Content uploaded by Ekhlas K. Gbashi
Author content
All content in this area was uploaded by Ekhlas K. Gbashi on Oct 13, 2022
Content may be subject to copyright.
International Journal of Electrical and Computer Engineering (IJECE)
Vol. 12, No. 5, October 2022, pp. 4988~5000
ISSN: 2088-8708, DOI: 10.11591/ijece.v12i5.pp4988-5000 4988
Journal homepage: http://ijece.iaescore.com
Novel lightweight video encryption method based on ChaCha20
stream cipher and hybrid chaotic map
Abeer Tariq Maolood, Ekhlas Khalaf Gbashi, Eman Shakir Mahmood
Department of Computer Science, University of Technology, Baghdad, Iraq
Article Info
ABSTRACT
Article history:
Received May 26, 2021
Revised Jun 24, 2022
Accepted Jul 4, 2022
In the recent years, an increasing demand for securing visual resource-
constrained devices become a challenging problem due to the characteristics
of these devices. Visual resource-constrained devices are suffered from
limited storage space and lower power for computation such as wireless
sensors, internet protocol (IP) camera and smart cards. Consequently, to
support and preserve the video privacy in video surveillance system,
lightweight security methods are required instead of the existing traditional
encryption methods. In this paper, a new light weight stream cipher method
is presented and investigated for video encryption based on hybrid chaotic
map and ChaCha20 algorithm. Two chaotic maps are employed for keys
generation process in order to achieve permutation and encryption tasks,
respectively. The frames sequences are encrypted-decrypted based on
symmetric scheme with assist of ChaCha20 algorithm. The proposed
lightweight stream cipher method has been tested on several video samples
to confirm suitability and validation in term of encryption–decryption
procedures. The performance evaluation metrics include visual test,
histogram analysis, information entropy, correlation analysis and differential
analysis. From the experimental results, the proposed lightweight encryption
method exhibited a higher security with lower computation time compared
with state-of-the-art encryption methods.
Keywords:
ChaCha20
Chaotic maps
Stream cipher
Video encryption
Video privacy
This is an open access article under the CC BY-SA license.
Corresponding Author:
Abeer Tariq Maolood
Department of Computer Science, University of Technology
52 Road, Al Senaa’h Street, Baghdad, Iraq
Email: abeer.t.maolood@uotechnology.edu.iq
1. INTRODUCTION
Recently, the encryption of multimedia data at the for front of research today, due to the continuous
increment in digital communication on the internet and increasing uses of video in a wide range of applications
security and privacy issues into serious attention. The nominal aim of multimedia data encryption is to make
the multimedia information robust against the unauthorized divulgence in transit and storage. Many presented
encryption schemes which are to preserve text data and image data are not adequate for video encryption
because of the real time restriction and huge data volume in the video. Therefore, a lot of works have been
presented for video encryption to improve the encryption quality and video security conditions based on
utilizing different encryption algorithms. Hence, by comparing the chaos-based encryption with several
encryption schemes, it has demonstrated an excellent performance with confirmed ability of increased
security and privacy based on utilizing variable keys [1]–[8]. These make the chaotic scheme is adequate to
implement video encryption for different applications. On the other hand, the ChaCha20 is stream cipher
algorithm depended on XORed the plaintext with stream of pseudo-random bytes as a key of ChaCha20
Int J Elec & Comp Eng ISSN: 2088-8708
Novel lightweight video encryption method based on ChaCha20 stream … (Abeer Tariq Maolood)
4989
[9]–[12]. ChaCha20 provides high confidence and much sensitive to change of the initial conditions, where a
simple flipping of single bit into the input stream will because unpredictable changes at the output stream.
The problems of the previous schemes for securing the video data can be defined as, where these
schemes were not taken into account the various parameters at simultaneously, such as quality of security,
efficient computational with execution time, and compression efficiency, where many video-encryption
schemes have been proposed to secure the local videos. Therefore, several of these schemes not adequate for
secure the video transmission at real-world applications. From this point of issue, the motivation of this paper
is to propose solution which can meet main requirements for secure transmitting the public and private videos
based on a new encryption scheme by combined of chaos-based encryption models and ChaCha20 symmetric
stream cipher algorithm. In this paper, the proposed scheme reduces the size of video data by extraction the
differences of sequences frames based on dynamic reference frame. The resultant video will be double
encrypted using two encryption phases, chaos-map and ChaCha20. The proposed scheme provides the main
requirements for transmitting videos which are compression by reduces the video data, computational,
quality of security by utilized double security phases, and efficient computations by utilizing two-model
chaotic maps, first model to generate encryption keys to encrypt individual frame and second model to
generate initial key of ChaCha20 algorithm. This paper is organized as follows: section 1 presents an
introduction of this paper, section 2 presents the related work, section 3 presents the objectives and
contributions of proposed video encryption scheme, the material and methods presented in section 4, the
proposed video encryption scheme presented in section 5, the experimental results presented in section 6, the
conclusion presented in section 7 at a final of this paper.
2. RELATED WORK
In this section, different related works on video encryptions are discussed and explained. The
security performance analysis of video encryption schemes is presented and investigated in different
research. In general, the video encryption was divided for selective encryption and full encryption, based on
the amount of encrypted data and the requirements of the security level [13]–[18].
A selective video encryption scheme was presented to encrypt the interested part of the frame which
of greatest information for the user, based on entropy measure of the frame data blocks and utilizing the
chaotic map [19]. The presented, method depends on encrypted video by selecting interesting macro-blocks
to be encrypted based on entropy measure. This method exhibited a strength security against an
entropy-based attack, and it is suitable for video on demand applications. The encrypted video outflow this
scheme has the same frames with distortion of several macro-blocks in each frame. In [20], a selective video
encryption scheme based on coding characteristics was presented based on two case of encryption video.
First case is complete selective video encryption using video encoding and 4D hyper-chaotic systems at the
small amount of encrypted data. Second case is selective video encryption was implemented based on video
characteristic encoding such as the motion vector difference (MVD), residual coefficient, intra prediction
mode (IPM) and delta-QP in each slice of H.264/AVC encoding algorithm. This method presented high
quality of encrypted video with high processing complexity for encrypted video. The encryption scheme was
depended on three phases of work progress such as first phase is constructing plaintext, second phase is
encrypting plaintext, and third phase is replacing the original bitstream. The H.264/AVC encoding algorithm
was adopted to encode the video into independent multiple slices. The advanced encryption standard (AES)
based cipher feedback mode (CFB) was utilized to encrypt these slices individually. The pseudo-random
number generator (PRNG) was used to generate the required keys and real time updated.
On the other hand, a lot of researchers have been presented different schemes to encrypt whole
video frame by adopting different encryption algorithms [21]–[26]. In work [24], to generate the key stream
for video frames encryption, two algorithms were utilized different chaotic. The video file had been
compressed before applied encryption process. The two proposed schemes were sensitive to slight variations
in any components of comprising the secret encryption key as apparent from the obtained differential
measures. In work [27], AES algorithm was adopted to encrypt video frame directly on the compressed
domain with motion picture expert group (MPEG) video compression. In this method, the time cycle of
compression and decompression was preserved.
In [28], the 2 dimension (2D) chaotic model was presented to apply encryption sachem for the
multimedia encryption at high security data transmission. Their work was presented a hybrid chaotic
structure based on multiple combined maps for different media encryption (text, voice, image, and video). At
video encryption scheme of this method, the whole video frames were encrypted as separated image. This
method exhibits high security data transmission with increases computation complexity and consuming in
processing time.
In [29], a three-level chaotic video encryption scheme was presented based on employed the
permutation and diffusion rounds. The maps of combination of logistic and tent (LTS) were utilized to obtain
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 12, No. 5, October 2022: 4988-5000
4990
the prefatory encryption parameters. Their method exhibited good competency for processing time; however,
the drawback was a compression reduction. The key generator block (KGB) was utilized to generate the
internal keys of the key generation technique in order to generate required keys for chaos system. The
internal keys were adopted to present the frame selection (FS) technique. On the bases of the presented
results, the key space of the presented scheme was 2212 which can prove good security level. Also, the
maximum deviation and deviation irregular were achieved good results compared with different related
works. In [30], the S-Box and two alternative schemes were utilized to present video encryption pipeline.
These schemes are higher dimensional chaotic map and Ikeda delay differential equation (DDE). The
associated limitations of this method were increasing the key complexity and processing time for encryption.
In [31], the region of interest (ROI) based on faster region based convolutional neural networks (R-CNN)
method for video encryption was presented to overcome the shortcoming of several video encryption
schemes. Different encryption schemes had been used to encrypt the non-ROI in order to implement fully
video frame encryption. This method presented an improved the performance of video encryption scheme
based on reduces the amount of data in video for encryption at ROI. However, the main limitation of this
method was increasing the encryption time and computation complexity by using two cases of encryptions
(ROI and non-ROI) with different encryption scheme.
On the other hand, the encryption of video traffic had been implemented based on design of
algorithms for identification/prediction attack from encrypted traffic. In [32], the practical mobile traffic had
been designed based on utilized the deep learning as a viable strategy. This classifier was proposed to extract
the features automatically and it has capability to deal with encrypted traffic, then reflecting their complex
patter of traffic, while in [33]–[36], multi classifier have been proposed for mobile application traffic based
on different approaches for instance Markov modeling, capture, and ground-truth creation. Finally, the
general comparisons of the previous video encryption schemes that demonstrated in the literature are
illustrated in Table 1.
Table 1. Comparisons of the previous video encryption schemes
3. OBJECTIVES AND CONTRIBUTIONS OF PROPOSED VIDEO ENCRYPTION SCHEME
Many researchers’ literature in this paper had been interested on the video encryption issue in order
to validate video encryption in real time by utilizing different techniques. Their works had been associated
with different weakness points. It is noteworthy, several video-encryption algorithms had been developed to
secure the local videos, but these encrypted videos are not appropriate for transmitting in real-world
applications. The main contribution of this paper, a new video encryption scheme has been proposed to
shorthand the computation time by reduction the data included in the frame. This contribution is carried out
based on computing the differences between sequences frames, in order to produce video frames, have less
video data than original video with availability to return original video at decryption end. The presented work
in this paper aims to improve the encryption quality, reduces processing time, reduces the execution time,
and increases the robustness against different attacks. In order to achieve these objectives, the proposed work
in this paper is adopted combinations of different algorithms to present light video encryption scheme which
are Henon chaotic map, ChaCha20 algorithm, and algorithm of differences of sequences frame. Finally, to
validate the mentioned aims of this paper, the proposed video encryption scheme will be tested by different
video files. Then, the finding of the experimental results will be described in sufficient details.
4. METHODS AND MATERIALS
The chaotic map and ChaCha20 algorithms are utilized in this paper to propose a new video
encryption scheme for real time applications. In order to implement the proposed video encryption scheme in
this paper, efficient secret keys are needed to obtain a robust and powerful encryption model against variant
attacks. Accordingly, hybrid chaotic maps are adopted and carried out in two phases models to generate the
secrete keys in this work. First chaotic model is employed to generate chaotic sequence-based permutation
Ref
Algorithm
Methods
Time
Encryption Ratio
Size
[31]
Hyper chaos and GF(17)
ROI and Fasted-CNN
0.2307 sec
100%
No change
[30]
Chaos system
Chaos maps and Ikeda
time delay system
0.8062 sec
100%
No change
[29]
1D Chaotic Maps
Chaos system
0.5403 sec
100%
No change
[19]
chaotic map and entropy measure
Chaotic system
------
100%
[16]
4D and 3D Arnold’s cat map and Chebyshev map
Arnold cat map
--------
100/%
No change
Int J Elec & Comp Eng ISSN: 2088-8708
Novel lightweight video encryption method based on ChaCha20 stream … (Abeer Tariq Maolood)
4991
process of light-weight encryption scheme. The second chaotic model is exploited to generate the initial keys
of Chach20 method as will be explained in algorithm 1.
Algorithm 1: keys generation -based encryption
Input: a, b, Xo, Yo, Mid, M
Output: Key2[M] // list of generated keys
Define basic model parameters: a and b
Initialize model parameters (Xo and Yo)
Start Iteration Mid
Update XMid+1 and YMid+1
End iteration Mid
Start iteration M
X,Y ← update XM+1 and YM+1 ///Generate encryption keys using (2.a) and (2.b)
Key[M]=select1(X, Y, 32)
Key1[M]=select2(Key(M),16)
Key2[M]=integrate(Key, Key1)
End iteration M
4.1. Hybrid chaos models-based keys generation
Henon chaotic map presented in [37] is adopted in the first model (model 1) to generate
two-dimension chaotic map Xi, Yi where (i=1, 2, ..., M) and used to perform the permutation process in the
lightweight encryption phase.
Xi+1=1+b1Yi-a1Xi2 (1a)
Yi+1=Xi (1b)
Where a1, b1 are the initial parameters of this model with initial values a1 [1.07, 1.4] and b1=0.3. Xo1, Yo1
initial values are stand to Xo1=0.5, Yo1=0.5. It is worth noting that the first model of Henon chaotic maps is
iterated by (H×W)+Mid, where H, W refereed to high and width of the current frame, respectively. Mid is a
predefined iteration number. The second model is iterated by M times. The second model (model 2) of
Henon chaotic map presented in [6] is adopted to generate the important initial key of chach20 encryption
method. The initial values of this model are, a2=1.4, b2=0.3, Xo2=1.2, Yo2=0.8 as illustrated in (2):
Xi+1=1-a2Xi2+Yi (2a)
Yi+1=b2Xi (2b)
where both functions select1 () and select2 () are implemented for selecting 32, 16 bytes from the original
chaos keys X, Y, respectively. Then we integrate the obtained keys of each iteration to construct the final key
Key2 used later for chach20 encryption method.
4.2. ChaCha20 encryption algorithm
ChaCha20 algorithm is stream cipher has naturally good performance, as well as itis fast, secure,
and simple encryption algorithm. ChaCha20 implemented in this paper is derived from the code and
algorithms available in different research works and reports [9]–[12]. Its input includes a 48 byte consists of
32-byte key, a 4-byte counter, a 12-byte nonce. The core is a pseudo-random number generator. The output
cipher data is obtained by XOR'ing the input plain data with a pseudo-random stream:
, at decryption, the same operation as
encryption: .
The ChaCha20 ciphering process is depended on 20 rounds of mathematical calculations using
XOR, addition and rotation using as inputs four 32-byte key, a 4-byte counter, a 12-byte nonce as input
matrix as shown in Table 2. The number of rounds depends on the application requirements such as 20 round
(maximum security), 8 rounds (maximum speed) or 12 rounds (balance between speed and security) [9]. In
our work, the 20 round of ChaCha20 algorithm is implemented to provide high data security for the
transmitted video.
Table 2. Input matrix
Const.
Const
Const
Const
Key
Key
Key
Key
Key
Key
Key
Key
Counter
Nonce
Nonce
Nonce
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 12, No. 5, October 2022: 4988-5000
4992
5. THE PROPOSED VIDEO ENCRYPTION SCHEME
This section describes the proposed encryption/decryption schemes for the acquired plain video
sequence Pv={F1, F2, ……Fn}. The main steps of the proposed encryption algorithm are schematized in
Figure 1 and illustrated in algorithm 2. Likewise, the main steps of the proposed decryption algorithm are
demonstrated in Figure 2 and stated in algorithm 3. The decryption phase is designed and implemented to
reconstruct the original video from the encrypted sequenced frames.
Algorithm 2: the proposed video encryption scheme
Input: Plain video Pv, Key2, a1, b1[1.07, 1.4], Xo1= 0.5, Yo1=0.5, a2 = 1.4, b2 = 0.3,
Xo2= 1.2, Yo2 =0.8, step
Output: Cipher video Cv
Begin
X1,Y1 ← Key Generation-based Permutation (a1, b1, Xo1, Yo1) // using model
Ke2, X2,Y2 ← Key Generation-based Encryption (a2, b2, Xo2, Yo2) // using model 2
Len=size(Pv)
Ref=Pv(1) // taken reference frame from Pv
Set i=1
while i+1<Len-step
Read (Pv, Fi) // read frames from Pv
DFi=Pre-processing (Fi, Ref) // find dissimilar frames using Algorithm (2)
LDFi=Lightweight Encryption (DFi, X1, Y1) // using Algorithm (4)
EPvi=Chacha-Encryption(LDFi, Key2)
Cv[i]=EPvi //Construct array of encrypted frames
if i% step==0 Ref=Fi end if
End while
End
Figure 1. Diagram of the proposed video encryption
algorithm
Figure 2. Diagram of the proposed video decryption
algorithm
Algorithm 3: video decryption scheme
Input: Cipher video Cv, Key2, X, Y, step
Output: Palin video Pv
Begin
Int J Elec & Comp Eng ISSN: 2088-8708
Novel lightweight video encryption method based on ChaCha20 stream … (Abeer Tariq Maolood)
4993
Len=size(Cv)
Ref=Cv(1)
While i+1<Len
Read (Cv, Fi) // read one frames from Pv
DCvi=Chacha-Decryption (Fi, Key2) // decryption phase
LDFi=Inverse-Lightweight (DCvi, X, Y) // inverse-lightweight phase according to model1
DFi=Pre-processing (LDFi, Ref)
Pv[i]=DFi //Construct encrypted video from encrypted frames
if i>=step then
Ref=Fi
end if
End while
End
5.1. Pre-processing of video using frames differences
In this section, the frame differences technique will be explained. This technique is applied at both
encryption and decryption sides in order to reduce the processing time. The main objective of utilizing
frames differences technique is to estimate the changes between sequenced frames (reference frame and
current frame) along video sample based on their dissimilarity measurement. In this context, the reference
frame is updated at a predefined interval of frames named step. Afterwards, the output video obtained from
pre-processing phase is passed into lightweight encryption phase with lower spatial information of each video
frame in order to minimize the computation time of the encryption/decryption processes. The main steps of
pre-processing approach are illustrated in Figure 3. The main contributions of adopting frame differences as
pre-processing phase in our video encryption scheme can be summarized as follows: i) minimize the spatial
information of video frame, ii) reduces the processing time and complexity of preprocessed video in
encryption and decryption sides, and iii) lower redundancy of video frames are acquired through adopting
step size for reference frame denoting.
Figure 3. Diagram of pre-processing approach
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 12, No. 5, October 2022: 4988-5000
4994
5.2. Lightweight encryption scheme
A lightweight encryption scheme is designed and implemented to encrypt dissimilar video frames
obtained from pre-processing video phase. The core of this algorithm is carrying out the data permutation for
individual frame based on the chaotic map model (1). The aim of lightweight encryption algorithm is to make
the encryption complexity is strong resistance against most measurable tests. The lightweight encryption
algorithm is illustrated in algorithm 4.
Algorithm 4: lightweight encryption
Input: Dissimilar video frame DFi, X1, Y1
Output: Cipher video frame LDFi
Begin
Examine the frame dimensions (DFi) and reshape into one dimensional vector of pixels
(DFvi=H×W×3) // H, W are height and width of DFi video frame)
Permute DFvi data frame (pixels) using chaotic sequence X1, Y1 // permutation phase
according to model1
LDFi=reshape (DFvi, H, W)
End
6. RESULTS AND DESCUSSION
The experimental results of many video files are tested to validate and investigate the objectives are
mentioned in this work. For this purpose, the proposed encryption scheme and the key generation algorithm
have been investigated on the platform (MATLAB 2018, windows 8.1-64-bit, CPU Cor i7, RAM 6 GB,
2.2 GHz) and tested with different video files. The strength and suitability of proposed scheme for video
encryption have been tested for key sensitivity, key space, computational time, randomness, and resistance
against statistical attacks.
The experimental results are demonstrated for different video files listed from any video dataset. For
the simulation experiments, the selected video files have properties are enlisted in Table 3. The important key
metrics are measured to validate the objectives of the proposed video encryption scheme. The following
progresses are implemented to establish the validation of the proposed scheme.
Table 3. Properties of video files used in simulation results
Attribute
Value
Number of frames
821
Height
640 pixels
Width
360 pixels
Frame rate
25-FPS
Bit per pixel
24-BPP
Video format
RGB24
6.1. Visual degradation measurements
To measure the perceptual distortion of the encrypted video frame compared with original video
frame, the visual degradation must be measured. The simulation results for selected video frames of different
videos are illustrated in Figure 4 (Figures 4(a) and 4(b)), where these videos have properties listed in Table 3.
From this figure, it can be inferred that, the proposed encryption scheme offers high scale of visual
degradation. Therefore, the proposed encryption/decryption scheme as shown in Figures 4(c) and 4(d)
respectively is very suitable to encrypt high class data-based multimedia applications.
6.2. Key space analysis
In crypto analysis system, the Brute force attack depends on trying out all possible keys to recover
the original video data. The size of key space for the encryption scheme provides its strength to make Brute-
force attack unfeasible. In this work two dimensional Henon chaotic systems were employed to provide a
larger key space. The 2D chaotic systems are utilized to generate the secret key for two encryption phases of
the proposed video encryption scheme. The encryption values of ChaCha20 depend on its initial seed keys
(key, nonce, counter and constants) and its related parameters associated with the second model of Henon
map. Likewise, the initial encryption values of light weight scheme depend upon the first model of Henon
map equation and its initial values. In addition, the key space depends upon the number of decimal places of
the mantissa which is approximately 215×215=230 for each of model1 and model2 which yields 260 key spaces
based chaotic map generation. This making the Brute force attack is computationally infeasible. This is a
main contribution of using two Henon chaotic maps to realize a larger key space. Further, the key space
regards the encryption phase based on ChaCha20 is denoted as 2348.
Int J Elec & Comp Eng ISSN: 2088-8708
Novel lightweight video encryption method based on ChaCha20 stream … (Abeer Tariq Maolood)
4995
(a)
(b)
(c)
(d)
Figure 4. Simulation results of proposed encryption scheme for selected frame taken from different videos:
(a) original frame, (b) differences frame, (c) encrypted frame, and (d) decrypted frame
6.3. Correlation coefficients analysis
The correlation coefficient reflects the pixels distribution and their relationships (dependency or
independency) in the adjacent video frames. This mean, the relationship between adjacent frames indicates
correlation value. The correlation is high when the correlation value is (approach to 1) at the linear
relationship, while correlation is implying low when the correlation value (approach to zero) at the nonlinear
relationship [8]. Therefore, the correlation value between adjacent frames of encrypted video must be
minimized. The correlation between adjacent frames is measured for the pixels located at the same position
of adjacent frames. In this paper, the differences between adjacent frames are adopted to provide a new video
frame with minimum data size. This technique makes the correlation value (approach to zero) and the
relationship between adjacent frames is nonlinear.
The experimental results of the correlation coefficients related to adjacent frames for the original
and encrypted frames are illustrated in Table 4. From these results, it is inferred that there exists negligible
correlation between original and encrypted frames, thereby providing no clue to make the statistical
cryptanalysis is possible. A scientific comparison was conducted between the proposed encryption scheme
and published research work presents in literature to evaluate the performance of the proposed encryption
scheme. This comparison has been implemented based on the properties of video file “dfs.avi”, which is
utilized in [14] and illustrated in Table 5. The comparison simulation results are illustrated in Table 6.
Table 4. Simulation results of correlation coefficient
Correlated
Frame No.
Correlation coefficient of red
component
Correlation coefficient of green
component
Correlation coefficient of blue
component
Original
Differences
Encrypted
Original
Differences
Encrypted
Original
Differences
Encrypted
1
1
1
1
1
1
1
1
1
1
1
1
2
0.9951
0.1077
-0.00027
0.9627
0.1065
-0.00053
0.9959
0.1106
-0.00074
2
3
0.9950
0.4121
-0.00049
0.9643
0.3800
-0.00310
0.9959
0.4276
-0.00043
3
4
0.9939
0.0219
0.00013
0.9670
0.0066
0.00051
0.9949
0.0089
0.00064
4
5
0.9932
0.0788
0.00025
0.9645
0.0570
0.00033
0.9942
0.0647
0.00073
5
6
0.9328
0.2263
0.00011
0.9610
0.2450
0.00201
0.9938
0.2348
0.00047
Table 5. Properties of video files “dfs.avi” [14]
Attribute
Value
Number of frames
102
Height
240 pixels
Width
320 pixels
Frame rate
15-FPS
Bit per pixel
24-BPP
Video format
RGB24
6.4. Histogram analysis
The histogram analysis reveals the graphical relationships between the original and encrypted frame
based on the frame pixels distribution. Statistical attacks are made by exploiting the predictable this
relationship to recover the original frames. The considerable changes in the histograms of encrypted frames
compared with the original frames reflect the strength of the encryption scheme against statistical attacks.
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 12, No. 5, October 2022: 4988-5000
4996
The histograms of the original frames are shown in Figure 5 (Figures 5(a) to 5(c)) and encrypted frames are
illustrated in Figures 5(d) to 5(f), for the red, green, and blue components. Obviously, the histograms of the
encrypted frame for three components red, green, blue are uniformly distributed and significantly different
from the original frames. Subsequently, they do not provide any evidence for exploiting the statistical attack.
Also, the differences frames are shown in Figures 5(g) to 5(i) and encrypted differences frames are illustrated
in Figures 5(j) to 5(l), for the red, green, and blue components.
Table 6. Comparison simulation results of correlation coefficient
Frame No.
Frame No.
Correlation coefficient of red
component
Correlation coefficient of green
component
Correlation coefficient of blue
component
Ref [14]
Proposed method
Ref [14]
Proposed method
Ref [14]
Proposed method
1
2
− 0.0065
-0.0082
− 0.0066
−0.0072
−0.0020
-0.0062
2
2
1
1
1
1
1
1
2
3
− 0.00740
-0.00930
− 0.0067
− 0.0085
0.0089
0.00036
3
4
0.00022
0.00012
0.0037
− 0.00042
0.0087
0.00042
4
5
0.0014
-0.00250
0.0036
0.00031
0.00087
0.00023
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
(k)
(l)
Figure 5. Histogram components (red, green, blue) of original frame, differences frame and encrypted frame
for (original and differences frame): (a) original frame (red), (b) original frame (green), (c) original frame
(blue), (d) encrypted original frame (red), (e) encrypted original frame (green), (f) encrypted original frame
(blue), (g) differences frame (red), (h) differences frame (green), (i) differences frame (blue), (j) encrypted
differences frame (red), (k) encrypted differences frame (green), and (l) encrypted differences frame (blue)
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Red value
Red value
Red value
Red value
Green value
Green value
Green value
Green value
Blue value
Blue value
Blue value
Blue value
Int J Elec & Comp Eng ISSN: 2088-8708
Novel lightweight video encryption method based on ChaCha20 stream … (Abeer Tariq Maolood)
4997
6.5. Randomness test
The randomness analysis is required in video cryptographic by measuring the entropy of each frame
at both original and encrypted video frames. From a security viewpoint, the entropy for the encrypted frame
should be close to an ideal value. Significantly, the higher values of entropy indicate results in high
randomness (statistically), and high distortion results in the encrypted frames.
(3)
Where pi [0.0, 1.0] and −log pi represents the information associated with a single occurrence of pi. Using
(3) [19], entropy is found for selected frame taken from original, encrypted, and decrypted frames of the
video. The simulation results and comparison are illustrated in Table 7. From the simulation results, the
entropy values obtained are close to an ideal value of eight. Therefore, the proposed scheme provides an
encrypted video has strength against an entropy-based attack due to lower effectiveness of this type of
attackers when the data outflow is minimized.
Table 7. Simulation results and comparison of entropy (randomness)
Proposed Scheme
Ref [15]
The entropy of the original frame
7.9996
7.4541
The entropy of the encrypted frame
7.9989
7.4535
The entropy of the decrypted frame
7.9987
7.4537
6.6. Differential attacks analysis
Differential attack could be called the choice of plaintext attack, where it is utilized to measure the
influence of modification one pixel with respect to whole video frame. For this purpose, two measures
(NPCR and UACI) can be adopted to evaluate and measure the differential attack. These two measures can
be computed using mathematical formulas explained in (4) and (5) [19]:
(4)
(5)
where F1(i, j) and F2(i, j) are the pixel intensity of two encrypted video frames F1 and F2 related to the
original video frames. The symbol (H) signifies the row count and (W) signifies the column count.
The D(i, j)=0 if encrypted F1(i, j)=original F1(i, j), otherwise it is equal to 1. Table 8 demonstrates
that the results of differential attack analysis for randomly chosen video frames. These results show that the
encrypted frame is 99.7% than the original video frames. Table 8 demonstrate that the results of differential
attack analysis for randomly chosen video frames. These results show that the encrypted frame is 99.7% than
the original video frames. Another type of comparison related to NPCR and UACI metrics was performed
which demonstrates the percentage of NPCR and UACI for different video samples as shown in Table 9.
Table 8. Simulation results of differential attacks
Sample frame
number
Video frame
chosen
NPCR %
UACI %
1
1
99.3
33.71
2
10
99.7
33.68
3
15
99.56
33.69
4
20
99.7
33.73
Table 9. Comparison results of NPCR and UACI metric
Test
video
Ref [28]
Ref [29]
Proposed
NPCR
UACI
NPCR
UACI
NPCR
UACI
Rhino
99.61
33.61
99.51
33.54
99.73
33.68
Flamingo
99.63
33.63
99.52
33.52
99.68
33.77
Train
99.63
33.63
99.61
33.50
99.71
33.78
VIP-train
99.61
33.60
99.58
33.62
99.75
33.67
6.7. Encryption speed time and encryption ratio analysis
In real time security applications, the encryption speed is a critical key to implement video
encryption scheme. Therefore, to increase the encryption speed of the encryption scheme, encryption ratio
should be minimized. Encryption ratio denoted as the ratio between the size of the encrypted frame and the
original frame size. Achieving the equilibrium between the strength of encryption and encryption speed by
completely encrypting the entire video frame and increases the encryption speed as well as adopting the differences
of video frames technique which proposed in this work as illustrated in pre-processing phase of this work.
The proposed scheme significantly reduces the time of encryption and decryption the whole video
by eliminating the similar information between the sequences frames making it suitable for real time video
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 12, No. 5, October 2022: 4988-5000
4998
encryption. In Tables 10 and 11, we have stated a comparison result related to consuming time of the
proposed scheme against literature research works. Obviously, the obtained results demonstrate that the
proposed video encryption scheme has less time consuming and faster compared to video encryption schemes
listed in in literature. It is owing to; our encryption scheme is based essentially on encryption the data
differences between sequences frames in video instead of encryption the whole video frames.
Table 10. A comparison results of encryption speed time (in frame/seconds) for randomly selected frames
Test video
Ref [30] 8D
Ref [30] 12D
Ref [30] IKeda
Ref [29]
Proposed scheme
Rhino
0.9124
1.7964
1.2147
0.5403
0.2311
Flamingo
0.8062
0.8402
1.1124
0.5982
0.2503
Train
0.8631
1.6493
0.9908
0.5178
0.1926
VIP-train
1.0588
2.3137
1.3574
0.4723
0.16821
Table 11. A comparison results of encryption time (second) per frame
Test video
Ref [29] (Time (second))
Proposed (Time (second))
Grandma
0.2927
0.0341
Akiyo
0.2994
0.0344
Foreman
0.3847
0.0427
6.8. Computational complexity of the proposed lightweight encryption scheme
This section discusses the computational complexity of the proposed lightweight encryption scheme.
Encryption and decryption of whole video frame is a one of the big problems to applied real time
cryptographic scheme at different video types. The proposed scheme in this paper contribute to mitigate this
problem by reduces the computational complexity of encryption scheme based on reduces the number of total
data points for each video frame through adopting the frame differences technique. This technique led to make
computational complexity of the cryptographic scheme is a light weight. Finally, the proposed encryption
scheme present reduces the computational complexity of encryption and improves the encryption speed.
7. CONCLUSION
In this paper, we have proposed, presented, and investigated a new light weight stream cipher
scheme of video frames based on hybrid chaotic system and ChaCha20 algorithm. A lightweight encryption
scheme-based permutation process is employed to scramble video frames contents using chaotic system-
model 1. We have used chaotic system for permutation task due to its robustness, randomness, and it is faster
than traditional scheme for stream ciphers. Thus, the chaos system was utilized to generate encryption keys at
the lightweight encryption phase and to generate the seed keys of ChaCha20 encryption algorithm. In
addition, to provide a larger key space, 2D chaotic systems have been employed to generate the secret keys
for two encryption phases of the proposed video encryption scheme. The permutation process associated with
the proposed encryption scheme is contributed to scramble the frame pixels based on chaos key streams to
provide the randomness distribution of the pixels’ positions. Our experimental results of UACI and NPCR
metrics have been shown that any small changes in the original video frame can be spread overall pixels in
the cipher video frame. The ChaCha20 stream cipher algorithm is implemented with 20 rounds (maximum
secure) in order to produces high quality, lightest and significant performance improvements of encryption
scheme. The ChaCha20 combined with chaos maps are utilized to provide robust video encryption scheme to send
public and private videos on unsecure communication network.
Further, we have presented a new technique for the purpose of encryption video contents through
exploiting the difference criteria between each sequences frames in order to minimize the computation time,
getting minimal correlation coefficient as well as obtaining image histogram with uniformly distribution. As
we shown in the experiments, the proposed method is more robust against the statistical attacks. Further, the
results of histograms and correlation coefficients proved that the proposed encryption scheme is resistance
against statistical attacks. Finally, the advantages of the proposed scheme have been highlighted by
comparing it against different state-of-the-arts methods from literature. Based on the acquired comparative
Int J Elec & Comp Eng ISSN: 2088-8708
Novel lightweight video encryption method based on ChaCha20 stream … (Abeer Tariq Maolood)
4999
performance results, the proposed scheme is obviously has extra efficacious in term of entropy, correlation
coefficients, NPCR and UACI metrics and encryption time(s). The suggested future works are implementing
the proposed encryption and decryption scheme using FPGA technique. Also, apply the proposed scheme in
this paper on the high quality 4k format for both image and video frames. Finally, the region of interest
extraction (ROI) will be utilized to implements the proposed scheme in order to reduce the important data
points of video frames. This leads to reduce the computation time and the computational complexity of video
encryption scheme, as well as making the encrypted video is very adequate for real time applications.
ACKNOWLEDGEMENTS
This article is supported by University of Technology/Computer Science Department, Baghdad, Iraq.
REFERENCES
[1] M. S. Azzaz, C. Tanougast, S. Sadoudi, and A. Bouridane, “Synchronized hybrid chaotic generators: Application to real-time
wireless speech encryption,” Communications in Nonlinear Science and Numerical Simulation, vol. 18, no. 8, pp. 2035–2047,
Aug. 2013, doi: 10.1016/j.cnsns.2012.12.018.
[2] A. Hasheminejad and M. J. Rostami, “A novel bit level multiphase algorithm for image encryption based on PWLCM chaotic
map,” Optik, vol. 184, pp. 205–213, May 2019, doi: 10.1016/j.ijleo.2019.03.065.
[3] J. Yu, S. Guo, X. Song, Y. Xie, and E. Wang, “Image parallel encryption technology based on sequence generator and chaotic
measurement matrix,” Entropy, vol. 22, no. 1, pp. 76–81, Jan. 2020, doi: 10.3390/e22010076.
[4] B. Yousif, F. Khalifa, A. Makram, and A. Takieldeen, “A novel image encryption/decryption scheme based on integrating
multiple chaotic maps,” AIP Advances, vol. 10, no. 7, Jul. 2020, doi: 10.1063/5.0009225.
[5] M. A. Murillo-Escobar, M. O. Meranza-Castillón, R. M. López-Gutiérrez, and C. Cruz-Hernández, “A chaoticencryption
algorithm for image privacy based on two pseudorandomly enhanced logistic maps,” in Multimedia Security Using Chaotic Maps:
Principles and Methodologies, Springer:, Germany, 2020, pp. 111–136.
[6] M. Suneel, “Cryptographic pseudo-random sequences from the chaotic henon map,” Indian Academy of Sciences, vol. 34, no. 5,
pp. 689–701, Apr. 2006.
[7] M. García-Martínez and E. Campos-Cantón, “Pseudo-random bit generator based on multi-modal maps,” Nonlinear Dynamics,
vol. 82, no. 4, pp. 2119–2131, Dec. 2015, doi: 10.1007/s11071-015-2303-y.
[8] F. A. Elizalde-Canales, I. D. J. Rivas-Cambero, L. F. Rebolledo-Herrera, and C. J. Camacho-Bello, “Pseudo-random bit generator
using chaotic seed for cryptographic algorithm in data protection of electric power consumption,” International Journal of
Electrical and Computer Engineering (IJECE), vol. 9, no. 2, pp. 1399–1409, Apr. 2019, doi: 10.11591/ijece.v9i2.pp1399-1409.
[9] M. Goll and S. Gueron, “Vectorization on ChaCha stream cipher,” in 2014 11th International Conference on Information
Technology: New Generations, Apr. 2014, pp. 612–615, doi: 10.1109/ITNG.2014.33.
[10] L. E. Kane, J. J. Chen, R. Thomas, V. Liu, and M. Mckague, “Security and performance in IoT: A balancing Act,” IEEE Access,
vol. 8, pp. 121969–121986, 2020, doi: 10.1109/ACCESS.2020.3007536.
[11] P. McLaren, W. J. Buchanan, G. Russell, and Z. Tan, “Deriving ChaCha20 key streams from targeted memory analysis,” Journal
of Information Security and Applications, vol. 48, Oct. 2019, doi: 10.1016/j.jisa.2019.102372.
[12] R. Velea, F. Gurzau, L. Margarit, I. Bica, and V.-V. Patriciu, “Performance of parallel ChaCha20 stream cipher,” in 2016 IEEE
11th International Symposium on Applied Computational Intelligence and Informatics (SACI), May 2016, pp. 391–396, doi:
10.1109/SACI.2016.7507408.
[13] H. Schwarz, D. Marpe, and T. Wiegand, “Overview of the scalable video coding extension of the H.264/AVC standard,” IEEE
Transactions on Circuits and Systems for Video Technology, vol. 17, no. 9, pp. 1103–1120, Sep. 2007, doi:
10.1109/TCSVT.2007.905532.
[14] L. S. Lian, Multimedia content encryption: techniques and applications. New York, NY, USA: Auerbach Publications, 2008.
[15] F. Sbiaa, S. Kotel, M. Zeghid, R. Tourki, M. Machhout, and A. Baganne, “A selective encryption scheme with multiple security
levels for the H.264/AVC video coding standard,” in 2016 IEEE International Conference on Computer and Information
Technology (CIT), Dec. 2016, pp. 391–398, doi: 10.1109/CIT.2016.53.
[16] J. S. Bowade, P. Khade, Raghuwansh, and D. M. M., “Technique of video encryption/scrambling using chaotic functions and
analysis,” Journal of Emerging Technologies and Innovative Research (JETIR), vol. 2, no. 6, pp. 1951–1958.
[17] S. Lian, J. Sun, J. Wang, and Z. Wang, “A chaotic stream cipher and the usage in video protection,” Chaos, Solitons & Fractals,
vol. 34, no. 3, pp. 851–859, Nov. 2007, doi: 10.1016/j.chaos.2006.03.120.
[18] B. Jovanović and S. Gajin, “An efficient mechanism of cryptographic synchronization within selectively encrypted H.265/HEVC
video stream,” Multimedia Tools and Applications, vol. 77, no. 2, pp. 1537–1553, Jan. 2018, doi: 10.1007/s11042-017-4389-3.
[19] R. Malladar and R. Sanjeev Kunte, “Selective video encryption based onentropy measure,” in Integrated Intelligent Computing,
Communication and Security, 2019, pp. 603–612.
[20] S. Cheng, L. Wang, N. Ao, and Q. Han, “A selective video encryption scheme based on coding characteristics,” Symmetry,
vol. 12, no. 3, pp. 332–348, Feb. 2020, doi: 10.3390/sym12030332.
[21] Z. Lin, S. Yu, J. Lu, S. Cai, and G. Chen, “Design and ARM-embedded implementation of a chaotic map-based real-time secure
video communication system,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 25, no. 7, pp. 1203–1216,
Jul. 2015, doi: 10.1109/TCSVT.2014.2369711.
[22] A. Alfalou, C. Brosseau, and N. Abdallah, “Simultaneous compression and encryption of color video imag es,” Optics
Communications, vol. 338, pp. 371–379, Mar. 2015, doi: 10.1016/j.optcom.2014.10.020.
[23] J. Chen and F. Peng, “A perceptual encryption scheme for HEVC video with lossless compression,” International Journal of
Digital Crime and Forensics, vol. 10, no. 1, pp. 67–78, Jan. 2018, doi: 10.4018/IJDCF.2018010106.
[24] H. Elkamchouchi, W. M. Salama, and Y. Abouelseoud, “New video encryption schemes based on chaotic maps,” IET Image
Processing, vol. 14, no. 2, pp. 397–406, Feb. 2020, doi: 10.1049/iet-ipr.2018.5250.
[25] M. Eid, E.-S. M. El-kenawy, and A. Ibrahim, “A fast real-time video encryption/decryption technique based on hybrid chaotic
maps,” Journal of Computer Science and Information Systems, vol. 2, no. 2, pp. 1–8.
[26] F. Yan, A. M. Iliyasu, S. E. Venegas-Andraca, and H. Yang, “Video encryption and decryption on quantum computers,”
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 12, No. 5, October 2022: 4988-5000
5000
International Journal of Theoretical Physics, vol. 54, no. 8, pp. 2893–2904, Aug. 2015, doi: 10.1007/s10773-015-2524-3.
[27] S. Keshav and A. B. Deshmukh, “Video frame encryption algorithm using AES,” International Journal of Engineering Research
& Technology (IJERT), vol. 5, no. 6, pp. 588–591, 2016.
[28] I. Yasser, M. A. Mohamed, A. S. Samra, and F. Khalifa, “A chaotic-based encryption/decryption framework for secure
multimedia communications,” Entropy, vol. 22, no. 11, pp. 1253–1276, Nov. 2020, doi: 10.3390/e22111253.
[29] R. Ranjith kumar, D. Ganeshkumar, A. Suresh, and K. Manigandan, “A new one round video encryption scheme based on 1D
chaotic maps,” in 2019 5th International Conference on Advanced Computing & Communication Systems (ICACCS), Mar. 2019,
pp. 439–444, doi: 10.1109/ICACCS.2019.8728443.
[30] D. Valli and K. Ganesan, “Chaos based video encryption using maps and Ikeda time delay system,” The European Physical
Journal Plus, vol. 132, no. 12, p. 542, Dec. 2017, doi: 10.1140/epjp/i2017-11819-7.
[31] L. Duan, D. Zhang, F. Xu, and G. Cui, “A novel video encryption method based on faster R -CNN,” in Proceedings of the 2018
International Conference on Computer Science, Electronics and Communication Engineering (CSECE 2018), 2018, pp. 100–104,
doi: 10.2991/csece-18.2018.21.
[32] G. Aceto, D. Ciuonzo, A. Montieri, and A. Pescape, “Mobile encrypted traffic classification using deep learning: Experimental
evaluation, lessons learned, and challenges,” IEEE Transactions on Network and Service Management, vol. 16, no. 2,
pp. 445–458, Jun. 2019, doi: 10.1109/TNSM.2019.2899085.
[33] G. Aceto, D. Ciuonzo, A. Montieri, and A. Pescapé, “Multi-classification approaches for classifying mobile app traffic,” Journal
of Network and Computer Applications, vol. 103, pp. 131–145, Feb. 2018, doi: 10.1016/j.jnca.2017.11.007.
[34] G. Aceto, G. Bovenzi, D. Ciuonzo, A. Montieri, V. Persico, and A. Pescape, “Characterization and prediction of mobile-app
traffic using markov modeling,” IEEE Transactions on Network and Service Management, vol. 18, no. 1, pp. 907–925, Mar.
2021, doi: 10.1109/TNSM.2021.3051381.
[35] G. Aceto, D. Ciuonzo, A. Montieri, V. Persico, and A. Pescape, “MIRAGE: Mobile-app traffic capture and ground-truth
creation,” in 2019 4th International Conference on Computing, Communications and Security (ICCCS), Oct. 2019, pp. 1–8, doi:
10.1109/CCCS.2019.8888137.
[36] S. A. Mahmood, K. A. Hussein, Y. N. Jurn, and E. A. Albahrani, “Parallelizable cipher of color image based on two-dimensional
chaotic system,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 18, no. 1, pp. 101–111, Apr. 2020,
doi: 10.11591/ijeecs.v18.i1.pp101-111.
[37] Y. Liu, J. Zhang, D. Han, P. Wu, Y. Sun, and Y. S. Moon, “A multidimensional chaotic image encryption algorithm based on the
region of interest,” Multimedia Tools and Applications, vol. 79, no. 25–26, pp. 17669–17705, Jul. 2020, doi: 10.1007/s11042-020-
08645-8.
BIOGRAPHIES OF AUTHORS
Abeer Tariq Maolood received the M.Sc. and Ph.D. in Computer Science
from University of Technology, Iraq, 2005 and 2010, respectively. She has around 15 years
of teaching experience. Her areas of interests are computer and network security,
neural networks, and web applications security. She can be contacted at email:
abeer.t.maolood@uotechnology.edu.iq
Ekhlas Khalaf Gbashi earned her Ph.D. in networks security from the
Department of computer sciences at the Technology University. Ekhlas earned her bachelor’s
and master’s degree in computer sciences from the University of Technology (UOT),
Baghdad, Iraq in 1998, 2005. Ekhlas is a faculty member in the computer sciences Department
at the University of Technology (UOT) since 2000; where she became a Head of computer
security Branch at the UOT from 2016 until 2020. Her research interested focus on in
networks security (intrusion detection system), data security, computer networks, comparative
education and computer architecture, image processing and artificial intelligence (AI). She can
be contacted at email: Ekhlas.K.Gbashi@uotechnology.edu.iq
Eman Shakir Mahmood received the MSc. in Computer Science from
University of Technology, Iraq, 2006. She has around 14 years of teaching experience. Her
areas of interest's Image processing, operating system, computer security and web applications
security. She can be contacted at email: 110036@uotechnology.edu.iq
