 35 a 0
25 — 26 36 — 26
26 gallon 17 37 dozen 21
27 — 26 38 — 26
28 of 19 39 eggs 34
29 — 26 40 · 27
30 milk 3 41 — 26
31 , 27 42 — 26
32 — 26 43 — 26
33 and 16 44 — 26

When the message exceeds 22 units, the segments are defined as follows: units 1 to 11 constitute M1, units 12
to 22 constitute M2. For units 23 to 33, they are also classified under M1, while units 34 to 44 fall under M2. The
calculation and detailed description of the numerical values of M1 and M2 are given in table 7.
Using table 7, S and R compute their respective keys based on Example 4 as follows:
X1 = 105.11(mod 256) , u1 = 105.11(mod 256) ,
X2 = 171.028(mod 256) , u2 = 171.028(mod 256).
To determine the secret key, S and R solve four examples based on four different cases.
S uses the secret key (Ξ1, Ξ2) to encrypt the image E(I) and transfer it to R. For encryption, S calculates the
following:
E (I ) = (I * X1) + X2 (mod 256).
R then uses the secret key (υ1, υ2) to decrypt E(I). Finally, R calculates a method to restore the original image I.
D (I ) = (E (I ) - u2) u1-1 (mod 256) = I .
Several images have been used to validate the above Examples 1−4. For each example, a different image has
been selected from USC-SIPI image database [51]. The encrypted and decrypted versions of images such as
House, Jelly beans, Mandrill and Clock are displayed in figure 14. These images are part of an experiment to
assess the effectiveness of an encryption algorithm. By comparing the original images with their encrypted and

32
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Table 7. Complete message of S and R of Example 4 with the help of table 6.

M1 Private message (S) Value M2 Private message (R) Value

M11 (Please , -) 14 + 26 = 2 M21 (apples, -) 25 + 26=13

M12 (- , 1) 26 + 29 = 17 M22 (, , a) 27 + 0 = 27
M13 (buy , -) 7 + 26 = 33 M23 (-, -) 26 + 26 = 14
M14 (- , gallon) 26 + 17 = 5 M24 (2.5, dozen) 14 + 21 = 35
M15 (the , -) 30 + 26 = 18 M25 (-, -) 26 + 26 = 14
M16 (- , of) 26 + 19 = 7 M 26 (lbs, eggs) 30 + 34 = 26
M17 (following , -) 28 + 26 = 16 M27 (-, · ) 26 + 27 = 15
M18 (: , milk) 27 + 3 = 30 M28 (of , -) 19 + 26 = 7
M19 (- , , ) 26 + 27 = 15 M29 (-, -) 26 + 26 = 14
M110 (3 , -) 31 + 26 = 19 M210 (bananas, -) 7 + 26 = 33
M111 (- , and) 26 + 16 = 4 M211 (, , -) 27 + 26 = 15

decrypted counterparts, we can evaluate the encryption quality and check for any information loss. This
experiment visually demonstrates the image encryption and decryption process, allowing us to assess the
security and efficiency of the encryption algorithms discussed below.

6.3. Security analysis of the proposed encryption and decryption scheme

6.3.1. Key space analysis
A large key space is important for increasing the security of the encryption algorithm. Key space refers to the total
number of possible secret keys used in the encryption process. For a secure cryptosystem, it is necessary to have a
wide key space to defend against brute force attacks. The secret key in this image encryption scheme consists of
parameters (a, b, c, d, h, k1, k2, a1, b1, c1, d1, h1, l1, l2), initial conditions x jmi (0), x jsi (0), i = 1, 2 & j = 1, 2, K, 11,
time delay, projective matrices P1 and P2, time parameter (t, t1) and solutions of synchronized hyperchaotic
systems. With a precision of 10−14, the estimated key space 101176 ¬ 23905.

6.3.2. Histogram analysis

Histogram shows the pixel intensity distribution for an image at different levels. In figure 15, the histogram of
color components of the original and encrypted images for House, Jelly beans, Mandrill and Clock are shown.
The encrypted images display significantly different histograms than their original counterparts, exhibiting a
uniform distribution. This uniformity ensures that no obvious statistical patterns or information of the original
image can be extracted from the encrypted image, thereby increasing the security of the encryption process. As a
result, encrypted images resist statistical attacks, further strengthening their robustness in secure image
transmission.

6.3.3. Information entropy analysis

Predictability in the data sequence indicates a failure in the encryption process. To assess the unpredictability of
encrypted results, this study introduces an information entropy test. Information entropy serves as a statistical
tool to measure the randomness within a data sequence, providing insight into the uncertainty of the image
content. A higher level of uncertainty indicates that the image contains more information. Ideally, the entropy
value for a well encrypted image should be close to 8.
The formula to calculate entropy is
L-1
H = - å P (i ) log 2 P (i ) ,
i=0

where L represents the number of gray levels in the image and P(i) is the probability of each gray value i.
The information entropy for a series of encrypted images has been calculated and listed in table 8. The result
show that the entropy of encrypted images is significantly higher, indicating that they contain a larger amount of
information and more unpredictability. This confirms that the encryption algorithm performs well, ensuring a
high level of security and randomness.

6.3.4. Correlation analysis

To assess the strength of the linear relationship between adjacent pixels, we calculated the correlation coefficient
for adjacent pixel pairs in both the original and encrypted images. The correlation coefficient is calculated using
the formula:

33
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 14. Move left to right, beginning at the top: The original, encrypted and decrypted images of House, Jelly beans, Mandrill and
Clock respectively.

å i= 0 (X i - X¯ )(Yi - Y¯ ) ,
n

r=
å i= (X i - X¯ )2å i= 0 (Yi - Y¯ )2
n n

where Xi and Yi represent the intensity values of adjacent pixels, X̄ and Ȳ are the mean intensity values of these
pixels and n is the number of pixel pairs used in the calculation. This coefficient r ranges from − 1 to 1, with r = 1
indicating a perfect positive correlation, r = − 1 indicating a perfect negative correlation and r = 0 indicating no
correlation. In plain images, adjacent pixels are often highly correlated, especially in the horizontal, vertical and
diagonal directions. However, an effective encryption algorithm should reduce these correlations to near zero,
disrupting the spatial relationships between pixels. Table 9 shows the correlation coefficients calculated for the
original and encrypted images in different directions.

6.4. Experimental analysis and comparison results

In this section, we present the results of the security analysis for the proposed image encryption algorithm based
on an 11-dimensional hyperchaotic system. The results covering key space analysis, histogram analysis,

34
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 15. Histograms of Red, Green and Blue color components for the original and encrypted images: (a) House, (b) Encrypted
House, (c) Jelly Beans, (d) Encrypted Jelly Beans, (e) Mandrill, (f) Encrypted Mandrill, (g) Grayscale Clock and (h) Encrypted Clock.

Table 8. Information entropy of

encrypted images.

Encrypted images Entropy value

House 7.9992
Jelly Beans 7.9991
Mandrill 7.9998
Clock 7.9975

information entropy and correlation analysis are compared with existing encryption techniques to demonstrate
the robustness and efficiency of the proposed scheme.
Key space analysis showed that the key space of the encryption algorithm is 39 times larger than the standard
size of 2100 which provides a strong protection against brute force attacks. This significant expansion in key space
is necessary to ensure long term security in modern cryptographic applications. The comparative analysis of key
space sizes with other encryption methods is presented in table 10, where the proposed method is showing clear
superiority.
The histogram analysis of both original and encrypted images shown in figure 15 shows a significant change.
For the original image, the histograms of the red, green and blue components show distinct patterns
corresponding to the pixel distribution of the image. However, after encryption, the histograms of all color
components become evenly distributed, which confirms that the encryption successfully hides the statistical
properties of the original image. This uniformity across all color channels (red, green and blue) demonstrates the
effectiveness of the proposed method in resisting statistical attacks.
In terms of information entropy, the proposed encryption algorithm results in a value close to 8 listed in
table 8, which is close to the ideal level for an encrypted image with true randomness. This high entropy ensures
that the pixel distribution of the encrypted image is highly unpredictable, reducing the probability of successful
cryptanalysis. Comparison of entropy values with previous work, as shown in table 10, highlights the improved
randomness of the proposed encryption method.
Finally, correlation analysis demonstrated that the correlation between adjacent pixels in the encrypted
image is close to zero, indicating a strong decorrelation of pixel values. This result depicted in figure 16 and
summarized in table 9 underlines the ability of the proposed method to resist statistical and differential attacks.

35
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Table 9. Correlation coefficients of adjacent pixels in different directions for

original and encrypted images.

Original Images

Horizontal Vertical Diagonal

Image Correlation Correlation Correlation

House 0.97823 0.95293 0.93605

Jelly Beans 0.97883 0.98244 0.96474
Mandrill 0.86654 0.75873 0.72619
Clock 0.95649 0.97408 0.93893

Encrypted Images

Image Horizontal Vertical Diagonal

Correlation Correlation Correlation
House − 0.0021 − 0.0024 − 0.0063
Jelly Beans − 0.0084 − 0.0014 0.0022
Mandrill 0.0004 0.0005 0.0017
Clock 0.0013 0.0051 0.0053

Table 10. Comparison of security measures.

Information Correlation
Algorithm Key Space Entropy (Horizontal)

[15] −−− 7.9972 0.0000104

[40] 21026 7.9968 0.0020
[41] 2697 7.9998 − 0.0182
[42] 2205 7.9994 0.0008
[43] 2541 7.9994 − 0.0025
Proposed Scheme 23905 7.9998 0.0004

As seen in table 9, the original image exhibits strong correlations in all directions, particularly in the vertical
direction with a correlation of 0.98244. In contrast, the encrypted image shows near zero correlations (e.g.,
0.0004 in the horizontal direction), confirming that the encryption process successfully randomizes the pixel
values and eliminates any discernible patterns. Comparison of the correlation results with existing encryption
methods as shown in table 10, further highlights the effectiveness of the proposed encryption algorithm. Low
correlation is a key factor in the overall security strength of encryption.

Remark 5. The private key of S, which cannot be used in any other part of the suggested ciphers, is used only for
encryption. This ensures non-repudiation, which means that the S cannot deny encrypting the image. As a
result, the integrity and authenticity of the encryption process is maintained.

Remark 6. Additionally, only the receiver can recover the image I by computing
D (I ) = (E (I ) - u2) * u1-1 (mod q ) using the private key (υ1, υ2) and the solutions of the response systems (20),
(21). This ensures that only S and R can access the communication, thereby maintaining confidentiality. This
secure method prevents unauthorized access and ensures that the transmitted information remains private and
protected.

Remark 7. If an attacker attempting to recover the original image must calculate (E (I ) - u2) u1-1 (mod q ). Since
υ1 and υ2 are private keys of R, obtaining them directly is impossible. Even with knowledge of M1 and M2, the
attacker cannot derive υ1 and υ2 because they are generated from the response systems (20) and (21) and the
public key (M1, M2). The solutions of the hyperchaotic delayed system at time t1 remain unknown, thus
preventing the attacker from recovering the original image.

Remark 8. Reconstructing the original image from (E (I ) - u2) u1-1 (mod q ) is practically impossible for an
attacker who possesses only (υ1, υ2). Computing u1-1 (mod q ) requires knowledge of q and solving integer
factorization problems is extremely challenging. Notably, q is shared exclusively between S and R and not with
others. Hence, any attempts by an unauthorized party to decrypt the image will be blocked.

36
Phys. Scr. 99 (2024) 125008 J K Bharti et al

Figure 16. Correlation analysis for original and encrypted images in horizontal, vertical and diagonal directions: (a) House, (a1)
Encrypted House, (b) Jelly Beans, (b1) Encrypted Jelly Beans, (c) Mandrill, (c1) Encrypted Mandrill, (d) Clock and (d1) Encrypted
Clock.

These components collectively ensure the security and integrity of the encryption scheme, protecting the
transmission of sensitive information.

7. Conclusion

The framework for matrix projective combination-combination synchronization of 11-dimensional

hyperchaotic systems is established by extending the 6-dimensional system by adding via complex variables and
time delay. Synchronization is achieved using an active nonlinear control approach involving four identical 11-
dimensional hyperchaotic delayed systems, two of which act as drive system and the other two as response
system. The controllers are designed to stabilize the error states and manage the delayed hyperchaotic system
which is verified through Lyapunov stability theory.
The dynamical analysis including time response, phase portraits, Lyapunov exponents, Kaplan Yorke
dimensions, symmetry, equilibrium points and dissipation demonstrated that the 11-dimensional system
exhibits hyperchaotic behavior under specific parameter values and initial conditions. Incorporating time delays
into the hyperchaotic system increased the complexity, making it particularly effective for secure
communications such as image encryption. This time delayed hyperchaotic system is applied to image
encryption using an affine cipher algorithm, in which secret key is generated from the private messages of the
sender and receiver. Depending on the number of words (including spaces and punctuation mark) in the private
message, four cases emerge, with examples given for each of them. To validate the algorithm, an original images
are used and numerical illustrations are performed to demonstrate the effectiveness of the proposed time
delayed hyperchaotic system in enhancing image encryption.
Security analysis such as key space, histogram, information entropy and correlation analysis and
experimental results showed that the proposed encryption method ensured a high level of security which
effectively protected the original image during the encryption and decryption process. This method also
provides flexibility to accommodate varying message lengths, further increasing its robustness.
However, the increased complexity of time delays introduces computational demands that may limit the
efficiency of the method in real time applications. Additionally, while the encryption method exhibits strong
security, it may be vulnerable to sophisticated cryptographic attacks, necessitating further research to enhance its
robustness against such threats. Future work will focus on enhancing the encryption algorithm by incorporating

37
Phys. Scr. 99 (2024) 125008 J K Bharti et al

fractional order time delayed chaotic systems and addressing model uncertainties and disturbances.
Additionally, the method will be extended to protect not only images but also video and audio during
transmission, aiming to provide stronger data security across multiple media formats in online
communications.

Declarations

• Funding: No funding was received to assist with the preparation of this manuscript.
• Conflict of interest: The authors declare that they have no conflict of interest.
• Ethical approval: This article does not contain any studies with human/animals participants performed by
any of the authors.
• Author contributions: Jyotsna Kumari Bharti: Conceptualization, Formal analysis, Methodology, Software,
Validation, Writing - Original Draft, Writing - Review & Editing. P Balasubramaniam: Investigation,
Validation, Supervision, Writing - Original Draft Preparation, Writing - Review & Editing, Project
administration, Funding acquisition. K Murugesan: Investigation, Validation, Supervision, Writing -
Original Draft Preparation, Writing - Review & Editing, Project administration, Funding acquisition. All
authors read and approved the final manuscript.
• Availability of data and materials: Data sharing is not applicable to this article as no new data were created or
analyzed in this study.

Acknowledgments

All authors would like to thank the National Board for Higher Mathematics (NBHM), Department of Atomic
Energy, Government of India, for supporting the research project (File No. 02011/21/2021 NBHM (R.P)/ R&D
II/9243). We sincerely thank the Editor and the anonymous reviewers for their thorough review and valuable
feedback, which significantly enhanced the quality of this manuscript.

Data availability statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

ORCID iDs

Jyotsna Kumari Bharti https://orcid.org/0009-0006-4561-7241

P Balasubramaniam https://orcid.org/0000-0001-9673-5949
K Murugesan https://orcid.org/0000-0002-6122-6455

References
[1] Pecora L M and Carroll T L 1990 Synchronization in chaotic systems Physical review letters 64 821
[2] Li H-L, Hu C, Zhang L, Jiang H and Cao J 2022 Complete and finite-time synchronization of fractional-order fuzzy neural networks via
nonlinear feedback control Fuzzy Sets and Systems 443 50–69
[3] Li G-H and Zhou S-P 2007 Anti-synchronization in different chaotic systems Chaos, Solitons and Fractals 32 516520
[4] Wang Y-W and Guan Z-H 2006 Generalized synchronization of continuous chaotic system Chaos, Solitons and Fractals 27 97–101
[5] Yang X, Li X and Duan P 2022 Finite-time lag synchronization for uncertain complex networks involving impulsive disturbances
Neural Computing and Applications 34 5097–106
[6] Mainieri R and Rehacek J 1999 Projective synchronization in three-dimensional chaotic systems Phys. Rev. Lett. 82 3042–5
[7] Lei Y, Xu W and Zheng H 2005 Synchronization of two chaotic nonlinear gyros using active control Physics Letters A 343 153158
[8] Chen X, Park J H, Cao J and Qiu J 2017 Sliding mode synchronization of multiple chaotic systems with uncertainties and disturbances
Applied mathematics and Computation 308 161–73
[9] Dai H, Jia L-X and Zhang Y-B 2012 Adaptive generalized matrix projective lag synchronization between two different complex
networks with non-identical nodes and different dimensions Chinese Physics B 21 120508
[10] Yan S, Wang J, Wang E, Wang Q, Sun X and Li L 2023 A four-dimensional chaotic system with coexisting attractors and its
backstepping control and synchronization Integration 91 67–78
[11] Qi F, Qu J, Chai Y, Chen L and Lopes A M 2022 Synchronization of incommensurate fractional-order chaotic systems based on linear
feedback control Fractal and Fractional 6 221
[12] Wu K, Ren C, Nan Y, Li L, Yuan S, Shao S and Sun Z 2024 Experimental research on vehicle active suspension based on time-delay
control International Journal of Control 97 1157–73

38
Phys. Scr. 99 (2024) 125008 J K Bharti et al

[13] Zhang C, Wang X, Ye X, Zhou S and Feng L 2020 Robust modified function projective lag synchronization between two nonlinear
complex networks with different-dimensional nodes and disturbances ISA transactions 101 42–9
[14] Liu S, Li Y, Ge X, Li C and Zhao Y 2022 A novel hyperchaotic map and its application in fast video encryption Physica Scripta 97 085210
[15] Zhang N, Liu J, Tong X, Jiao W and Gan H 2023 A novel lorenz-sine coupling chaotic system and its application on color image
encryption Physica Scripta 98 095217
[16] Farah M B, Farah A and Farah T 2020 An image encryption scheme based on a new hybrid chaotic map and optimized substitution box
Nonlinear Dynamics 99 3041–64
[17] Joshi A B, Kumar D, Gaffar A and Mishra D 2020 Triple color image encryption based on 2d multiple parameter fractional discrete
fourier transform and 3d arnold transform Optics and Lasers in Engineering 133 106139
[18] Shi X and Wang Z 2023 Hybrid projective synchronization of fractional-order extended hindmarsh-rose neurons with hidden
attractors Axioms 12 157
[19] Hao C, Wang B and Tang D 2024 Finite-time and fixed-time function projective synchronization of competitive neural networks with
noise perturbation Neural Computing and Applications 1–17
[20] Zhang H, Chen X, Ye R, Stamova I and Cao J 2023 Quasi-projective synchronization analysis of discrete-time focvnns via delay-
feedback control Chaos, Solitons & Fractals 173 113629
[21] Shukla V K, Mbarki L, Shukla S, Vishal K and Mishra P K 2023 Matrix projective synchronization between time delay chaotic systems
with disturbances and nonlinearity International Journal of Dynamics and Control 11 1926–33
[22] Hao Y, Fang Z and Liu H 2024 Adaptive ts fuzzy synchronization for uncertain fractional-order chaotic systems with input saturation
and disturbance Information Sciences 666 120423
[23] Baluni S, Sehgal I, Yadav V K and Das S 2024 Exponential synchronization of a class of quaternion-valued neural network with time-
varying delays: a matrix measure approach Chaos, Solitons & Fractals 182 114794
[24] Razmara S and Yahyazadeh M 2022 Design of an analog time-varying audio cryptography system based on sliding mode
synchronization of non-identical chaotic systems described with time-delayed fractional-order dynamics Arabian Journal for Science
and Engineering 47 10471–84
[25] Yang Y, Gao J and Imani H 2023 Design, analysis, circuit implementation, and synchronization of a new chaotic system with
application to information encryption AIP Advances 13
[26] Yao X, Chen X, Liu H, Sun L and He L 2022 Adaptive sliding-mode synchronization of the memristor-based sixth-order uncertain
chaotic system and its application in image encryption Frontiers in Physics 10 269
[27] Yadav V K, Prasad G, Srivastava M and Das S 2019 Triple compound synchronization among eight chaotic systems with external
disturbances via nonlinear approach Differential equations and dynamical systems 30 549–572
[28] Singh A K, Yadav V K and Das S 2017 Dual combination synchronization of the fractional order complex chaotic systems Journal of
Computational and Nonlinear Dynamics 12 011017
[29] Pan W, Li T and Wang Y 2022 The multi-switching sliding mode combination synchronization of fractional order non-identical
chaotic system with stochastic disturbances and unknown parameters Fractal and Fractional 6 102
[30] Adeniji E A, Ogundipe S O, Ajayi A A, Ojo K S and Akinrinlola S A 2023 Combination-combination synchronization of chaotic
fractional order systems Nigerian Journal of Physics 32 57–65
[31] Khattar D, Agrawal N and Singh G 2023 Triple compound combination synchronization of eleven n-dimensional chaotic systems
International Journal of Dynamics and Control 11 2499–513
[32] Ansari S P and Das S 2015 Projective synchronization of time-delayed chaotic systems with unknown parameters using adaptive
control method Mathematical methods in the applied sciences 38 726–37
[33] Li Z, Tang Y, Xu F and Zhang Y 2024 Full states pseudo-random projective synchronization of hyperchaotic system and corresponding
secure communication algorithm Multimedia Tools and Applications 1–42
[34] Han J, Chen G, Wang L, Zhang G and Hu J 2023 Direct approach on fixed-time stabilization and projective synchronization of inertial
neural networks with mixed delays Neurocomputing 535 97–106
[35] Lei H and Jia N 2023 Synchronization of complex networks with dynamic parameters uncertainty and mixed delays coupling
International Journal of Dynamics and Control 12 382–95
[36] Yadav N and Pallav Handa H 2023 Projective synchronization for a new class of chaotic/hyperchaotic systems with and without
parametric uncertainty Transactions of the Institute of Measurement and Control 45 1975–85
[37] Zheng P, Guo X and Wen G 2022 Fixed-time adaptive time-varying matrix projective synchronization of time-delayed chaotic systems
with different dimensions CMES-Computer Modeling In Engineering & Sciences 131 1451–63
[38] He J and Pei L 2023 Dual function matrix projective synchronization for fractional-order hyperchaotic systems Journal of
Computational and Nonlinear Dynamics 18
[39] Khan A, Ali S, Nigar U and Khan A 2023 Combination difference synchronization between hyperchaotic complex lü time-delay
systems via adaptive control Palestine Journal of Mathematics 12 147–58
[40] Ullah S, Liu X, Waheed A, Zhang S and Li S 2024 Novel grayscale image encryption based on 4d fractional-order hyperchaotic system,
2d henon map and knight tour algorithm Physica Scripta 99 095248
[41] Yousfi H, Islam Y, He S, Gasri A and Hassan M M 2024 Advanced medical image encryption techniques using the fractional-order
halvorsen circulant systems: dynamics, control, synchronization and security applications Physica Scripta 99 055208
[42] Li H, Yu S, Feng W, Chen Y, Zhang J, Qin Z, Zhu Z and Wozniak M 2023 Exploiting dynamic vector-level operations and a 2d-
enhanced logistic modular map for efficient chaotic image encryption Entropy 25 1147
[43] Feng W, Wang Q, Liu H, Ren Y, Zhang J, Zhang S, Qian K and Wen H 2023 Exploiting newly designed fractional-order 3d lorenz
chaotic system and 2d discrete polynomial hyper-chaotic map for high-performance multi-image encryption Fractal and Fractional
7 887
[44] Wen H and Lin Y 2023 Cryptanalyzing an image cipher using multiple chaos and dna operations Journal of King Saud University-
Computer and Information Sciences 35 101612
[45] Wen H and Lin Y 2024 Cryptanalysis of an image encryption algorithm using quantum chaotic map and dna coding Expert Systems with
Applications 237 121514
[46] Sabaghian A, Balochian S and Yaghoobi M 2020 Synchronisation of 6d hyper-chaotic system with unknown parameters in the presence
of disturbance and parametric uncertainty with unknown bounds Connection Science 32 362–83
[47] Evans D J, Cohen E, Searles D J and Bonetto E 2000 Note on the kaplan-yorke dimension and linear transport coefficients Journal of
Statistical Physics 101 17–34

39
Phys. Scr. 99 (2024) 125008 J K Bharti et al

[48] Hale J K 2006 Functional differential equations Analytic Theory of Differential Equations: The Proceedings of the Conference at Western
Michigan University (Springer) 9–22
[49] Silverman J H, Pipher J and Hoffstein J 2008 An introduction to mathematical cryptography Vol 1 (Springer)
[50] Muthukumar P, Balasubramaniam P and Ratnavelu K 2015 Fast projective synchronization of fractional order chaotic and reverse
chaotic systems with its application to an affine cipher using date of birth (DOB) Nonlinear Dynamics 80 1883–97
[51] Usc: Usc-sipi database (http://sipi.usc.edu/database/)

40