Computers, Materials & Continua Tech Science Press

DOI: 10.32604/cmc.2023.036861
Article

A Novel Lightweight Image Encryption Scheme

Rawia Abdulla Mohammed1, *, Maisa’a Abid Ali Khodher1 and Ashwak Alabaichi2

1
Department of Computer Science, University of Technology, Baghdad, Iraq
2
Department of Biomedical Engineering, University of Kerbala, Kerbala, Iraq
*Corresponding Author: Rawia Abdulla Mohammed. Email: [email protected]
Received: 14 October 2022; Accepted: 14 December 2022

Abstract: Encryption algorithms are one of the methods to protect data

during its transmission through an unsafe transmission medium. But encryp-
tion methods need a lot of time during encryption and decryption, so it
is necessary to find encryption algorithms that consume little time while
preserving the security of the data. In this paper, more than one algorithm
was combined to obtain high security with a short implementation time. A
chaotic system, DNA computing, and Salsa20 were combined. A proposed
5D chaos system was used to generate more robust keys in a Salsa algorithm
and DNA computing. Also, the confusion is performed using a new S-
Box. The proposed chaos system achieves three positive Lyapunov values.
So results demonstrate of the proposed scheme has a sufficient peak signal-
to-noise ratio, a low correlation, and a large key space. These factors make
it more efficient than its classical counterpart and can resist statistical and
differential attacks. The number of changing pixel rates (NPCR) and the
unified averaged changed intensity (UACI) values were 0.99710 and UACI
33.68. The entropy oscillates from 7.9965 to 7.9982 for the tested encrypted
images. The suggested approach is resistant to heavy attacks and takes less
time to execute than previously discussed methods, making it an efficient,
lightweight image encryption scheme. The method provides lower correlation
coefficients than other methods, another indicator of an efficient image
encryption system. Even though the proposed scheme has useful applications
in image transmission, it still requires profound improvement in implementing
the high-intelligence scheme and verifying its feasibility on devices with the
Internet of Things (IoT) enabled.

Keywords: Chaotic system; lightweight; confusion; diffusion; encryption;

Salsa20

1 Introduction
Recent technological advances have profusely altered global communication, where One of the
most pivotal advancements of this century is the frequent use of small computing devices. These
devices are becoming essential for both individual customers and global organizations. Hence, these

This work is licensed under a Creative Commons Attribution 4.0 International License,
which permits unrestricted use, distribution, and reproduction in any medium, provided
the original work is properly cited.
2138 CMC, 2023, vol.75, no.1

technological advances led to immersing the general public in wired and wireless communication
routes. The widespread use of such devices compels cryptographers to devise methods for safeguarding
the data transmission process. Correspondence is used to transmit official and private communication
between organizations and individuals. The sensitivity of some of these data makes a secure trans-
mission form necessary due to concerns about unauthorized access to data [1]. Recently, image-based
communication has become increasingly popular in private and public scenarios. This connection can
be seen in money transactions, medical reports, and military exercises. Images transferred between
parties without a secure connection are constantly at risk of stealing or losing their data. Therefore
image encrypting has of major importance [2].
Salsa20 appears to be an appropriate candidate for image encoding due to its intriguing
structure [3].
The chaotic encryption technique has been used extensively for image encryption due to its proper-
ties (initial conditions sensitivity, pseudo-randomness, and non-periodicity). These properties conform
to the required encryption characteristics, which can be found in the chaotic encryption technique
[4,5]. DNA cryptography techniques are becoming a viable network method. They provide a high
level of parallelism, which helps regulate computational speed. Its capacity for storing information via
DNA molecules is very large. This capacity is accompanied by low power consumption. Compared
to standard encryption algorithms like data encryption standard (DES) and advanced encryption
standard (AES), DNA cryptography is a new method for unbroken data that provides exceptional
network security. Scientists exploit the benefits of DNA coding and chaotic systems by combining
them into image cryptosystems to create more powerful and secure systems that are difficult to access
or break [6]. Since the 2010s, many image encryption schemes have been proposed, but many of them
have not been effective due to their lower key space and computational complexity. According to
previous literature [7,8], a secure cryptosystem must contain confusion and diffusion. Some schemes in
the literature [9,10] do not fulfill the abovementioned requirements. Some schemes are also unsuitable
for Internet of Things (IoT) devices and real-time applications.
In this paper, a novel lightweight image encryption approach is suggested. This approach inte-
grates the chaotic system with Salsa20 and DNA. This integration improves the security and efficiency
of image encryption by utilizing the high diffusion supplied by the XOR operation and the Lorenz
system. It also incorporates the high confusion provided by a new S-Box. The contributions of this
research are as follows:
• Present a new lightweight image encryption/decryption scheme that requires less time and less
memory and has high efficiency. The scheme is also designed to be more secure and ensure lower
correlation coefficients between adjacent pixels of the encrypted image.
• Creating a highly sensitive key for encrypting and decrypting images based on the innovative 5D
Lorenz map system has more advantages than the simple chaotic system. A greater parameter,
space, high randomizations, and a vast number of chaotic sequences are among the many
advantages of this system.
• Significantly increase the key space with a proposed novel 5D Lorenz map system; this will
protect images from unauthorized access (Henon map, sin map, ten maps, etc.).
• Propose new S-Boxes to synthesize efficient 16 ×16 S-Boxes based on the 2D Henon map. These
boxes are compared with recent methods.
• Evaluating the proposed scheme with many metrics, such as entropy, correlation, and differen-
tial attack.
CMC, 2023, vol.75, no.1 2139

The rest of the paper consists of the following sections: literature review, theoretical background,
and proposed encryption and decryption scheme. The practical system performance is evaluated; the
last section concludes the paper.

2 Literature Review
Images are a key part of communication. The general public and organizations both rely on
transmitting a massive amount of data in the form of images. Some sensitive and important images
necessitate a secure way of transmission. Numerous encryption algorithms are available in the
literature to obtain security and confidentiality. A concise literature review is provided hereafter.
Alireza et al. [9] proposed an efficient digital image encryption based on Salsa20. The results and
analysis demonstrate that the algorithm has an acceptable level of security. Zheng et al. [10] proposed
a lightweight authenticated encryption scheme for railway cloud service based on a novel discrete
chaotic S-Box coupled map lattice (SCML). Zhang et al. [11] an image encryption scheme based on
bit permutation and dynamic DNA encoding. Security analysis indicated that the algorithm could
withstand attack operations such as statistical analysis and exhaustive analysis. Janakiraman et al. [12]
proposed a chaotic lightweight algorithm and its implementation on a 32-bit microcontroller to
encrypt grayscale images. The scheme’s performance is sufficient for real-time applications, but it is
not robust because it is based on spatial domain techniques. Patro et al. [13] proposed a combined
hyper-chaos and chaos-based encryption technique to secure images. One round of diffusion and
multi-stage bit-plane permutation operations are performed to obtain better encryption results, but
this method is insecure due to the limited key space. Qasim et al. [3] proposed a hybrid encryption
algorithm that consisted of modified Salsa20 and chaos theory; it has been applied by encrypting
medical information. Most tests demonstrate that the messy Salsa proposed is faster than the original.
An improved algorithm was designed by Lin [14]. Its purpose was to analyze the existing cryptographic
methods based on chaotic maps and resist the chosen plaintext attack (CPA). In this scenario, an
improved CIES-UBPRPD (chaotic map-based image encryption system employing both plaintext-
related permutation and diffusion) approach is used to achieve higher plaintext sensitivity than the
original CIES-UBPRPD method. Despite its resilience and high security, it is more time-consuming
than the original CIES-UBPRPD method. Guan et al. [15] developed image encryption in the
frequency domain by combining the techniques of 4D hyper chaotic maps and DNA encoding.
Dagadu et al. [16] proposed an image encryption scheme based on a pseudo-randomly enhanced
logistic map, random permutation, and DNA. Zheng et al. [17] proposed an image encryption
scheme based on a multi-chaotic system and DNA coding. This method has high security, but it is
not suitable for color images. Liu et al. [18] designed a cost-effective, lightweight image encryption
scheme regarding time and storage, which was based on message passing (MP) and chaotic maps.
Gupta et al. [19] proposed a fast, secure, and lightweight symmetric image cryptographic algorithm
based on the session key. Ravichandran et al. [20] Suggested using integer wavelet transforms (IWT),
DNA computing, and chaos to form an effective medical image encryption method. This plan is
impervious to CPA, but it is not used for color images. Ferdush et al. [21] presented a standard
framework and algorithm based on two chaotic maps, such as Arnold and logistic, for lightweight
image encryption. This scheme is implemented for a grayscale image only. Abdallah et al. [22]
suggested incorporating numerous shuffling operations based on the 3D-Lornez chaos theory, the
initial permutation (IP) and S-Box ideas, and the confusion and diffusion operations in the key process.
State-of-the-art research indicates that some encryption algorithms are slow or unsuitable for
color images, and those others are less resistant to plaintext and differential assaults. Some encryption
2140 CMC, 2023, vol.75, no.1

algorithms are also unsuitable for Internet of Things devices and real-time applications. The cryptosys-
tem must be resistant to statistical and differential assaults as a result. The cryptosystem must also be
extremely sensitive to its secret keys and use a larger key space to prevent data from being retrieved by
an unauthorized user.

3 Theoretical Background
3.1 Chaos System
Chaos theory is a mathematical discipline that includes the study of complex systems. The main
components of chaos theory are sensitivity to the initial value, parameters, and randomness. Applying
minor input adjustments lead to greatly alters the systems’ outputs. Systems based on chaos theory
are more secure for image encryption; unauthorized individuals cannot predict the chaos sequence if
they are unaware of the proper control parameters and initial values [23].

3.1.1 Lorenz Chaotic Mapping

Lorenz mapping is a typical example of chaotic mapping in chaotic systems, and the system
dynamic equations are:
x = α (y − x) (1)

y = xz + βx − y (2)

z = xy − γz (3)

System parameters are part of the mapping; their typical values are 10, 28, and 8/3. When they
remain constant, the system collapses when the criterion of 24.74 is met. The Lorenz system generates
chaotic sequences with a more complicated system structure than the low-dimensional one, which
can combine chaotic sequences with either one or more variables. The sequence design is highly
flexible [24].

3.1.2 H’enon map

The term Hénon-Pomeau attractor/map is frequently used to refer to a discrete-time dynamic
system. It is among the most studied models of chaotic behavior in dynamic systems. The H’enon
system maps a point in the plane (Xn ,Yn ) to a new point [25]. The H’enon map is a 2D discrete chaotic
map, which is defined by the following equation:
Xn+1 = 1 − aXn2 + Yn (4)

Yn+1 = bxn (5)

When parameter b = 0.3 and a ∈ [1.06,1.22]∪ [1.27,1.29] ∪ [1.31,1.42], the H’enon map is
chaotic [26].

3.2 Salsa20 Algorithm

The stream cipher Salsa(20) uses a counter mode for encryption. Salsa(20) initial seed is an array
(4, 4) with 512 bits. Edition, XOR, and rotation are the fundamental operations of Salsa(20), and they
are used on an array of Salsa(20) for 10 cycles. Salsa(20) refers to the fact that its array is altered twice
CMC, 2023, vol.75, no.1 2141

in each round. At the end of the Salsa(20), the final adjustment of the array and its initial seed are
added [27].

3.3 DNA Rules

The nucleic acid bases that make up a DNA sequence are broken down into four categories:
adenine (A), cytosine (C), thymine (T), and guanine (G). The letter A and the letter T, as well as
the letters G and C, are complementary. Because the binary numbers 0 and 1 are complimentary to
one another, the binary numbers 00 and 11 and 01 and 10 are also complementary. A total of 24
possible coding combinations may be achieved by encoding the numerals 00, 01, 10, and 11 using the
four bases A, T, C, and G. Only eight of these combinations can satisfy the Watson-Crick complement
requirement [24]. Those combinations are shown in Table 1.

Table 1: a-Coding, b-Add, c-Subtract operation in DNA [24]

a-Coding b-Add c-Subtract operation

Code Bits Add A T C G Sub A T C G

A 00 A A T C G A A T G C
T 01 T T A G C T C A T G
C 10 C C G A T C G C A T
G 11 G G C T A G T G C A

4 Proposed S-Box Algorithm and its Performance

4.1 Construction of S-Box
There is a viable theory about the construction of a superior S-Box. According to the outcomes of
the final operation of the supplied scheme requirements for creating a new dynamic S-Box based on
chaos theory principles, an S-Box that satisfies this criterion is resistant to differential cryptanalysis.
This operation provides increased protection and complexity based on the two-dimensional chaotic
H’enon system. Using the H’enon chaotic map (2D) and the initial value X0 for the chaotic system
and numbers created, one can generate a big 1616 S-Box. It similarly consists of a range of numbers
(0–255) and uses the S-Box output method to produce the H’enon system values. All values contained
within the S-Box must be distinct and not repeated. The responsive dependency on the initial State with
chaos theory changes the construction of the S-Box, and the result of the dynamic S-Box is inversed
with every slight change in the initial value. S-Box and its inversed form are listed in Tables 2 and 3.

Table 2: The S-Box generated by the proposed scheme

0 1 2 3 4 5 6 7 8 9 A B C D E F
0 01 DD B8 7A 51 A9 B3 16 AE A7 9B 45 DC 7B 13 7E
1 CB 71 8D 92 12 43 40 A0 C0 BB C4 77 FD BA FC 9D
2 C1 75 D5 6E AD A2 D6 D8 6D C7 50 2E 66 2F 5B 49
(Continued)
2142 CMC, 2023, vol.75, no.1

Table 2: Continued
0 1 2 3 4 5 6 7 8 9 A B C D E F
3 39 85 63 BC 21 10 C6 06 37 4E 04 B0 B7 DB 79 23
4 11 70 1A 3B 24 94 18 C9 97 8F 31 02 4D 32 8B 9C
5 F3 E9 68 E5 54 65 EA 0D 62 60 47 ED AC FE 41 A5
6 22 F8 29 9A 81 EE D3 0F 56 5D E2 8C 64 D7 FA DA
7 B6 C3 53 05 3A A1 99 A6 74 57 AB 17 F9 6C C8 F2
8 84 CA 98 1D 4B 59 0A B2 D9 86 61 BD 48 8E D0 7F
9 46 E8 0C 38 AF 0B 3F 14 91 F6 DF 03 2C 73 90 25
A C2 D1 82 E6 19 5A 6F 8A 4C 83 27 3E 72 E3 F0 CE
B 3D 26 96 2A 08 52 BF 2B 6B F1 9E 07 F5 5E A8 BE
C 09 9F C5 4F 7D 89 EF 44 D2 DE 36 34 6A 30 7C D4
D F4 EC E7 76 F7 FB 93 CC 42 1F 4A B4 E0 B5 28 CF
E 2D 20 88 B9 87 5C A4 A3 1E 3C B1 69 0E 67 55 95
F 1B 5F 00 80 E4 1C 15 33 E1 EB CD 35 58 78 AA FF

Table 3: The inverse of the proposed S-Box

0 1 2 3 4 5 6 7 8 9 A B C D E F
0 F2 00 4B 0B 3A 73 37 BB B4 C0 86 95 92 57 EC 67
1 35 40 14 0E 97 F6 07 7B 46 A4 42 F0 F5 83 E8 D9
2 E1 34 60 3F 44 9F B1 AA DE 62 B3 B7 9C E0 2B 2D
3 CD 4A 4D F7 CB FB CA 93 30 74 43 E9 B0 AB 96
4 16 5E D8 15 C7 0B 90 5A 8C 2F DA 84 A8 4C 39 C3
5 2A 04 B5 72 54 EE 68 79 FC 85 A5 2E E5 69 BD F1
6 59 8A 58 32 6C 55 2C ED 52 EB CC B8 7D 28 23 A6
7 41 11 AC 9D 78 21 D3 1B FD 3E 03 0D CE C4 0F 8F
8 F3 64 A2 A9 80 31 89 E4 E2 C5 A7 4E 6B 12 8D 49
9 9E 98 13 D6 45 EF B2 48 82 76 63 0A 4F 1F BA C1
A 17 75 25 E7 E6 5F 77 09 BE 05 FE 7A 5C 24 08 94
B 3B EA 87 06 DB DD 70 3C 02 E3 1D 19 33 8B BF B6
C 18 20 A0 71 1A C2 36 29 7E 47 81 10 D7 FA AF DF
D 8E A1 C8 66 CF 22 26 6D 27 88 6F 3D 0C 01 C9 9A
E DC F8 6A AD F4 53 A3 D2 91 51 56 F9 D1 5B 65 C6
F AE B9 7F 50 D0 BC 99 D4 61 70 6E D5 1E 1C 5D FF

4.2 Performance Analysis of Proposed S-Box

The strength of the suggested S-Box is assessed using multiple conventional performance analyses.
The assessments in this work include the following criteria: balanced criterion (BC), completeness
criterion (CC), avalanche criteria (AC), strict avalanche criteria (SAC), nonlinearity, and bijective
properties.
CMC, 2023, vol.75, no.1 2143

4.2.1 BC
One of the most critical S-Box tests is to ensure that the distribution of 0s and 1s in the output
sequences is balanced. The results of this test (which employed two words with the new S-Box) reveal
that the new S-Box is balanced, as presented in Table 3.

4.2.2 CC
This standard ensures comprehensiveness. When something is complete, every bit of output is
tested against every bit of input [28]. As seen in Tables 1 and 2, the created S-Box passes this test.

4.2.3 AC
The non-relationship between input bits and the output sequence is essential to a good block
cipher. A block cipher is evaluated using the avalanche criterion. In the avalanche criterion, a small
change in plaintext results in a large change in cipher text. A small change may include flipping a
single bit from 0 to 1 or vice versa, which causes a large change in output. The value of this criterion
is calculated using Eq. (6) and should be in the range of 0–1, with 0.5 being the best value.
Number of Flipped Bits in Cipher Text
AC = (6)
Number of All Bits in Cipher Text

To test the proposed S-Box, a single bit is changed from the letter “L” to the letter “M” Both
are then replaced with data from the proposed S-Box. In five of the original eight bits, the result of
“L” differed from “M” As a result of Eq. (6), the AC is 0.625. The formed S-Box meets the avalanche
condition. The results of this test were compared to the findings of other relevant studies, as presented
in Table 4.

Table 4: Comparative BC-AC results

Method Balanced criterion (BC) Avalanche criteria (AC)

Words

Computer BMAOPRN

0’s 1’s 0’s 1’s

Ref. [29] 34 30 27 37 0.5
Ref. [30] 35 29 28 36 0.375
Ref. [31] 28 36 40 24 0.375
Proposed S-Box 32 32 32 32 0.625

4.2.4 SAC
Whenever the AC and CC are obtained, the SAC is obtained. The S-Box meets the SAC when
switching one bit from the input results in a 50% change in the output bits [28]. Our proposal meets
the requirements AC, CC, and SAC.
2144 CMC, 2023, vol.75, no.1

4.2.5 Nonlinearity Property

To resist linear cryptographic attacks, the nonlinearity of the Boolean function must be large [32].
The nonlinearity scores of the S-Box obtained via H’enon chaotic map are 110, 106, 106, 108, 108, 106,
and 106, with an average of 107.25. This average was the objective function’s goal .When compared to
other techniques, the H’enon chaotic map method is considered. acceptable outcomes for achieving
high nonlinearity.

4.2.6 Bijective Property

Generated S-Boxes for the H’enon chaotic map have different output values from the interval
[0,255] with no repetition. Both S-Boxes satisfy the bijectivity property.

5 The Proposed Lightweight Scheme

This work aims to design a Lightweight Image Encryption/Decryption Scheme that can be used
to transfer a secure image in an untrusted channel. This scheme is inspired by a combination of some
functions from Salsa20 and DNA algorithms. This scheme involves breaking the link between the
pixels and is mostly concentrated on concepts such as confusion and dispersion. This scheme was
developed to achieve high encryption efficiency, resistance to various assaults with an appropriate
execution time, memory savings, and complexity reduction. Fig. 1 presents the structure of the
proposed scheme.

Figure 1: General structure of the proposed scheme

5.1 The Proposed Encryption Scheme

The step-by-step proposed lightweight image encryption scheme is presented in Fig. 2.

5.1.1 Key generation using 5D chaotic system

A new 5D chaotic system of differential dynamic, chaotic equations has been proposed to examine
chaotic properties and produce a series of numerical output sequences. The proposed innovative 5D
chaotic system equations are:
X[i + 1] = x [i] + (−s × x [i] + y [i] × k [i] − r × p [i]) × dt (7)

Y[i + 1] = y [i] + (−y [i] − x [i] × z [i] + r × x [i] − u × p [i]) × dt (8)

Z [i + 1] = z [i] + (z [i] × x [i] × y [i] − 1.5 × s × p [i] − k [i]) × dt (9)

K [i + 1] = k [i] + (s × x [i] + (u × y [i] − r × k [i]) × dt (10)

CMC, 2023, vol.75, no.1 2145

Figure 2: The proposed lightweight image encryption scheme

P[i + 1] = p[i] + (b × (x [i] + k [i])/z [i] + y [i]) × dt (11)

where b, r, s, u, and dt comprise the chaos parameter, and x, y , z , k, and p comprise the initial
conditions for the chaos map. The proposed 5D chaotic system was implemented and evaluated, and
the Lyapunov exponents for the initials and parameters were calculated. With the maximal Lyapunov
values (x = 2.1, y = 0.5, z = 1.1, k = 1.1, and p = 0.1) and parameters (b = 0.01, r = 0.5, s = 0.95, u =
1.1, and dt = 0.01), the suggested novel 5D chaotic system possesses super chaotic Lyapunov values
that include five positive values. All proposed systems use the generated chaos keys (K1, K2 . . . K5),
which are then placed in the file for ease of use and referenced in subsequent operations. Fig. 3 displays
the chaotic attractors of each plane of the 5D Lorenz chaotic map.

Figure 3: Chaotic attractors of each plane of the 5D Lorenz chaotic map

2146 CMC, 2023, vol.75, no.1

5.1.2 Salsa20
In the proposed scheme, Salsa20 was applied to generate 512 bits block of keystream (Matrix 4 ∗
4; 16 words). The first four numbers are selected from the initial values of keys to k [0], the second set
of four numbers from the initial keys to k [1], and the third set of four numbers from the initial keys
to k [2], and this pattern is repeated. The final four numbers were selected from the initial keys to k
[7]. The first four numbers are selected from the initial values of the nonce to n [0], and the second set
of four numbers from the initial values of the nonce to n [1]. The first four numbers are selected from
the initial values of block counters to b [0], and the second set of four numbers from the initial values
of block counters to b [1]. For Constants, c [0] = 61707865, c [1] = 3320646e, c [2] = 79622d32, and c
[3] = 6b206574. We applied a Little-endian Function to keys (k [0], k [1] . . . k [7]), nonce (n [0] and n
[1]), and block counters (b [0], b [1]). To generate the initial matrix (S), 512 bits, the block of keystream
(Matrix 4 ∗ 4; 16 words) is ordered as follows:
⎛ ⎞
c0 k0 k1 k2
⎜k3 c1 v0 v1 ⎟
S=⎜ ⎝ i0 i1 c3 k4 ⎠
⎟

k5 k6 k7 c4
In each iteration, 20 rounds came next (r-value), and S was selected as the input (only for the
first iteration). Quarter rounds equations were applied in Salsa20. The columns round equation was
then applied. Every row is transposed as a column to generate (S1) distinct 512 bits, the keystream
block. Note that S1 is the next iteration’s input until the twenty rounds are completed. The Adding
operation is also applied between S1 (after 20 rounds) and S for every word; the result is the Final
Salsa. The inputs in the Add Round Process are State (block of image = 46-bits) and 32 Chaotic keys.
Final Salsa20 has the initial values i = 0 and j = 0. The State is converted from integer to binary for
32 rounds. If the number of the round (i) is even: first, XOR is applied between keys (Z [i] and K [i])
into (CK). Then, the XOR operation is applied between (State and CK) into (State). After this, the
XOR operation applies between keys (X [i] and Y [i]) into (CK). Next, the XOR operation is applied
between (State and CK) into (State). We apply XOR between key P [i] and Final Salsa20 [j] into (CK)
and j = j + 1. Finally, the XOR operation is applied between (State and CK) into State and i = i +1. If
the number of the round (i) is odd, XOR is applied between keys (Z [i] and K [i]) into (CK). Then, the
XOR operation is applied between (State and CK) into (State). The XOR operation is applied between
keys (X [i] and Y [i]) into (CK). Next, XOR is applied between (State and CK) into (State). Finally, if
the value of iis less than or equal to 32, then i = i + 1 or the Final State equals the State.
State1 and a random key from salsa20 are the initial values for DNA encryption. A random key
is selected from Final Salsa20 and converted to the binary form. Padding was applied to ensure the
length of the key was exactly 64 bits. Zeros are added on the left of the key in case the length must
be adjusted. Next, every two-bit block is split and encoded with DNA code using Table 1a. Finally,
the result is named the DNA key. The Final State was also encoded with DNA using Table 1a. The
result is named DNA State. Subsequently, the DNA-Addition operation is applied to the DNA state
and DNA key using Table 1b. The output from this stage is Final State. The Final State input is added
to the new S-Box, where the values replace the pixel values of the Final State matrix in the proposed
S-Box.

5.2 The Proposed Decryption Scheme

The decryption process is the inverse of its encryption process of (LIES). The five-dimension
randomness chaotic system generates 32 keys, and the Salsa20 algorithm generates 512 bits blocks
CMC, 2023, vol.75, no.1 2147

of keystream. Also, the sub-DNA operation is used, as presented in Table 1c. Add Round is used to
decrypt data to avoid threats. The inverse of each previous process was completed; Fig. 4 presents the
decryption process of the proposed scheme.

Figure 4: The proposed lightweight image decryption scheme

6 Simulation Experiments and Performance Analysis

The experiment was carried out on a personal computer with Intel core i5, 2.40 GHz CPU, 8 GB
memory, windows 10, and Python 3.7.4. The test images (Lena, Baboon, and Pepper) have dimensions
of 128 × 128 with 24-bit color. Fig. 5 shows that some of the cipher images are noise-like images, which
makes it impossible to infer anything valuable from them. A key space analysis, histogram analysis,
correlation coefficient analysis, information entropy analysis, and differential analysis are all included
in the performance evaluation of the suggested design.

Figure 5: Simulation results of the proposed encryption/decryption scheme

2148 CMC, 2023, vol.75, no.1

6.1 Key Space

The key space volume can ensure that the proposed scheme resists the attack. The encryption
method must have a wide key space to withstand an extensive attack. A key space over 2100 enables a
high level of protection [33]. The secret keys used in the proposed algorithm can be summarized as
follows: the five initial values of the 5D chaos system are denoted; iteration time N0 is used as the
secret key in the proposed scheme; the number of keys is 32; the ranges of these keys are key-x0 (−40,
40), key- y0 (−40, 40), key- z0 (1, 81), key- k0 (−250, 250), and key− p0 (1000, 2500). If all the initial
values have the precision of 1015, the key space can calculate as key space = 80 × 1015 × 80 × 1015
× 80 × 1015 × 500 × 1015 ×1500 ≈ 2 237 > 2 128.

6.2 Histogram Analysis

A histogram displays the distribution of an image’s pixel intensity. A safe encryption system must
include an encrypted image with a uniform histogram to withstand statistical attacks [34]. Fig. 6
presents the histograms of the plain image (Lena) and the encrypted image (b, d). In Fig. 6, the values
of the encrypted image have equal distribution (d). As a result, the distribution of the plain image Lena
in Fig. 6a differs significantly.

Figure 6: Results of the proposed scheme. (a) Original a plain image. (b) Histogram of plain image. (c)
The complete encrypted image. (d) Histogram of the encrypted image.

6.3 Correlation between Plain and Ciphered Images

Within this portion, an examination of the association between plain and encrypted images is
carried out. The calculation of the correlation coefficient follows the format of Eq. (12) [34].

(xi − xm )(yi − ym )
r=  
i
 (12)
i
( i (xi − xm )2 ( i (yi − ym )2
Both xm and ym averages plain and ciphered images. Because the correlation coefficients between
the plain and ciphered images are so close to zero, it may be deduced that the suggested encryption
scheme possesses the correlation quality sought after. Table 5 lists the findings of the comparison of
the correlations.
CMC, 2023, vol.75, no.1 2149

Table 5: Correlation comparison of encryption schemes

Image test Ref. [11] Ref. [13] Ref. [17] Ref. [18] Ref. [19] Ref. [21] Proposed
scheme
Lana 0.0082 0.0064 0.6578 0.00199 0.0065 0.011 0.003901
Baboon – – – – 0.0029 0.0087 0.000824
Pepper – – – 0.00220 – – 0.011421

6.4 Information Entropy Analysis

Information entropy (IE) is an important index for measuring randomness [35]. The information
entropy is calculated according to the following formula [36]:
m−1
H(e) = − P(ei ) log2 P(ei ) (13)
i=0

where e is the origin of the information, m is the number of bits that must be present for the symbol
mi and p (ei ) is the probability that the symbol ei Will be present. The entropy of a 24-bit color image
can reach its maximum value when all of the pixels are laid out in an even distribution. This maximum
entropy value means that the information is random. The information entropy must be near eight in a
ciphered image because it decreases the chance of an attacker decrypting it. Because the entropies of the
encrypted images are quite close to the ideal value of eight, it can be deduced that the suggested method
possesses the required information entropy properties. Calculating the information entropies of both
the plain and the cipher images can be accomplished using Eq. (13). Table 6 presents the findings.

Table 6: Comparison of information entropy of encryption schemes

Test Plain Ciphered image Proposed scheme
image image
Ref. [15] Ref. [16] Ref. [18] Ref. [19] Ref. [21] Ref. [22]
Lena 7.4463 7.9923 7.9968 7.9967 7.9962 7.9642 7.9980 7.9982
Baboon 7.4649 7.3583 – 7.9970 7.9375 7.9982 7.9965
Pepper 7.4743 – 7.99744 7.9993 – – 7.9977 7.9978

6.5 Differential Analysis

An effective encryption strategy must ensure that each slight change to the plain image results in a
considerable difference in the encrypted images to protect against a differential attack. The suggested
encryption approach can make two ciphered images completely distinct, even if their plain images
differ by one pixel. Let C1 and C2 be the two encrypting images, and compute the measured value of
the plain image’s sensitivity to a slight modification using Eqs. (14) and (15) [37,38]:
N M D(i, j)
NPCR = × 100% (14)
i=1 j=1 M ×N
N M abs C1 (i, j) − C2 (i, j)
UACI = × 100% (15)
i=1 j=1 255 × M × N
2150 CMC, 2023, vol.75, no.1

where D(i, j) = 0 if C1 (i, j) = C2 (i, j). Otherwise, D(i, j) = 1. NPCR focuses on the total number of pixels,
which fluctuates during a differential attack. Table 8 presents an NPSR comparison of differential
schemes. The Lena, Baboon, and Pepper images were chosen as the test images, and the NPCR and
UACI values were computed. Table 7 presents the NPCR and UACI results. The suggested algorithm
achieves high NPCR and UACI scores, with an NPCR of 0.99710 and a UACI of 33.68. These scores
indicate that the suggested encryption method is resistant to a differential attack.

Table 7: NPCR and UACI scores of lena, baboon, and pepper with only a one-pixel change
Pixel change NPSR UACI
(Position)
Lena Baboon Pepper Lena Baboon Pepper
(1,1) 0.99561 0.99710 0.99689 33.64 33.60 34.60
(256,256) 0.99647 0.99592 0.99497 33.61 33.68 33.62
(180,220) 0.99573 0.99710 0.99624 33.60 33.62 33.59

Table 8: Comparison of the NPSR and UACI scores of encryption algorithms

Test image Ref. [15] Ref. [16] Ref. [18] Ref. [19] Ref. [21] Ref. [22] Ref. [39] Proposed
scheme
Lena NPSR 0.9963 0.9961 0.9961 0.9958 0.9954 0.9962 0.9946 0.99647
Baboon – – – – – 0.9969 – 0.99710
Pepper – – – – 0.9969 – 0.99689
Lena UACI 33.61 33.445 33.46 28.48 26.51 33.62 33.06 33.64
Baboon – – – – – 33.92 33.20 33.68
Pepper – – – – – 33.60 – 33.62

6.6 Mean Square Error and Peak Signal-to-Noise Ratio

MSE can calculate the difference in pixel values between two pictures. MSE values between these
plaintext and encrypting images are most likely sufficient for a reliable encryption system. The peak
signal-to-noise ratio (PSNR) quantifies the difference in peak error between plain and encrypted
pictures. The PSNR of plain and ciphered images should be low due to their large disparity. The MSE
and PSNR values are obtained from Eqs. (16) and (17) [40].
M N
(Pij − Cij )2
MSE =
i=1 j=1
(16)
M ×N
Imax
PSNR = 20 log10 √ (17)
MSE
Both Pij and Cij Are the pixels of plain and ciphered images at the position (i, j), and Imax , the
maximum pixel estimation of the image. Table 9 presents the results of MSE and PSNR for the
ciphered images. The proposed encryption scheme achieves the desired effect.
CMC, 2023, vol.75, no.1 2151

Table 9: The results of MSE and PSNR for ciphered images

Image MSE PSNR
Lena 8929.5358 0.005452
Baboon 9488.7627 0.005019
Pepper 8781.314 0.004727

6.7 Encryption and Decryption Time

The main evaluation criterion is how long the cryptographic strategy takes to encrypt and decrypt
any image. A slight discrepancy between encrypting and decryption time is discovered. As a result,
decrypting the photos takes longer than encrypting them. Table 10 compares our proposed solution
with existing lightweight image encryption methods. As a result, our approach takes less overall time
in seconds (s) to execute than current methods.

Table 10: Comparison of computational complexity in time (s)

Parameter Ref. [14], Lena Ref. [18], Lena Ref. [39], Bird The proposed
128 × 128 128 × 128 128 × 128 scheme, Lena
128 × 128
Encryption time (s) – – – 1.234
Decryption time (s) – – – 1.553
Total time (s) 3.7944 18.20037 6.14 2.787

7 Conclusions
A large amount of communication is in the form of images. A fast, secure, and lightweight
encryption technique is required for this communication. This paper proposes LIES. The proposed
method is a unique, fast, safe, and lightweight encryption/decryption technique that exploits confusion
and diffusion principles. We not only provide an effective S-Box building approach developed using
the Hénon system, but also an image encryption scheme has been discussed that uses a unique 5D
chaotic system, Salsa20, and S-Box.
S-Box modeling and experimental data demonstrate that the proposed S-Boxes have more desired
features. The building approach of the suggested S-Boxes and the proposed image encryption method
are more efficient because LIES has a high peak signal-to-noise ratio and is highly sensitive to the
secret key. The encrypted image provides the attacker with no information. The suggested approach is
resistant to heavy attacks and takes less time to execute than previously discussed methods, making it
an efficient, lightweight image encryption scheme. The method provides lower correlation coefficients
than other methods, another indicator of an efficient image encryption system.
Even though the proposed scheme has useful applications in image transmission, it still requires
profound improvement in implementing the high-intelligence scheme and verifying its feasibility on
devices with the Internet of Things (IoT) enabled.

Funding Statement: The authors received no funding for this study.

2152 CMC, 2023, vol.75, no.1

Conflicts of Interest: The authors declare they have no conflicts of interest to report regarding the
present study.

References
[1] C. Paar and J. Pelzl, Understanding cryptography: A textbook for students and practitioners, Midtown
Manhattan, New York City: Springer Publishing Company, 2010. [Online]. Available: https://link.springer.
com/book/10.1007/978-3-642-04101-3.
[2] A. Javeed, T. Shah and A. Attaullah, “Lightweight secure image encryption scheme based on chaotic
differential equation,” Chinese Journal of Physics, vol. 66, no. 2, pp. 645–659, 2020.
[3] K. R. Qasim and S. S. Qasim, “Encrypt medical image using CSalsa20 stream algorithm,” Medico-Legal
Update, vol. 20, no. 3, pp. 1248–1256, 2020.
[4] G. Alvarez and S. Li, “Some basic cryptographic requirements for chaos-based cryptosystems,” Interna-
tional Journal of Bifurcation and Chaos, vol. 16, no. 8, pp. 2129–2151, 2006.
[5] A. N. Telem, C. Segning, G. Kenne and H. B. Fotsin, “Image encryption using multi-scroll attractor and
chaotic logistic map,” Computers, Materials & Continua, vol. 72, no. 2, pp. 3448–3463, 2022.
[6] S. Paul, P. Dasgupta, P. K. Naskar and A. Chaudhuri, “Secure image encryption scheme based on DNA
and new multi chaotic map,” Journal of Physics, vol. 1447, no. 1, pp. 1–12, 2020.
[7] M. Khan and F. Masood, “A novel chaotic image encryption technique based on multiple discrete
dynamical maps,” Multimedia Tools and Applications, vol. 78, no. 18, pp. 26203–26222, 2019.
[8] J. Ahmad and S. O. Hwang, “A secure image encryption scheme based on chaotic maps and affine
transformation,” Multimedia Tools and Applications, vol. 75, no. 21, pp. 13951–13976, 2015.
[9] A. Jolfaei and A. Mirghadri, “Survey: Image encryption using Salsa20,” IJCSI International Journal of
Computer Science Issues, vol. 7, no. 5, pp. 213, 2010.
[10] Q. Zheng, X. Wang, M. K. Khan, W. Zhang, B. B. Gupta et al., “A lightweight authenticated encryption
scheme based on chaotic SCML for railway cloud service,” IEEE Access, vol. 6, pp. 711– 722, 2017.
[11] X. Zhang, F. Han and Y. Niu, “Chaotic image encryption algorithm based on bit permutation and dynamic
DNA encoding,” Computational Intelligence and Neuroscience, vol. 2017, no. 10, pp. 1–11, 2017.
[12] S. Janakiraman, K. Thenmozhi, J. B. Rayappan and R. Amirtharajan, “Lightweight chaotic image encryp-
tion algorithm for real-time embedded system: Implementation and analysis on 32-bit microcontroller,”
Microprocessors and Microsystems, vol. 56, no. 1, pp. 1–12, 2018.
[13] K. A. Patro, B. Acharya and V. Nath, “A secure multi-stage one-round bit-plane permutation operation
based chaotic image encryption,” Microsystem Technologies, vol. 25, no. 7, pp. 2331–2338, 2019.
[14] C. Lin, “Cryptanalysis and improvement of a chaotic map-based image encryption system using both
plaintext related permutation and diffusion,” Entropy, vol. 22, no. 589, pp. 1–23, 2020.
[15] M. Guan, X. Yang and W. Hu, “Chaotic image encryption algorithm using frequency-domain DNA
encoding,” IET Image Processing, vol. 13, no. 9, pp. 1535–1539, 2019.
[16] J. Dagadu, J. P. Li and P. C. Addo, “An image cryptosystem based on pseudorandomly enhanced chaotic
DNA and random permutation,” Multimedia Tools and Applications, vol. 78, no. 17, pp. 24979–25000,
2019.
[17] J. Zheng, Z. Luo and Z. Tang, “An image encryption algorithm based on multichaotic system and DNA
coding,” Discrete Dynamics in Nature and Society, vol. 2020, no. 10, pp. 1–16, 2020.
[18] H. Liu, B. Zhao, J. Zou, L. Huang and Y. Liu, “A lightweight image encryption algorithm based on message
passing and chaotic map,” Security and Communication Networks, vol. 1, no. 7151836, pp. 1–12, 2020.
[19] M. Gupta, K. K. Gupta and P. K. Shukla, “Session key based fast, secure and lightweight image encryption
algorithm,” Multimedia Tools and Applications, vol. 80, no. 7, pp. 10391–10416, 2021.
[20] D. Ravichandran, A. Banu, S. B. Murthy, V. Balasubramanian, S. Fathima et al., “An efficient medical
image encryption using hybrid DNA computing and chaos in transform domain,” Medical & Biological
Engineering & Computing, vol. 59, no. 3, pp. 589–605, 2021.
CMC, 2023, vol.75, no.1 2153

[21] J. Ferdush, M. Begum and M. S. Uddin, “Chaotic lightweight cryptosystem for image encryption,”
Advances in Multimedia, vol. 2021, no. 5, pp. 16, 2021.
[22] A. A. Abdallah and A. K. Farhan, “A new image encryption algorithm based on multi chaotic system,”
Iraqi Journal of Science, vol. 63, no. 1, pp. 324–337, 2022.
[23] A. Arab, M. J. Rostami and B. Ghavami, “An image encryption method based on chaos system and AES
algorithm,” The Journal of Supercomputing, vol. 75, no. 10, pp. 6663–6682, 2019.
[24] Y. Niu, X. Zhang and F. Han, “Image encryption algorithm based on hyperchaotic maps and nucleotide
sequences database,” Computational Intelligence and Neuroscience, vol. 2017, no. 40, pp. 9, 2017.
[25] M. Li, M. Xu, J. Luo and H. Fan, “Cryptanalysis of an image encryption using 2D Henon-Sine Map and
DNA approach,” IEEE Access, vol. 7, pp. 63336–63345, 2019.
[26] Z. Hua and Y. Zhou, “Image encryption using 2D Logistic-adjusted-Sine map,” Information Sciences, vol.
339, no. 8, pp. 237–253, 2016.
[27] D. P. Schmid and A. Biryukov, “Slid pairs in Salsa20 and Trivium,” Mathematics, Computer Science, vol.
22, no. 2, pp. 1–17, 2014.
[28] R. S. Salman, A. K. Farhan and A. Shakir, “Creation of S-Box based one-dimensional chaotic logistic map:
Colour image encryption approach,” International Journal of Intelligent Engineering and Systems, vol. 15,
no. 5, pp. 378–389, 2022.
[29] A. A. Alzaidi, M. Ahmad, M. N. Doja, E. Al Solami, M. M. Sufyan et al., “A new 1D chaotic map and
β-hill climbing for generating Substitution-Boxes,” IEEE Access, vol. 6, no. 10, pp. 55405–55418, 2018.
[30] Q. Lu, C. Zhu and G. Wang, “A novel S-Box design algorithm based on a new compound chaotic system,”
Entropy, vol. 21, no. 10, pp. 1004, 2019.
[31] G. Hanchinamani, D. G. Narayan and R. Savakknavar, “Construction of S-Box based on parametric
mixing of chaotic maps,” in Proc. ICAECT, Bhilai, India, pp. 1–4, 2021.
[32] A. H. AL-Wattar, “A review of block cipher’s S-Boxes tests criteria,” Iraqi Journal of Statistical Sciences,
vol. 16, no. 29, pp. 91–104, 2019.
[33] S. Mohammad Seyedzadeh and S. Mirzakuchaki, “A fast color image encryption algorithm based on
coupled two-dimensional piecewise chaotic map,” Signal Processing, vol. 92, no. 5, pp. 1202–1215, 2012.
[34] M. A. Khodher, A. Alabaichi and A. S. Abbas, “Dual method cryptography image by two force secure and
steganography secret message in IoT,” Telecommunication, Computing, Electronics and Control, vol. 18, no.
6, pp. 2928–2938, 2020.
[35] M. Ragab and E. B. Ashary, “Metaheuristic lightweight cryptography for security enhancement in Internet
of Things,” Computers, Materials & Continua, vol. 72, no. 2, pp. 3010–3023, 2022.
[36] A. M. Alabaichi, “True color image encryption based on dna sequence, 3D chaotic map, and key-dependent
DNA S-Box of AES,” Journal of Theoretical and Applied Information Technology, vol. 96, no. 2, pp. 304–
321, 2019.
[37] H. Liu and X. Wang, “Colour image encryption based on one-time keys and robust chaotic maps,”
Computers & Mathematics with Applications, vol. 59, no. 10, pp. 3320–3327, 2010.
[38] M. Ahmad, M. N. Doja and M. M. S. Beg, “Security analysis and enhancements of an image cryptosystem
based on hyperchaotic system,” Journal of King Saud University–Computer and Information Sciences, vol.
33, no. 1, pp. 77–85, 2021.
[39] B. Yousif, F. Khalifa, A. Makram, A. Takieldeen, “A novel image encryption/decryption scheme based on
integrating multiple chaotic maps,” AIP Advances, vol. 10, no. 7, pp. 1–9, 2020.
[40] Y. Q. Zhang, J. L. Hao and X. Y. Wang, “An efficient image encryption scheme based on S-Boxes and
fractional-order differe der differential logistic map,” IEEE Access, vol. 8, pp. 54175–54188, 2020.