https://doi.org/10.31449/inf.v48i14.6073  Informatica 48 (2024) 117-130        117 
 
Chaotic Encryption-Based Network Robot for Indoor Security and 
Remote Video Monitoring 
 
*
Man Zhang, Ning Chen 
Chongqing Technology and Business Institute, Chongqing, 401520, China 
E-mail: zhangman@cqtbi.edu.cn; chenning@cqtbi.edu.cn 
*
Corresponding author 
 
Keywords: video encryption, video encrypted transmission, home security robot, multi-hazard information acquisition, 
active sniffing 
Recieved: April 20, 2024 
As society develops and living standards improve, people are putting forward higher and higher requirements 
for their living environment. To fulfill interior security requirements, fire security, and building automation, 
intelligent robotics has come into being. Video data under intelligent network remote monitoring is different 
from traditional data, as video is a kind of data with a large amount of information, highly complex coding 
and decoding, difficult to process, and requires real-time operation. Therefore, it is important to focus on the 
format characteristics of H.264 when designing the encryption scheme. Due to the heavy computational 
burden caused by the large amount of video data, it is not possible to encrypt all data when encrypting video 
data, and at the same time to ensure that the traditional encryption algorithms, such as Data Encryption 
Standard (DES) and Rivest-Shamir-Adleman (RSA), are no longer suitable for video encryption applications 
to ensure that the format specificity and good operability are not compromised. Therefore, this paper aims to 
design a sufficiently secure and practical video encryption scheme for the remote monitoring of network robots 
based on the H.264 codec standard by introducing a chaotic iterative system, and finally to design and 
implement an indoor security method for remote monitoring of video technology based on chaotic encryption 
for network robots with certain practical application value.  
Povzetek:Študija predlaga razvoj varnega video enkripcijskega sistema za oddaljeno spremljanje z uporabo 
robotske tehnologije, temelječe na H.264 in kaotičnem iterativnem sistemu za izboljšanje notranje varnosti. 
 
1 Introduction 
With the rapid development of China's economy, people's 
living standards continue to improve, and people also put 
forward a higher level of requirements for the family living 
environment [1]. Indoor security system is an indispensable 
part of the modern residential security system, in the process 
of modern urban construction, a lot of homes use advanced, 
scientific, and technical means to carry out community 
management and maintenance [2-3]. However, the 
traditional sense of the community access control system is 
due to its complex functions and was built to reduce its use 
[4-5], in addition to many intelligent management systems 
although it achieves some conventional control, but cannot 
fully meet the people of the family living environment 
intelligent monitoring of this feature. The network robot 
remote video technology under the indoor security 
monitoring system, can be on people's living environment, 
home security, and other aspects of real-time monitoring [6-
7]. It can not only achieve remote control and centralized 
management functions, but also send alarm signals to the 
security center through the alarm device when some 
unexpected events occur, to achieve rapid processing 
purposes [8], and at the same time can also upload the 
dangerous situation information to the dispatching room, so 
that the management staff can take corresponding measures 
in time. Therefore, it is important to explore new methods of 
indoor security based on chaotic encryption of the network 
robot remote monitoring video technology, for today's 
indoor security [9-10].   
 
 
 
 
 

118        Informatica (2024) 117-130  M. Zhang et al. 
 
Table 1:  Related work
Reference Objective Method Result Limitations  
[11] The paper 
investigated chaotic 
encryption as an 
option for security 
issues related to 
video leakages, 
testing the viability 
of the idea by 
establishing a 
working prototype 
system in place for 
real-time video 
communications. 
A prototype system for 
transmitting video in 
real-time was created to 
verify the efficacy of 
approach, which uses tent 
connecting to improve a 
chaotic encryption 
algorithm for enhanced 
performance. 
The result demonstrated 
the viability and 
efficacy of 
the optimized chaotic 
encryption method in 
real-time video 
transmission by 
showing that it 
outperforms distinct 
techniques and 
incorporates logistic and 
tent mapping. 
The following 
section indicates 
several possible 
drawbacks, such as 
the need for more 
testing in various 
scenarios, unfixed 
vulnerabilities, or 
scalability 
problems with 
bigger datasets or 
in various systems. 
[12] The research 
focused on 
integrated 
programming 
features in robotics 
boards, the work 
attempts to enhance 
the effectiveness and 
safety of algorithms 
for recognizing face
s in robotics 
applications. 
The study suggested 
using a cloud server to 
enhance the robustness, 
efficiency, and speed of 
facial recognition 
processing in robotics 
boards and proposes a 
secure hybrid encryption 
method using bit slicing 
(BS) and discrete Fourier 
transform (DFT). 
The study result shows 
that the suggested 
approach significantly 
outperforms the current 
models and encryption 
algorithms, such 
as advanced Encryption 
Standard (AES), BS-
DFT encryption, 
Genetic Algorithm 
(GA), and 
deoxyribonucleic acid 
(DNA) algorithm. 
A potential BS-
DFT encryption 
algorithm was 
presented in the 
work, although it 
lacks 
comprehensive 
information on 
computational or 
resource expenses, 
as well as deep 
insights into 
potential limits and 
negative aspects. 
[13] The objective solves 
the issues of 
significant 
redundancy and 
inadequate 
communication 
safety for industrial 
robots' multi-
channel real-time 
vibration sensor 
information. 
The study suggested a 
compressed sensing 
(CS)-based approach that 
uses a separate cosine 
transform matrix, a 
chaotic matrix, and cross-
correlation function 
integration for sparse 
decomposition 
evaluation, robot data 
integration, and effective 
encrypted transmission. 
The research results 
show that the suggested 
approach improves 
encryption efficiency 
during transmission 
while also drastically 
reducing the quantity of 
data transferred. That 
maintains signal 
transmission security 
and efficiency without 
compromising useful 
information. 
The suggested 
method's 
shortcomings are 
not specifically 
mentioned in the 
section. 
Difficulties with 
computing 
expenses, 
complexity of 
implementation, or 
certain situations 
where the 
approach could not 
work well are 
illustrations of 
possible 
drawbacks. 

Chaotic Encryption-Based Network Robot for Indoor Security...    Informatica (2024) 117-130    119 
 
[14] The objective of the 
research suggested a 
lightweight 
encryption strategy 
for the Internet of 
Robotic Things 
(IoT) robots, 
utilizing discrete-
time chaotic maps 
like Cubic Map and 
Ricker's Population 
Model Map for 
effective data 
encryption. 
The study developed a 
new chaotic encryption 
mechanism using 
discrete-time maps, 
tested it on the NVIDIA 
Jetson Orin evolution 
package, and evaluated 
its effectiveness using 
statistical tests and 
security evaluations. 
The research 
demonstrated that a high 
rate of image encryption 
per second was possible 
due to the developed 
encryption mechanism's 
suitability for parallel 
processing, providing 
robust security for 
Internet of Things (IOT) 
applications, 
particularly when 
encrypting large files. 
The paper 
suggested an 
encryption 
method, but 
acknowledges its 
limitations, such as 
scalability and 
suitability for 
various IoRT 
scenarios, and calls 
for further analysis 
to assess its 
resilience and 
suitability for real-
world 
implementation. 
[15] The objective of the 
research was to solve 
security issues with 
portrait data by 
setting the proposed 
identity security 
technology platform 
for remote sensing 
(RS) images 
provided by 
Unmanned aerial 
vehicles (UAVs). 
The suggested approach 
makes use of dynamic 
chain Deoxyribonucleic 
Acid (DNA) encoding, 
phase-change 
discontinuous 
propagation, lightweight 
bit-level disorientation 
hash eigenvalue 
extraction, edge 
identification face 
detection technology, 
and selected matrix 
encryption. 
The study's 
experimental results that 
chaotic pseudo-random 
sequence and dynamic 
DNA chained 
encrypting increase the 
resilience of 
cryptographic systems, 
decrease their 
complexity, and 
increase their security 
and efficiency. 
Though efficient 
and secure, the 
suggested system 
lacks information 
regarding possible 
drawbacks such as 
processing 
complexity or 
particular 
situations, and 
further research 
has been done to 
determine how 
well it works in a 
variety of UAV 
and RS 
communication 
contexts. 
[16] The objective of the 
investigation was to 
examine security 
concerns related 
to IoT-based smart 
home systems 
(SHSs) and present a 
thorough synopsis of 
previous studies 
conducted in the 
domain. 
The methodology tackles 
two security issues by 
analyzing SHSs, 
proposing precise 
definitions, analyzing 
design, extracting 
features, and discussing 
cyber-attack methods, 
defenses, security 
frameworks, assessment 
methods, technology 
scenarios, and real-world 
research. 
The study explored 
security concerns in 
SHS through analysis of 
architecture, 
cyberattack techniques, 
defenses, structures, and 
combination scenarios. 
The study makes 
the argument that 
resolving security 
problems can 
prove difficult due 
to the quickly 
growing and 
complicated nature 
of SHSs and that 
the emphasis on 
study cannot 
adequately account 
for real-world 
constraints. 
[17] The primary 
objective of the 
The research proposed 
encryption protocols for 
Comparing the modified 
AES algorithm to the 
The section does 
not mention 

120        Informatica (2024) 117-130  M. Zhang et al. 
 
study propose and 
evaluate the 
lightweight crypto-
encryption protocols 
and content-based 
video encryption 
standards for real-
time video 
surveillance, 
focusing on object 
tracking 
enhancement and 
data security in 
video surveillance 
(VS) technologies. 
object tracking and video 
data protection, including 
weight average 
background reduction, an 
entropy adaptable object 
learning model, and a 
modified Advanced 
Encryption Standard 
(AES) method for multi-
object tracking.  
current encryption 
standards, it 
demonstrates virtually 
NCPR value and 
quickest time. Weight 
average background 
removal techniques help 
the entropy adaptive 
object learning model 
conserve memory and 
enhance multi-object 
tracking performance. 
specific 
limitations, but 
potential 
drawbacks include 
integration 
difficulties, 
additional 
validation 
requirements, and 
balancing security 
and computing 
performance. 
[18] The study 
investigated 
enhancing the safety 
features of a multi-
chaotic systems-
based color image 
encrypting technique 
by addressing 
defects such as poor 
statistical 
characteristics and 
vulnerability to 
known-plaintext and 
chosen-plaintext 
attacks. 
The approach involves 
altering the pixel-
chaotic-shuffle method 
of image encoding, 
establishing a new pixel-
chaotic-diffusion 
structure, extracting keys 
from chaotic 
configurations, and 
adjusting shuffling time 
based on plain images. 
The modified 
cryptosystem has 
greater statistical 
properties and more 
resilient to known and 
specific plaintext 
attacks, according to the 
study findings. 
The suggested 
approach can be 
impacted by 
possible 
drawbacks such as 
higher computing 
complexity, speed 
trade-offs for 
security, and 
difficulties 
incorporating 
modifications in 
current systems. 
[19] The study explored 
the challenge of 
preventing 
unauthorized access 
to sensitive image 
information using a 
new privacy-
preserving method 
combining a chaotic 
dynamical system, 
Arnold 
transformation (AT), 
and the code for 
DNA sequencing. 
The proposed privacy-
preserving method 
utilizes an Arnold 
transformation, DNA 
sequencing code, and a 
chaotic dynamical 
system to create an initial 
S-box. Various 
techniques, such as the 
National Institute of 
Standards and 
Technology, histogram 
analysis, nonlinearity 
analysis, and strict 
avalanche criterion 
(NIST, HA, NL, SAC)   , 
are used in tests to 
validate the randomness 
and security of the S-box. 
The effectiveness of 
suggested system was 
demonstrated by its 
strong security 
measures and resilience 
to different types of 
assaults, which 
havebeen evaluated and 
comparedwith current 
approaches. 
The suggested 
privacy-preserving 
scheme can have 
drawbacks in real-
world 
implementation 
and scalability, and 
its efficiency in 
practical scenarios 
can require further 
evaluation. 

Chaotic Encryption-Based Network Robot for Indoor Security...    Informatica (2024) 117-130    121 
 
[20] The study examined 
the security issues in 
SHS through 
consideration of 
cyberattack tactics, 
defenses, 
architecture, 
structures, and 
combination 
situations. 
The methodology used 
involves analyzing 
several SHS-related 
factors, including their 
architecture, cyberattack 
tactics used, defense 
systems in place, 
structural weaknesses, 
and possible situations in 
which numerous attack 
vectors are merged. 
The study of the results 
indicates possible weak 
points, attack routes, 
and defense systems for 
SHS, offering insights 
into the difficulties and 
dangers involved in 
protecting against 
cyberattacks. 
The analysis 
provides insightful 
information on 
SHS security 
issues, but it could 
be affected by the 
study's breadth, the 
accuracy of data, 
the evolution of 
cyber threats, and 
the efficacy of 
suggested 
defenses. 
 
There are issues with the suggested BS-DFT encryption 
algorithms scalability, weaknesses, and computational 
complexity. Chaotic encryption is suggested as a solution, 
providing increased efficiency and security. To guarantee its 
resilience in IoRT and SHS settings, more investigation is 
required. 
2 Video encryption algorithm based on 
chaos principle 
2.1  Analysis of STC principle 
Here the author uses one of the most widely used models in 
the study of spatio-temporal chaotic systems, i.e. the coupled 
image lattice to analyse the spatio-temporal chaotic 
behaviour. For the reaction-diffusion process.  
𝜕 ₜ𝜇 =𝐹 (𝜇 )+𝛿𝛻 ²𝑢                                                      (1) 
Here it is divided into two processes, one of which is the 
local chaotic reaction process; the local reaction process can 
be formulated as a parallel non-linear image in one 
dimension, and its constraints are as follows. 
𝑥 (𝑖 )−>𝑥 ′(𝑖 )=𝑓 (𝑥 (𝑖 ))                                                     (2) 
where x is the state of the current process, i are the 
coordinates of the lattice and L is the number of lattices of 
the system in the current STC. The second process is the 
overall diffusion process. For the diffusion process, we 
assume that each lattice is only associated with the lattice 
nearest to it, and here the Labrador operator discretization is 
used to calculate the final connected image lattice model in 
one dimension was found to be 
𝑋 ₙ₊₁(𝑖 )=(1−𝜀 )𝑓 (𝑋 ₙ(𝑖 ))+𝜀 /2[𝑓 (𝑋 ₙ(𝑖 +1))+
𝑓 (𝑋 ₙ(𝑖 −1))                                                      (3) 
It is shown that the coupled image lattice system is 
considered a discretized dynamical system, both from the 
temporal and spatial perspectives. 
2.2  H.264 STC-based video encryption 
algorithm 
The residual data is encoded in the basic grade specified in 
the H.264 standard using the CAVLC encoding mode, which 
comes after the intra-frame inter-frame predicted method. 
This study presents a method that uses two modules: the 
space-time chaos encryption module and the encryption 
selection management module, to encrypt a tiny quantity of 
data with great security. Two chaotic sequence systems 
provide two pseudo-random sequences that are used to 
encrypt the data and regulate the encryption preference, 
respectively. The primary video stream proceeds to the 
encoder for compressing coding. Here, the pattern data 
following intra-frame forecasting pattern choosing and inter-
frame chunking sequence selection is chosen and protected 
by the encryption choosing manage module. The remaining 
coded residual data proceeds to the CAVLC for the entropy 
coding manipulation, where the pertinent coding parameters 
in the coder process are additionally chosen and secured by 
the encryption preference control section utilizing the 
pseudo-random sequence flow produced by the STC system 
[21]. Through incorporating randomness into encryption 
algorithms and strengthening their resistance to brute-force 
assaults, chaotic encryption improves indoor security. This 
strategy strengthens security measures 
against unauthorized access and possible compromises by 
protecting sensitive data in indoor locations and 

122        Informatica (2024) 117-130  M. Zhang et al. 
 
guaranteeing the integrity and security of the data 
transferred.  
2.2.1 Selective encryption control algorithm design 
The DCT transformation procedure can provide a high-
security encryption impact when the data is encrypted, it is 
simple to lose the video data's original compression 
properties, this method decides to encrypt the data, which 
has a significant influence on the decrypted video's ratio of 
compression intra-frame inter-frame prediction mode and 
the scanning sequence of the residual data and related 
parameters in the CAVLC encoding process.  
H.264 uses context-based coding in both the entropy coding 
process, i.e. during the coding process, instead of coding the 
current data singularly, it refers to the situation of the data 
already coded before and after, while there are multiple code 
tables used for coding in the coding process, the coding only 
needs to select the corresponding data in the code table 
dynamically according to the coding value, ensuring high 
efficiency of coding by.  
• The number of trailing coefficients and the sum of 
non-zero coefficients are first encoded in the 
residual coefficients.  
• Encoding the sign bits of trailing coefficients in the 
range (0 - 3).  
• Encode the magnitude of the other non-zero 
coefficients excluding the trailing coefficients.  
• Next, the total number of zeros preceding the last 
non-zero coefficient in the residual coefficient 
block is encoded.  
• Finally, the number of zeros preceding each non-
zero coefficient in the residual block is encoded. 
 
Figure 1: Encryption selection control module design 
Except for (1) and (4) where the modification of the 
encoding control parameters will result in changing the 
coefficient control parameters of the original message 
format, the encoding parameters in the other three processes 
without compromising the functionality of the original file 
format, certain factors, such as the sign bits of the following 
parameters, the amplitude for the non-zero coefficients, and 
the number of zeros preceding each non-zero coefficient, can 
be utilized as encryption options; To improve the encryption 
speed, some schemes only select the trailing coefficient 
symbol bits for encryption operation, which can achieve fast 
encryption but also sacrifice part of the security. To solve 
the above problems, the encryption algorithm in this paper 
is designed as a selective encryption control module, that 
produces a pseudo-random pattern of streaming key using 
the more widely utilized logistic equation [𝑖 ]. These sites are 
encrypted using a pseudo-random sequence that is utilized 
to make a decision. Data encryption and when the key value 
should be used are determined by the parity of the pseudo-
random sequence [𝑖 ]. Figure 1, if the read is 1, the relevant 
data is encrypted at that moment; if it is 0, no encryption is 
chosen. 
The encryption selection control pseudo-random sequence 
generator adopts equation (4) as the generation equation of 
the pseudo-random sequence, and uses the mean threshold 
method to generate the sequence according to equation (5) 
as follows. 
𝑥 ᵢ=𝜇𝑥 ᵢ₋₁(1−𝑥 ᵢ₋₁)                                               (4) 
{
 
 
 
 
𝑎𝑣𝑒𝑟 =∑
𝑖 =0
𝑛 𝑥 𝑖 /𝑛 𝑘𝑒𝑦 [𝑖 ]=0; 𝑖𝑓 𝑥 𝑖 <𝑎𝑣𝑒𝑟 ;
𝑘𝑒𝑦 [𝑖 ]=1; 𝑖𝑓 𝑥 𝑖 >𝑎𝑣𝑒𝑟 ;
                          (5) 
where n is the overall number length of the unique floating-
point sequence and aver is an average value of the randomly 
generated original floating-point sequences. A binary 
sequence considers key[i] equal to 0 when the current 
sequence considers 𝑥 𝑖 is smaller than ever, and 1 when 𝑥 𝑖 is 
larger than ever. The security control is selected using the 
final produced binary pseudo-random sequence key. The 
location to be encrypted includes intra-frame prediction 
pattern parameters, inter-frame chunking pattern 
parameters, CAVLC pre-encoding zigzag rearrangement 
residual data, encoding trailing coefficient sign bits, and 
non-zero coefficient amplitude in five parts. encryption 
module, the scheme becomes dynamic encryption, and the 
encrypted data and the encryption effect of each encryption 
sub-process are different. This makes the encryption 

Chaotic Encryption-Based Network Robot for Indoor Security...    Informatica (2024) 117-130    123 
 
algorithm more secure and can achieve better encryption 
results. 
2.2.2 STC model design  
This work presents the design of an STC encryption module 
depending on the STC system, based on the encryption 
management module design. The module is designed to 
generate a pseudo-random sequence stream of 512 bits in 
length for encryption through the initial value acquisition, 
data truncation, and transformation of the STC box by the 
user inputting the key and feeding the key into the STC 
system. According to the analysis in the previous section, the 
coupled image lattice as an STC model is the most common, 
which has the characteristics of high computational 
efficiency, high degree of parallelism, compatibility with 
traditional theories, easy analysis, etc. Here we namely 
choose the neighboring STC model (Equation 6), and the 
algorithm is chosen as Equation (7). 
𝑋 ₙ₊₁(𝑖 )=(1−𝜀 )𝑓 (𝑋 ₙ(𝑖 ))+𝜀 /2[𝑓 (𝑋 ₙ(𝑖 +1))+
𝑓 (𝑋 ₙ(𝑖 −1))]                                                       （6 ） 
𝑋 𝑖 +1
=8
∗
𝑋 𝑖 4
−8
∗
𝑋 𝑖 2
+1                                           （7 ） 
as the local chaos equation, with x taking values in the range 
(1,1). 
Table 2: An estimated table of entropy indices and 
Lyapunov coefficients 
Local mapping Logistic X,+₁=8*X₄-
8*X²+1 
Lyapunov 
exponent 
0.6501 1.3871 
Local mapping Logistic X,+₁=8*X₄-
8*X²+1 
approximate 
entropy 
0.6567 1.2904 
Chaotic 
equation 
Logistic Formula (7) 
Lyapunov 
exponent 
0.6503 1.2816 
Local mapping Logistic X,+₁=8*X₄-
8*X²+1 
Lyapunov 
exponent 
0.6136 1.2816 
 
By testing the data associated with the local chaos mapping 
(Eq. 7) in this STC, Table 2 shows the Lyapunov exponent 
and the approximate entropy index. 
It can be seen that the performance associated with the local 
chaotic mapping (7) is better than that of the Logistic 
equation. Compared to ordinary chaotic systems, STC has 
extended the chaotic properties to space, ensuring that it is 
more complex and the resulting pseudo-random chaotic 
sequences for encryption are more difficult for an attacker to 
break. Additionally, based on the attributes of the connected 
image lattice model, lattices are quite closely related to each 
other, and the one-way robustness ensures that it is very 
difficult to break through by a reverse attack, greatly 
increasing the difficulty of attackers to break through and 
increasing the security of the encryption system. 
3 Encoding encryption algorithm 
design 
3.1 Predictive pattern dislocation encryption 
The intra-frame chunking patterns and the inter-frame 
predictive structure cause a fixed pattern categorization in 
the encrypted information. The 4x4 intra-frame forecast 
component, for instance, has nine prediction options. 
Therefore, the encryption cannot be scrambled directly, 
otherwise, the encoding pattern will be illegal and cannot be 
decoded properly. In this paper, a pseudo-random sequence 
is generated by the STC system to scramble the prediction 
patterns of the INTRA_4x4 block. The STC system 
repeatedly generates the pseudo-random sequences, which 
improves the security of the encryption. The encryption 
process is as follows. 
𝑢𝑅𝑎𝑛𝑑𝑆𝑒𝑞 =𝑆𝑝 𝑎 −
𝐶 ℎ𝑎𝑜 𝑠 −
𝐺𝑒 𝑛 𝑒𝑟𝑎𝑡𝑜𝑟 (𝑘 ′
3) ； 
𝐸𝑛𝑐𝑟𝑝 _𝑀𝑜𝑑𝑒 =𝑀𝑜𝑑𝑒 ⊕𝑢𝑅𝑎𝑛𝑑𝑆𝑒𝑞 ;  
As previously examined, the inter-frame chunking manner 
predicts that the chunking mode inside the frame could be 
split into seven variable-sized block modes 16x8, 8x16, 
16x16, 4x4, 8x4, and 8x8. There is a protected motion vector 
for each of these modes, and the total amount of motion 
vector values varies between blocks. To prevent the 
decoding errors caused by the scrambling of the block 
patterns, this paper does not directly encrypt the inter-frame 
block patterns, but adopts a scrambling swap method, by 

124        Informatica (2024) 117-130  M. Zhang et al. 
 
creating a 1-bit pseudo-random pattern containing zero, we 
may decide whether to jumble the two blocks patterns that 
have the same motion vector. Using an encryption control 
selection module, the method in this study selects to jumble 
the pseudo-random sequence produced by the STC system 
during the CAVLC encoding phase, while ensuring that the 
amount of data encrypted is as small as possible and the 
security of the encryption is as high as possible, without 
destroying the original encoding format. The security of the 
encryption is improved as much as possible while ensuring 
that the amount of data encrypted is as small as possible. 
3.2 Residual data zigzag scanning sequence 
encryption  
After the DCT transformation and quantization of the H.264 
data encoding process, the energy of the residual data is 
mainly concentrated in the low-frequency and DC regions. 
After quantization, the coefficients of the low-frequency and 
DC components are mostly zero except for a small number 
of larger values. Therefore, to encode more efficiently, 
before entropy encoding, the residual coefficients can be 
zigzag scanned according to the statistical characteristics 
from high to low, to achieve better encoding efficiency. 
In this paper, the zigzag scan of the residual data is 
scrambled to prevent the cryptographic scrambling from 
seriously damaging the characteristics of the DCT after 
coding, resulting in a significant reduction in the efficiency 
of entropy coding and compression. As shown in Figure 2, 
we choose to scramble the scanning direction of each 
diagonal line in the zigzag scanning process. The parity of 
the 1-bit frequency of the pseudo-random sequence 
produced by the spatio-temporal chaotic pseudo-random 
sequence generator is used to determine the read direction of 
each diagonal, thereby encrypting the zigzag scan sequence. 
Figure 3 shows the encryption method is illustrated. 
 
Figure 2: Zigzag saw tooth scan rule 
 
 
Figure 3: Zigzag data scanning scrambling process 
In Figure 3, the 8x8 data block scan is used as an example. 
The scanning order of each diagonal data is determined by 
the parity of the sequence value of each 1bit, and when the 
read sequence value is 1, the scanning order of the 
corresponding diagonal data block is inverted to obtain the 
scrambled scanning order of the residual data. This ensures 
that the overall data scanning order is still from top-left to 
bottom-right, but disrupts the specific scanning order, and 
achieves encryption of the residual data with fewer 
operations and less impact on the compression efficiency.  
3.3  CAVLC encoding process encryption   
The several straggling coefficients in the CAVLC encoding 
process are between 0 and 3, so encryption of trailing 
coefficients requires at most 3 bits of pseudo-random 
sequence to heterodyning with the T1_Signs. The non-zero 
coefficients other than the trailing coefficients are divided 
into prefixes and suffixes, and the encoding process starts 
with the conversion of 𝑙𝑒𝑣𝑒𝑙𝐶𝑜𝑑𝑒 =(𝑙𝑒𝑣𝑒𝑙 <<1)−
2  𝑖𝑓  𝑙𝑒𝑣𝑒𝑙 >0; 
𝑙𝑒𝑣𝑒𝑙𝐶𝑜𝑑𝑒 =−(𝑙𝑒𝑣𝑒𝑙 <<1)−1  𝑖𝑓   𝑙𝑒𝑣𝑒𝑙 <0;  
𝑙𝑒𝑣𝑒𝑙 ₋𝑝𝑟𝑒𝑓𝑖𝑥 =𝑙𝑒𝑣𝑒𝑙𝐶𝑜𝑑𝑒 /(1<<𝑠𝑢𝑓𝑖𝑥𝐿𝑒𝑛𝑔𝑡 ℎ)  
𝑙𝑒𝑣𝑒𝑙 _𝑠𝑢𝑓𝑓𝑖𝑥 =𝑙𝑒𝑣𝑒𝑙𝐶𝑜𝑑𝑒 %(1<<𝑠𝑢𝑓𝑓𝑖𝑥𝐿𝑒𝑛𝑔𝑡 ℎ);  
If the level Code value is encrypted, the compression 
efficiency of the encoding may be seriously affected, and on 
the other hand, the encrypted data may not be legal and 
cannot be decrypted. Therefore, in this paper, the encryption 
operation is carried out for level _ prefix. Since the encoding 
is not directly encoding the amplitude value but encoding the 
value in the corresponding code table through the level _ 
prefix lookup table, the encryption of the non-zero 
coefficient amplitude value is achieved indirectly by 
scrambling the level _ prefix here, and the encryption 
process is as follows 
𝑢𝑅𝑎𝑛𝑑𝑆𝑒𝑞 =𝑆𝑝𝑎 ₋𝐶 ℎ𝑎𝑜𝑠 ₋𝐺𝑒𝑛𝑒𝑟𝑎𝑡𝑜𝑟 (𝑘 ,
4);  
𝐸𝑛𝑐𝑟 𝑝 _𝑙𝑒𝑣𝑒𝑙 _𝑝𝑟𝑒𝑓𝑖𝑥 = 𝑙𝑒𝑣𝑒𝑙 _𝑝𝑟𝑒𝑓𝑖𝑥 ⊕𝑢𝑅𝑎𝑛𝑑𝑆𝑒𝑞 ;  

Chaotic Encryption-Based Network Robot for Indoor Security...    Informatica (2024) 117-130    125 
 
4  A new approach to indoor security 
design and implementation of a 
networked robot remote monitoring 
system under chaotic algorithm 
4.1  Overall design and functional design 
The system is capable of implementing basic video data 
playback, video encryption, and decryption functions, video 
encoding and decoding functions, as well as video 
encapsulation and encryption-related performance testing 
functions in several parts.   
4.1.1  Encryption security level selection module   
If a video encryption system only provides a single 
encryption scheme, then the system may not be able to cope 
with the encryption needs of users with different needs. 
Some users do not want their video data to be encrypted too 
well, instead, they want to be able to encrypt only part of the 
effect, or even be able to see part of the original video 
information from the encrypted video data, for example, 
commercial video applications, businessmen in attracting 
users to buy For example, in commercial video applications, 
merchants want to entice users to buy video services by 
allowing them to see some of the information but not all of 
it, so that they can pay for it. Some users, on the other hand, 
do not care about the time and expense of encryption but 
want their video data encrypted as effectively and securely 
as possible, such as military, government, confidential 
commercial video conference data, etc. Consequently, to 
satisfy the various customers' expectations for video 
encryption, as a video encryption system is bound to provide 
a variety of security encryption schemes for users to choose 
from, the system designed in this paper provides three levels 
of encryption schemes according to the different levels of 
encryption security, namely primary, intermediate and 
advanced. 
The primary encryption scheme uses an STC model with 
fixed local equations, which only encrypts the magnitude of 
non-zero coefficients of the residual coefficients and the sign 
bits of the trailing coefficients. This is generally suitable for 
video data with small motion information and low 
requirements for encryption effect; the intermediate 
encryption scheme uses the space-time chaotic model with 
fixed local equations and adopts a selective encryption 
control module. Based on the primary one, the scanning 
order of the residual coefficients is encrypted and the 
prediction pattern between frames is encrypted. It is 
generally suitable for video data with a large amount of 
motion information and also with certain requirements for 
encryption security. The advanced encryption scheme 
selects the STC model with dynamically changing local 
equations, introduces motion vector directional values and 
sign bit encryption based on the intermediate level, encrypts 
the sign bits of non-zero coefficients, and does dynamic 
encryption frame by frame. It is generally applicable to 
video data with very high-security requirements. 
4.1.2  Key matching symmetric module  
As the video encryption and decryption scheme in this paper 
is based on the H.264 encoding and decoding process, how 
to ensure the synchronization of the encrypted and decrypted 
data is a crucial point of the whole encryption scheme, if the 
decryption process cannot be synchronized with the 
encryption process, the correct decryption cannot be carried 
out. Since the H.264, the encoder has many time-spending 
calculation functions in the encoding process, these 
functions also perform encoding operations during the 
encoding process, but their main function is to budget the 
encoding overhead and does not perform encoding 
operations on the encoded data, while there are no such 
functions on the decoding side, so if the encryption is 
directly targeted at the encoding process, the decryption will 
not correspond to the encryption, so it is necessary to, 
Therefore, each encryption operation must be operated 
separately, i.e. the encryption process is only implemented 
in the real encoding operation on the encoding side, while 
the function for predicting the encoding overhead is left 
unchanged, to assure the encoded and decoded sides' 
encryption and decryption synchronization. The integration 
of advanced encryption and decryption techniques is 
achieved in indoor security through the use of chaotic 
encryption network robotic remote monitoring video 
technology. To protect video data both during transmission 
and preservation, the system uses chaotic encryption 
techniques. Increasing data security, these algorithms 
produce inconsistent patterns.   
More computing resources may be needed for both 
encryption and decryption of stronger encryption 
techniques, which are generally weighed against 
computational complexity when determining effective 
security encryption. Selected encryption methods must 
balance resilience and controllable computing complexity in 
order to provide the greatest possible security. Users can 
observe real-time video feeds from their computers or 
smartphones by using remote monitoring. Interior 
conditions are well protected against potential security 
concerns by encryption, which makes sure 
that unauthorized persons cannot access or interfere with the 
video data. In the H.264 codec, the functions for the intra-
frame prediction mode coding operation are 

126        Informatica (2024) 117-130  M. Zhang et al. 
 
Write Syntax  
Element_Intra4x4Prediction Mode (Syntax Element*se, 
Data Partition *this_data Parr). 
Under normal circumstances, the function is called as shown 
in Figure 4. We can see from the diagram that three of the 
four calls are to functions with RDCOST, i.e. they all act as 
budget encoding overheads and do not involve specific 
encoding, so for the operation of intra-frame predictive 
mode encryption, these three calls need to be avoided, i.e. as 
shown in Figure 5. 
 
Figure 4:  The in-frame prediction mode encoding function is called 
 
 
Figure 5: Budget coding attrition process calls the in-frame predictive mode coding process 
4.2 Video encoding and encryption module 
design 
First, the intra-frame prediction and motion vector 
difference dislocation encryption operation. In the 
scrambling process, since the performance of inter-frame 
chunking mode encryption is relatively not very high, so I 
discarded the operation of inter-frame chunking mode in the 
previous paper and only encrypted the intra-frame prediction 
mode in the prediction mode information encryption, 
meanwhile, to ensure the scrambling of inter-frame motion 
information, I chose to introduce motion vector values for 
encryption here. 
In the inter-frame prediction mode, each chunk will have an 
MV to be encoded, assuming that the current chunking mode 
is 4x4, then 16 MVs are needed for encoding, and each MV 
contains both horizontal and vertical values, so it would take 
a lot of space to encode the MVs directly, so the H.264 
encoding process makes use of the fact that the MVs 
between adjacent chunks have contextual relationships to 
the current chunk's Here, the actual MV values is deducted 
from its closest neighbouring block's MV value, and the 
remaining amount is seized, i.e. the motion vector difference 
value is encoded. Therefore, this chapter generates a 1-bit 
pseudo-random sequence to determine whether the 
horizontal and vertical components of the motion vector 

Chaotic Encryption-Based Network Robot for Indoor Security...    Informatica (2024) 117-130    127 
 
difference are to be encrypted interchangeably, as shown in 
the following equation. 
uRandSeq = Spa_Chaos_Generator(k,1); 
 if(uRandSeq) 
{ 
temp=MVD-x;  
MVD-x=MVD; 
 MVD-y=temp; 
} 
where uRandSeq is the generated pseudo-random sequence; 
Spa _Chaos _ Generator is the spatio-temporal chaotic 
pseudo-random sequence generation system, k is the user 
key, and 1 represents the generation of a 1-bit length pseudo-
random sequence.  
temp is the temporary storage value, and MVD-x and MVD-
y are the horizontal and vertical values of the motion vector 
difference, respectively. By encrypting the directional 
values of the motion vectors, the motion information 
appearing in the encrypted video is severely scrambled, 
which plays a very good encryption effect and security 
performance. 
Secondly, encrypting the motion vector difference sign bits. 
In the previous section, this scheme encrypts the directional 
value of the motion vector difference, and to better disrupt 
the motion information in the encrypted video, the motion 
vector difference symbolic bit encryption operation is 
continued here. In the H.264 encoding process, the motion 
vector difference is used as a signed value for signed 
exponential Columbus encoding, in which a signed bit 
encryption operation can be performed for the write stream 
process, as shown in the following equation. 
uRandSeq = Spa_Chaos_Generator(k,1); 
Encrp_MVD_sign = MVD_sign uRandSeq; 
where uRandSeq is the generated pseudo-random sequence; 
Spa _Chaos _ Generator is the spatio-temporal chaotic 
pseudo-random sequence generation system, k is the user 
key, and 1 represents the generation of a 1-bit length pseudo-
random sequence. Encrp _ MVD _ sign is the sign bit of the 
current motion vector difference MVD after encryption, and 
MVD _ sign  
is the sign bit of the motion vector difference MVD before 
encryption. By encrypting the sign bits of the motion vector 
difference, the improved scheme has a more confusing 
encryption effect for motion information. 
Thirdly, the CAVLC entropy coding encryption operation. 
The leftover block data's zigzag scanned sequence before to 
encoding remains mixed in the CAVLC entropy decoding 
procedure, and the sign bits of the trailing coefficients in the 
CAVLC process are also encrypted. This improvement 
improves the efficiency of the encryption scheme without 
reducing the quality of the encryption. 
4.3 Implementation of the video surveillance 
module of the robotic encryption system with 
chaotic algorithms 
Video surveillance assumes a very important responsibility 
in home security. The cost and risk of crime increase with 
video surveillance, which can theoretically lead to a 
reduction in the crime rate, and although it does not directly 
prevent criminal activity, it can provide a strong line of 
defense to deter crime and provide strong evidence. 
Moreover, even when traveling on business or on the way to 
work, you can inspect everything at home remotely at any 
time. The video surveillance module in this article consists 
of a gimbal and a motion camera. The gimbal uses a three-
axis brushless head for aerial photography, which is stable 
and shake-proof, with a wide range of angle adjustments.  
 
Figure 6: Video monitoring software flow chart 
The sports camera has a high-resolution wide-angle lens 
with high resolution and a US2.0 direct plug interface for the 
output. The designed video surveillance program, when the 
video surveillance task is ready, first initialize the camera 

128        Informatica (2024) 117-130  M. Zhang et al. 
 
and rudder, after confirming that the camera and rudder are 
normal, start acquiring the video stream, during which the 
direction and angle of the video surveillance can be 
controlled by the rudder driver after the video acquisition 
task is finished, carry out video encoding transmission to 
complete the video surveillance task. The flow of the video 
monitoring software is shown in Figure 6. 
4.4  Encryption algorithms  
This investigation offers a chaotic encryption-based DES-
improved security solution for interconnected robots in 
remote monitoring systems. The method provides strong 
data protection by combining chaotic encryption with DES. 
This strategy improves security dependability and security 
while providing an efficient and secure of remote 
environment monitoring. Chaotic encryption provides 
unpredictability to the network, increasing security, whereas 
RSA encryption is used for secure transmission of 
information in interior domains. This combination 
encryption strategy is used by robots that are outfitted with 
remote monitoring video technology to protect critical data 
and guarantee the integrity of video updates, therefore 
strengthening protections against any intrusions. Figure 7 
shows the outcome of the RSA and DES. 
 
 
Figure 7: Outcome of the RSA and DES 
4.5  Sensitivity  
Sensitive information is protected by security encryption, 
which encrypts it to thwart illegal access or interception. 
This protective mechanism maintains the security and 
privacy of sensitive data in a variety of situations by 
preventing from possible dangers, such as 
hackers, unauthorized users, or hostile entities. AlexNet [22] 
scored 0.64 %, VggNet [22] scored 0.79%,  and GoogLeNet 
[22] scored 0.24%, the proposed approach Chaotic 
encryption has the greatest sensitivity level of 0.92%.  Figure 
8 shows the outcome of sensitivity.  
 
Figure 8: Outcome of sensitivity 
4.6  Discussion  
The comparatively small 56-bit key length of the DES is one 
of its drawbacks. This key length is susceptible to brute force 
assaults in the modern computer environment, in which an 
attacker continuously tries every potential key before the 
right one is discovered. These kinds of assaults are doable in 
a fair amount of time given the capabilities of current 
processors. Furthermore, improved algorithms for 
encryption such as providing longer key lengths and greater 
cryptographic features, have eclipsed DES. For this reason, 
in settings where higher encryption requirements are 
necessary, DES is seen as being less appropriate for 
protecting sensitive data. The processing complexity of RSA 
encryption is one of its drawbacks, especially when using 
large key sizes. Since RSA depends on intricate 
mathematical processes such as modular exponentiation, it 
can be resource-intensive, particularly on appliances with 
minimal computing power or in situations where encryption 
and decryption are required instantly. Performance and 
usability can be negatively impacted by difficult key 
handling, especially with greater key sizes. Moreover, 
perfect forward secrecy is not provided by RSA, which 
means that previous communications can be decrypted if a 
private key is compromised. Consequently, while using 
RSA, trade-offs between performance, security, and key 
management must be carefully considered. Sensitivity 
analysis has a few disadvantages, including the potential for 
conclusions to be unclear due to their dependence on 

Chaotic Encryption-Based Network Robot for Indoor Security...    Informatica (2024) 117-130    129 
 
assumptions. Furthermore, inadequate or false insights into 
the robustness of a model or selection can result from 
sensitivity analysis's inability to appropriately capture all 
plausible sources of variability or interactions among 
variables. A suggested technique that makes use of chaotic 
encryption can improve security by overcoming the 
shortcomings of conventional encryption techniques like 
DES and RSA. By introducing unpredictability through 
chaotic systems, chaotic encryption reduces the viability of 
brute force assaults. This technique overcomes the 
drawbacks of conventional encryption algorithms and 
increases security against contemporary threats while 
increasing encryption efficiency. 
5 Conclusion 
In summary, based on the previous work, this paper designs 
a complex STC model to construct a pseudo-random 
sequence suitable for video encryption, thereby improving 
the resistance to attack of this random sequence and thus the 
resistance to attack of the whole video encryption algorithm. 
Based on the encryption algorithm of H.264 codec standard, 
the encryption scheme is designed under the requirement of 
ensuring the encryption efficiency and real-time 
performance, and the encrypted data and location with better 
encryption effect and less impact on the compression 
performance and higher real-time performance are selected 
through multiple sets of experiments, to improve the real-
time performance and operability of the encryption 
algorithm as much as possible while ensuring high enough 
security and less impact on the compression performance. 
The encryption algorithm is designed to ensure high enough 
security while maximizing the encryption algorithms 
operability and real-time performance, with the least amount 
of influence on compression efficiency. Using the H.264 
codec standards as a foundation, a more workable video 
encryption system is created. Based on the above, the 
scheme is implemented in the codec platform, and provides 
different encryption options with different security levels 
and different encryption performance according to the 
different encryption requirements, to better provide different 
encryption methods for different scenarios. 
Data availability 
The data used to support the findings of this study are 
included in the article. 
Conflicts of interest 
The authors declare no conflicts of interest. 
Funding statement 
This research study is sponsored by the Chongqing Natural 
Science Foundation Project. The name of the project is 
Research on Chaos-Based Video Encryption Technology for 
Remote Surveillance. The project number is 
CSTB2022NSCQ-MSX1510. Thank the project for 
supporting this article! 
References 
[1]      Fotovvat, Amir. Selective Encryption of Video 
Data for IoT Environments. PhD diss., University 
of Saskatchewan, 2021. 
[2]      Noura, Mohamad. Efficient and secure 
cryptographic solutions for medical data.  PhD 
diss., Université Bourgogne Franche-Comté, 2019. 
[3]      Tolba, Zakaria. Cryptanalysis and improvement of 
multimodal data encryption by the machine-
learning-based system. arXiv preprint arXiv, 2024, 
2402.15779. 
[4]      Ahmad, Rami, Raniyah Wazirali, and Tarik Abu-
Ain. 2022, Machine learning for wireless sensor 
networks security: An overview of challenges and 
issues. Sensors, 22, no. 13: 4730. 
https://doi.org/10.3390/s22134730  
[5]      Williams Phillip, Indira Kaylan Dutta, Hisham 
Daoud, and Magdy Bayoumi. 2022, A survey on 
security in the Internet of Things with a focus on 
the impact of emerging technologies. Internet of 
Things, 19: 100564. 
https://doi.org/10.1016/j.iot.2022.100564  
[6]      Quinga-Socasi, Francisco, Luis Zhinin-Vera, and 
Oscar Chang. A deep learning approach for 
symmetric-key cryptography system. 
In Proceedings of the Future Technologies 
Conference, 2020, pp. 539-552. Cham: Springer 
International Publishing. 
https://doi.org/10.1007/978-3-030-63128-4_41  
[7]      Radwan, Ahmed Salem. Security of IoT Systems: 
AES and Cellular Automata, 2022. 
[8]      Banerjee, Jyoti Sekhar, Siddhartha Bhattacharyya, 
Ahmed J. Obaid, and Wei-Chang Yeh, eds. 2022, 
Intelligent Cyber-Physical Systems Security for 
Industry 4.0: Applications, Challenges and 
Management, CRC Press. 
https://doi.org/10.1201/9781003241348  
[9]      Ramya, P. Sai, and Durgesh Nandan. Analysis of 
security issues and possible solutions in the Internet 
of things for home automation system. In Soft 
Computing: Theories and Applications: 
Proceedings of SoCTA, 2019, pp. 825-836. 
Springer Singapore.  https://doi.org/10.1007/978-
981-15-4032-5_74  

130        Informatica (2024) 117-130  M. Zhang et al. 
 
[10]      Fkirin, Alaa, Gamal Attiya, and Ayman El-Sayed. 
2021. Two-level security approach combining 
watermarking and encryption for securing critical 
colored images. Optical and Quantum Electronics, 
53, no. 6: 285. https://doi.org/10.1007/s11082-021-
02875-2  
[11]      Liu, Beisheng, Xiaodong Li, Haoyang Yu, and 
Jianwen Lv. A light chaotic encryption algorithm 
for real-time video encryption. In 4th EAI 
International Conference on Robotic Sensor 
Networks, 2022, pp. 111-118. Springer 
International Publishing.    
https://doi.org/10.1007/978-3-030-70451-3_10  
[12]      Antonijevic, Milos, Ivana Strumberger, Sasa 
Lazarevic, Nebojsa Bacanin, Djordje Mladenovic, 
and Dijana Jovanovic. 2022, Robust encrypted face 
recognition robot based on bit slicing and Fourier 
transform for cloud environments. Journal of 
Electronic Imaging, 31, no. 6: 061808-061808. 
https://doi.org/10.1117/1.jei.31.6.061808  
[13]      Yu, Xiaojie, Qiao Hu, Dan Xu, Xingju Xie, and 
Yaohui Liu. 2021, multi-channel monitoring data 
compression method for industrial robot based on 
compressed sensing. Measurement Science and 
Technology, 33, no. 1: 
014007.  https://doi.org/10.1088/1361-
6501/ac329c  
[14]      Kiran, Harun Emre, Akif Akgul, Oktay Yildiz, and 
Emre Deniz. 2023, Lightweight encryption 
mechanism with discrete-time chaotic maps for 
Internet of Robotic Things. Integration, 93: 
102047.  https://doi.org/10.1016/j.vlsi.2023.06.00
1  
[15]      Wen, Heping, Zhiyu Xie, Zhuxi Wu, Yiting Lin, 
and Wei Feng. 2024, Exploring the future 
application of UAVs: face image privacy protection 
scheme based on chaos and DNA cryptography. 
Journal of King Saud University-Computer and 
Information Sciences, 36, no. 1: 
101871.  https://doi.org/10.1016/j.jksuci.2023.101
871  
[16]      Yang, Jian, and Liu Sun. 2022, A Comprehensive 
Survey of Security Issues of Smart Home System: 
Spear and Shields, Theory and Practice. IEEE 
Access, 10: 124167-124192.   
https://doi.org/10.1109/access.2022.3224806  
[17]      Kumar, Chandan, and Shailendra Singh. 2024, 
Security standards for real time video surveillance 
and moving object tracking challenges, limitations, 
and future: a case study. Multimedia Tools and 
Applications, 83, no. 10: 30113-
30144.  https://doi.org/10.1007/s11042-023-
16629-7  
[18]      Sharma, Sanjeev, Tarun Kumar, Ravi Dhaundiyal, 
Amit Kumar Mishra, Nitin Duklan, and Ashish 
Maithani. 2019, Improved method for image 
security based on chaotic-shuffle and chaotic-
diffusion algorithms. International Journal of 
Electrical & Computer Engineering, (2088-8708) 
9, no. 1. https://doi.org/10.11591/ijece.v9i1.pp273-
280  
[19]      Masood, Fawad, Junaid Masood, Lejun Zhang, 
Sajjad Shaukat Jamal, Wadii Boulila, Sadaqat Ur 
Rehman, Fadia Ali Khan, and Jawad Ahmad. 2022, 
A new color image encryption technique using 
DNA computing and Chaos-based substitution box. 
Soft Computing,1-17.  
[20]      Radhi, Batool M., and Mohammed A. Hussain. 
2023, Smart Building Security using ESP32 based 
AES One Bio-key and Owner's Biometrics 
Encryption Technology. J. Basrah Res. (Sci.), 49, 
no. 2: 30-47. https://doi.org/10.56714/bjrs.49.2.4  
[21]      Kunhoth, Jayakanth, Nandhini Subramanian, 
Somaya Al-Maadeed, and Ahmed Bouridane., 
2023. Video steganography: recent advances and 
challenges. Multimedia Tools and Applications, , 
82, no. 27: 41943-
41985.  https://doi.org/10.1007/s11042-023-
14844-w  
[22]      Lee, Min-Fan Ricky, and Zhih-Shun Shih. 2022, 
Autonomous surveillance for an indoor security 
robot. Processes, 10, no. 11: 
2175.https://doi.org/10.3390/pr10112175