=Paper=
{{Paper
|id=Vol-3826/paper12
|storemode=property
|title=Integrated protection strategies and adaptive resource distribution for secure video streaming over a Bluetooth network
|pdfUrl=https://ceur-ws.org/Vol-3826/paper12.pdf
|volume=Vol-3826
|authors=Yuliia Kostiuk,Pavlo Skladannyi,Nataliia Korshun,Bohdan Bebeshko,Karyna Khorolska
|dblpUrl=https://dblp.org/rec/conf/cpits/KostiukSKBK24
}}
==Integrated protection strategies and adaptive resource distribution for secure video streaming over a Bluetooth network==
https://ceur-ws.org/Vol-3826/paper12.pdf
Integrated protection strategies and adaptive resource
distribution for secure video streaming over a Bluetooth
network ⋆
Yuliia Kostiuk1,†, Pavlo Skladannyi1,2,*,†, Nataliia Korshun1,†, Bohdan Bebeshko1,†
and Karyna Khorolska1,†
1
Borys Grinchenko Kyiv Metropolitan University, 18/2 Bulvarno-Kudriavska str., 04053 Kyiv, Ukraine
2
Institute of Mathematical Machines and Systems Problems of the National Academy of Sciences of Ukraine, 42 Ac. Glushkov
ave., 03680 Kyiv, Ukraine
Abstract
This paper presents an integrated approach to enhancing the security of adaptive video streams transmitted
over Bluetooth wireless networks with increased data transfer rates using adaptive modulation and a three-
zone buffer. The study addresses key security challenges such as confidentiality, integrity, and availability,
proposing a comprehensive strategy that combines adaptive encryption, dynamic modulation, traffic
multiplexing, and buffer management. The adaptive encryption mechanism allows for real-time
adjustments to encryption levels based on network conditions, ensuring both security and transmission
efficiency. A three-zone buffer policy is introduced, prioritizing the transmission of video data packets (I-
frames, P-frames, and B-frames) according to buffer occupancy and data importance. The use of traffic
multiplexing across multiple transmission paths enhances the availability of video streams, mitigating the
effects of network congestion and packet loss. The paper also explores future directions for video stream
security, including the potential of quantum encryption for unbreakable security and AI-driven techniques
for real-time threat detection and dynamic security adaptation. The relevance of lightweight encryption
methods and edge computing solutions in securing video streams within the Internet of Things (IoT)
environments is also discussed. Overall, the proposed approach balances security and performance, making
it suitable for modern multimedia applications. This research contributes to advancing video stream
protection strategies in wireless networks, ensuring continuous, secure, and high-quality video
transmission.
Keywords
adaptive video streaming, Bluetooth wireless networks, video stream security, artificial intelligence, IoT
security, multimedia applications, real-time video transmission 1
1. Introduction Improving Bluetooth video streaming with adaptive
modulation and a three-zone buffer addresses both data
Video transmission in wireless networks is a key transmission efficiency and security concerns. Ensuring the
component for many modern multimedia applications, such confidentiality, integrity, and availability of video data is
as monitoring systems, video telephony, and personalized critical when using open communication channels [5, 6, 17,
television. Streaming video, using compression and 18, 26–29, 34–36]. Confidentiality limits access to
buffering technologies, enables real-time transmission over authorized users, and the growing risks of unauthorized
local networks and the Internet. This requires high access in the globalized Internet heighten the need for
bandwidth, minimal delays, and acceptable data loss [1–15]. robust security solutions [1–6, 15–18, 37].
However, the Internet does not always guarantee the
required quality of service due to the heterogeneity of 2. Analysis of recent studies and
network structures and video systems. Developing effective
standards for video compression, conversion, and
publications
transmission methods is a significant challenge in In video streaming over Bluetooth networks with increased
information technology [4–6, 14–33]. data rates using adaptive modulation and a three-zone
buffer, protecting confidentiality, integrity, and availability
CPITS-II 2024: Workshop on Cybersecurity Providing in Information 0000-0001-5423-0985 (Y. Kostiuk);
and Telecommunication Systems II, October 26, 2024, Kyiv, Ukraine 0000-0002-7775-6039 (P. Skladannyi);
∗
Corresponding author. 0000-0003-2908-970X (N. Korshun);
†
These authors contributed equally. 0000-0001-6599-0808 (B. Bebeshko);
y.kostiuk@kubg.edu.ua (Y. Kostiuk); 0000-0003-3270-4494 (K. Khorolska)
p.skladannyi@kubg.edu.ua (P. Skladannyi); © 2024 Copyright for this paper by its authors. Use permitted under
Creative Commons License Attribution 4.0 International (CC BY 4.0).
n.korshun@kubg.edu.ua (N. Korshun);
b.bebeshko@kubg.edu.ua (B. Bebeshko);
karynakhorolska@gmail.com (K. Khorolska)
CEUR
Workshop
ceur-ws.org
ISSN 1613-0073
129
Proceedings
is critical [38–43]. Video data encryption is a promising projections to each participant has become outdated.
method, but due to the limitations of some encryption Modern technologies now use compression and advanced
algorithms, adaptive encryption and visual cryptography encoding to optimize data sizes, reduce traffic, and address
are needed to ensure both security and efficiency for large bandwidth limitations, enhancing video stream security
data volumes [7–11, 15, 27, 34–36]. over Bluetooth networks [5, 7–9, 21–27, 34, 35].
Integrity and availability are equally important. While A more advanced approach uses a data segmentation
methods like redundancy and mirroring protect data at the algorithm, dividing image points into non-overlapping
physical level, they do not ensure availability during classes and distributing video data across multiple channels
transmission. Cryptographic methods may secure integrity without adding redundancy, minimizing transmitted data
but are not always available [2, 20–24, 28–31]. Therefore, volume [2–6, 18, 26, 27, 34]. However, if projections are
combining visual cryptography and traffic multiplexing to intercepted, partial information could be exposed, so
secure both confidentiality and availability is essential. segmentation methods must reduce the value of intercepted
As Bluetooth networks evolve, confidentiality and data.
availability protection methods are increasingly vital [31, To ensure availability, a method is needed that allows
32, 44, 45]. Specialized encryption algorithms help secure recipients to restore the image even if some projections are
video streams but may lack sufficient cryptographic blocked or altered by an attacker. Balancing security and
strength. Alternatives like data hiding or digital signatures availability is crucial for protecting video transmissions
address integrity but not availability. Load balancing and over wireless networks [5, 7–10, 13, 14, 17, 19–25].
redundancy improve availability at lower network layers A study of video confidentiality methods over Bluetooth
but do not address the content level, reducing their networks with increased data transfer rates, using adaptive
effectiveness in Bluetooth networks. modulation and a three-zone buffer, proposed a
segmentation method based on pixel brightness values.
3. The purpose of the study Projections would include pixels from a specific brightness
range, with the number of participants determining the
The purpose of this study is to develop and evaluate an class division. For example, in a two-participant scenario,
integrated security framework for adaptive video streaming pixels would be divided into corresponding brightness-
over Bluetooth wireless networks with enhanced data based classes:
transfer rates. The framework aims to address critical
security concerns—confidentiality, integrity, and 1
𝑃 = {(𝑥, 𝑦)|𝑓(𝑥, 𝑦) < 𝑓(𝑥, 𝑦)},
availability—while maintaining high transmission 𝑀𝑀
efficiency. By incorporating adaptive encryption, dynamic (1)
modulation, traffic multiplexing, and a novel three-zone 1
𝑃 = {(𝑥, 𝑦)|𝑓(𝑥, 𝑦) ≥ 𝑓(𝑥, 𝑦)}.
buffer management strategy, the research seeks to ensure 𝑀𝑀
the secure, reliable, and uninterrupted transmission of
For a scenario with an arbitrary number of participants
multimedia content in resource-constrained environments.
where 𝑛 = 2 (𝑘 > 1), the definition of classes is recursive:
The study also explores future trends, such as quantum
2
encryption and AI-driven security adaptations, to further 𝑃 = {(𝑥, 𝑦)|𝑓(𝑥, 𝑦) < 𝑓(𝑥, 𝑦)},
strengthen video stream protection in wireless networks. 𝑀𝑀
( , )∈
(2)
2
4. Methods and Models 𝑃 = {(𝑥, 𝑦)|𝑓(𝑥, 𝑦) ≥
𝑀𝑀
𝑓(𝑥, 𝑦)}.
( , )∈
To address emerging challenges in protecting digital video,
where 𝑃 is the class obtained for the scheme with 2
specialized methods must be developed, focusing on
ensuring confidentiality and availability without relying participants, for the scheme with 2 , i.e 𝑗 ∈ [0; 2 −
solely on encryption algorithms. This can be achieved 1], а 𝑖 = 2𝑗.
through analytical transformation of the data sequence and At the same time, to ensure maximum availability of
route-hiding techniques during exchange, enhancing transmitted video information, a method of uniform image
resilience against unauthorized access using existing segmentation was proposed. Suppose the digital image 𝑃
physical security measures. with dimensions 𝑀 × 𝑀 corresponds to the following
A key approach involves traffic multiplexing systems matrix:
that distribute data fragments across multiple physical
𝑓(0,0) 𝑓(1,0) ⋯ 𝑓(𝑀 − 1,0)
channels, increasing the difficulty of reconstructing the 𝑓(0,1) 𝑓(1,1) ⋯ 𝑓(𝑀 − 1,1)
original data based on specific needs [8, 17]. This method (3)
⋮ ⋮ ⋮
aligns with modern encryption techniques, including 𝑓(0, 𝑀 − 1) 𝑓(1, 𝑀 − 1) ⋯ 𝑓(𝑀 − 1, 𝑀 − 1)
quantum cryptography and AI algorithms, applied at
various network transmission layers. These systems include It is necessary to divide it into nnn projections so that a
intermediate nodes and specialized devices that process and minimal number of projections is required for restoring the
protect data, implementing advanced security measures original image [10–12, 15, 16]. To achieve this, it is evident
across the network [1, 2, 5, 7, 8, 28–32, 44, 45]. The traffic that each projection should contain points evenly
multiplexing model uses sophisticated cryptographic distributed across the entire frame, meaning the projection
protection integrated with advanced data processing will represent a grid with equidistant nodes, and 𝑛 = 2 ,
techniques. The classical method of sending equal-sized 𝑘 ≥ 1:
130
𝑓(0,0) 0 ⋯ 𝑓(√𝑛, 0) 0 ⋯ stream security over a Bluetooth wireless network with
⎡ ⎤
0 0 increased data transfer rates using adaptive modulation
⎢ ⎥
⎢ ⋮ ⋱ ⎥ with a three-zone buffer. Since brightness segmentation
(4)
⎢𝑓(0, √𝑛) 𝑓(√𝑛, √𝑛) ⎥ maximizes the difference between values of 𝑓 the value of
⎢ 0 0 ⎥ the expression (𝑓 − 𝑓 ) in this case will also be maximized.
⎣ ⋮ ⋱⎦
In securing video streams over Bluetooth networks with Let the initial image 𝑓(𝑥, 𝑦) be segmented into 𝑛 projections
increased data rates using adaptive modulation and a three- 𝑓 (𝑥, 𝑦), … , 𝑓 (𝑥, 𝑦) using the brightness segmentation
zone buffer, restoring the original image from incomplete method. Then the mean square error 𝜀 , of the restored
projections is crucial for maintaining integrity and image ℎ (𝑥, 𝑦) based on an arbitrary projection 𝑓 (𝑥, 𝑦)
confidentiality [1, 5, 7–9]. This is achieved by constructing using interpolation methods has the following lower bound:
interpolation functions for known points to estimate 𝜀 ≥ (𝑓 ̅ − 𝑓 ) (6)
unknown ones. The more projections received, the more where 𝑓 is the mean value of 𝑓(𝑥, 𝑦) for projection 𝑖, and 𝑓
accurate the reconstruction, enhancing data protection [2–
is the mean value of 𝑓(𝑥, 𝑦) for the initial image. The proof
6, 10–13, 15–17, 19–22].
of this theorem is conducted by applying the method of
Security risks arise if attackers access a single physical
mathematical induction and transitioning from
channel, enabling them to analyze or modify data. Since
ℎ (𝑥, 𝑦) 𝑡𝑜𝑓 (𝑥, 𝑦).
each channel carries only part of the frame, an attacker
The implementation of methods for ensuring
could reconstruct the original frame from an intercepted
confidentiality and availability in digital video transmission
projection, compromising the stream’s confidentiality [1, 3,
over distributed networks was achieved using advanced
7, 8, 17, 18]. Blocking or altering a projection could further
video processing technologies [1–6, 8–11, 15, 17]. Microsoft
affect integrity and availability, raising concerns about
DirectShow was employed to create components that
system robustness against attacks.
transform the original video stream into projections.
The study examined various methods for restoring
Confidentiality protection relies on modern image
frames from projections, focusing on Bluetooth streaming
segmentation, while the availability filter applies uniform
security with adaptive modulation and a three-zone buffer
segmentation [1–6, 8–11, 15, 17]. These filters process the
[11–16, 19, 20, 46].
stream and generate projections with predefined
To evaluate the effectiveness of image restoration
parameters.
methods and, consequently, video stream protection
These filters were integrated into a filter graph for
methods over a Bluetooth wireless network with increased
secure transmission over Bluetooth networks with adaptive
data transfer rates using adaptive modulation with a three-
modulation and a three-zone buffer. The video source can
zone buffer, the following criteria were identified [19–24].
be any capture device or file, and the Infinite Pin Tee Filter
One of these criteria is the minimum square mean error of
generates the necessary stream copies based on the number
the restored image. This criterion measures the accuracy of
of participants [7, 22–24, 26, 27, 34]. Each copy is segmented
image reproduction after the restoration procedure. It
according to the participant's class number, and streams are
assesses the deviation between the original image and its
transmitted via the ASF Writer filter. The final recipient
restored version. The smaller the value of this criterion, the
combines these streams, ensuring data security and
better the image restoration, and therefore, the more
integrity [19–25, 27, 28, 35, 36].
effective the video stream protection against possible
The system was tested in a high-speed Bluetooth
attacks or distortions.
network with transmission rates up to 1 Gbps, in a large
organization’s network, meeting modern Bluetooth
1 standards [1–7, 9–19, 46]. Intel Core i9 workstations
𝜀 = (ℎ(𝑥, 𝑦) − 𝑓(𝑥, 𝑦)) ⟶ 𝑚𝑖𝑛, (5)
𝑀𝑀 (4.5 GHz) were used to assess system performance. The
main objectives were to determine the optimal number of
The psychovisual criterion for image reproduction system participants based on video stream frame size and
quality is essential when evaluating image restoration evaluate the efficiency of protection methods. Processing
methods for securing video streams over Bluetooth large volumes of video data in the traffic multiplexing
networks with adaptive modulation and a three-zone buffer system could limit overall efficiency.
[7–11, 17]. Several methods were analyzed, including Radio Channel Resource Allocation Algorithms for
bilinear interpolation, bicubic spline interpolation, and Adaptive Video Streaming:
linear extrapolation. Bicubic spline interpolation proved The radio channel resource management algorithms for
most effective in minimizing restoration error for uniformly securing video streams over Bluetooth networks with
segmented projections. A clear link was found between increased data rates, using adaptive modulation and a three-
increased gap size and higher relative error. While gap zone buffer, are designed to ensure reliable transmission
extrapolation worked in some cases, it was ineffective for while meeting security requirements. These algorithms
brightness-based segmentation [27, 28, 30–32, 34, 35, 44, 45]. adjust to network changes such as signal fluctuations and
These results align with the psychovisual quality criterion, congestion, optimizing transmission quality [14, 15, 19–22,
though expert evaluation is advised for specific cases. 28, 35, 36]. Key features include adaptive modulation and
Furthermore, the conclusion was formulated that traffic management, with machine learning methods used to
justifies the advantages of the brightness segmentation predict and optimize transmission, enhancing security and
method over other methods in the context of ensuring video reliability.
131
In parallel, systems for storing and transmitting video over proportional to the value of the buffering factor and directly
HTTP have grown in popularity due to increased mobile proportional to the average bitrate of the stream. Two QoE
device usage, social media integration, and the demand for criteria characterize the buffering factor for a specific user 𝑖:
distance learning. Video transmission is a The normalized ratio of buffering and viewing durations:
telecommunications priority, utilizing two main HTTP- 𝑏
based technologies: non-adaptive (HTTP Progressive 𝑔 = lim (7)
⟶ 𝜔 +𝑏
Download) and adaptive (HTTP Adaptive Streaming),
where 𝑏 is the total buffering duration of user 𝑖 during
defined by the DASH standard. As user mobility and
time 𝑇; 𝜔 is the total video viewing duration by user 𝑖
reliance on wireless networks grow, these networks face
during time 𝑇.
strain, leading to issues like buffering and playback
The ratio of buffering and viewing durations:
interruptions [14, 15, 19–22, 28, 35, 36].
The performance of centralized wireless networks is 𝑏
𝑞 = lim , (8)
significantly impacted by how frequency-time resources are ⟶ 𝜔
allocated among users. Scheduling algorithms at the base In studying wireless systems for video stream security
station handle this resource allocation, but since these over Bluetooth networks with increased data transfer rates,
algorithms vary by manufacturer, they directly affect using adaptive modulation and a three-zone buffer,
overall system performance [9–16, 19, 20, 37, 46]. Research differences were noted between adaptive and non-adaptive
into efficient scheduling algorithms is essential for transmission technologies [22–26, 28–31, 36]. While many
maintaining high performance and ensuring optimal studies focus on wireless communication performance,
Quality of Experience (QoE) during video transmission via there remains a gap in data for certain QoE aspects. This
the HTTP protocol. research evaluates video quality using both technologies in
Key aspects of HTTP-based video transmission include a centralized wireless HTTP transmission model, where
analyzing technologies and assessing user QoE, which is subscribers connect via radio channels. The model includes
vital for evaluating playback satisfaction. Models and local components such as a video server, buffering, and
criteria for evaluating video quality perception are playback processes, accounting for data transfer speeds and
developed based on this analysis. A common QoE clip selection [15, 34–37].
evaluation method is the Mean Opinion Score (MOS), which For Bluetooth video streaming with adaptive
rates user satisfaction on a 1–5 scale, depending on video modulation and a three-zone buffer, the system operates in
playback statistics and whether the transmission is adaptive equal-duration periods (slots), optimizing the allocation of
or non-adaptive [5–8, 11, 17, 18]. frequency-time resources. A wireless channel model and
In the analysis of MOS functions and their resource scheduler at the base station play critical roles in
approximations, two key factors affecting video quality resource allocation, directly influencing data transfer speeds
perception were identified: buffering during playback, and user satisfaction [15, 20–24, 28, 35, 36]. A software-
which influences both adaptive and non-adaptive hardware setup, including a video player that sequentially
technologies, and the average bitrate, which plays a crucial downloads segments from the server, is also key to securing
role in adaptive technologies. These factors directly impact video streams (Fig. 1). Below are the main assumptions used
user satisfaction and overall video transmission quality. in these models.
User satisfaction with viewing and, consequently, the
performance of the telecommunications system is inversely
Figure 1: Video Data Transmission Model in a Wireless Network
Source: Developed by the author in the LibreOffice environment
Each video data segment 𝑗 is presented in a continuous The user is considered active at time 𝑡 if they are
range of bit rates: downloading at that moment; otherwise, the user is
𝑅 , ∈ [𝑅 , 𝑅 ], 𝑖 = 1, 𝑁, (9) considered inactive.
The user behavior model is characterized by the user In the wireless communication channel, signal
video stream sparsity coefficient 𝑖—which is the ratio of the attenuation during propagation occurs uniformly across the
total durations of viewing and pauses to the viewing entire bandwidth for a specific user at a particular time. Let
duration of user 𝑖 over a given time interval 𝑇 ⟶ ∞: us introduce the variable 𝐶 (𝑇), which equals the data
transmission speed through the wireless channel if all
𝜔 +𝑏
𝛾 = lim (10) available resources were allocated to the user 𝑖 at time 𝑡.
⟶ 𝜔
132
This is called the maximum achievable channel speed. The The presented analytical model describes the key
assumptions used for the wireless channel model are: components and parameters of a real video data
During the transmission of one packet 𝑘 from segment transmission system over the HTTP protocol.
𝑗 by user 𝑖 in the downlink, the maximum achievable For all possible scheduling and video adaptation
wireless channel speed is constant: algorithms that meet the assumptions, the following
𝐶 (𝑇) = 𝐶 , , , 𝑡 , , ≤ 𝑡 ≤ 𝑡 , , + ∆𝑡 , , , (11) inequality holds:
where 𝑡 , , is the moment when user 𝑗, ∆𝑡 , , is the duration 𝐸[𝑅 ] 𝐸 𝐶
(1 − 𝜈 𝜈 ) ≤ 1, (18)
of downloading packet 𝑘 by user 𝑖 from segment 𝑗, 𝐶 , , is 𝑞 +𝛾
being the maximum achievable channel speed during the The relationship between all parameters of the video
packet download. data transmission network is reproduced by dependencies
The sequence of random variables: on the features of the user behavior model (𝛾 , 𝑞 ) video
𝐶 , , 𝐶 , , … , 𝑖 = 1, 𝑁, (12) stream properties (𝐸[𝑅 ]) and the operation of the wireless
where 𝐶 , = ∑ , , forms an ergodic random channel (𝐸 𝐶 ) for each network participant.
, , ,
The radio channel resource allocation algorithms for
process with finite mathematical expectations 𝐸[𝐶 ] and adaptive video streams, which are expressed by the ratio of
variation coefficients 𝜈 С . buffering duration to viewing duration 𝑞 , while
At each moment 𝑡, the scheduling algorithm distributes considering the average bit rate of the viewed stream, are of
portions of the channel resources 𝛼 (𝑡) for all users: great importance. It is worth noting that due to the presence
𝐴(𝑡) = {𝛼 (𝑡), 𝑖 = 1, 𝑁} (13) of video stream adaptation to wireless channel conditions,
An evident constraint on the operation of the the scheduling and video adaptation algorithms, denoted as
scheduling algorithm is the finite volume of available 𝐴 and 𝐵, вrespectively, influence the buffering factor.
resources: According to the research on quality criteria conducted in
the first section, it is proposed to evaluate system
∀𝑡: 𝛼 (𝑡) ≤ 1 (14) performance using two criteria. The average value of the
ratio of buffering duration to viewing duration:
To optimize resource allocation in a Bluetooth wireless
1
network with increased data transfer rates using adaptive 𝑞 (𝐴, 𝐵) = ( 𝑞 (𝐴, 𝐵)), (19)
modulation with a three-zone buffer, the scheduler has 𝑁
access to prior data, including portions of allocated channel The average bitrate of the viewed video stream, which
resources, maximum achievable channel speeds, and the is a key parameter:
volume of transmitted data for each individual user: 1
𝐴(𝑡) = 𝒜(𝛼 (𝜏), 𝐶 (𝜏); 𝜏 < 𝑡, 𝑖 = 1, 𝑁), (15) 𝑅(𝐴, 𝐵) = ( 𝐸 [𝑅 (𝐴, 𝐵)]), (20)
𝑁
where 𝐴 (∙) is represents the scheduling algorithm. The primary goal is to determine the lower bound for all
To ensure the security of video streams over a Bluetooth
possible scheduling and video adaptation algorithms that
wireless network with increased data transfer rates using
meet the conditions introduced in the second section for the
adaptive modulation with a three-zone buffer, the following
average ratio of buffering duration to viewing duration for
assumptions are considered for the scheduling algorithm
all users in the system, provided that the average bitrate of
[10, 17, 46]:
the viewed stream for all users is not less than the given
The scheduler allocates all available channel resources
value 𝑅 .
at every moment.
𝑄= inf 𝑞 (𝐴, 𝐵) (21)
The scheduler does not allocate resources to inactive , : ( , ) , ∈𝒜, ∈ℬ
users. The task of finding the lower bound of the quality of
Each active user is guaranteed a minimum portion of the experience (QoE) criterion 𝑄 is formulated as an
channel resources. optimization problem aimed at minimizing this criterion
When a new video data segment is requested by the under certain constraints and conditions. The objective is to
video player, the problem of choosing the bit rate for the minimize 𝑄 = ∑ 𝑞 subject to:
requested segment is solved by the following expression:
𝑅 , = ℬ(𝑅 , , 𝐶 (𝜏), 𝛼 (𝜏); 𝑘 < 𝑗, 𝜏 < 𝑡 , ), (16) ⎧ 𝑅𝐶
1−𝜈 𝜈 − 1 ≤ 0,
where 𝐵 (∙) is represents the algorithm for calculating the ⎪ 𝑞 +𝛾
⎪
⎪
bit rate of segment 𝑗 for user 𝑖, 𝑅 , is the bit rate of segment 1
− 𝑅 +𝑅 ≤0 (22)
𝑗 for user 𝑖, 𝑡 , is the time when user 𝑖 requests segment 𝑗. ⎨ 𝑁
The frequency of switching bit rates is limited: ⎪
⎪ 𝑅 ∈ [𝑅 , 𝑅 ], 𝑖 = 1, 𝑁
𝜎[𝑅 ] ⎪
∀𝑖: ≤𝜈 , (17) ⎩ − 𝑞 ≤ 0, 𝑖 = 1, 𝑁
𝐸[𝑅 ]
where 𝐸[𝑅 ] = 𝑙𝑖𝑚 ⟶ 𝐸 𝑅 , , 𝜎[𝑅 ] = 𝑙𝑖𝑚 ⟶∞ 𝐸 𝑅 , . 𝑅 = 𝐸 [𝑅 ] та 𝐶 = 𝐸[𝐶 ], (23)
The sequences 𝑅 , , 𝑅 , , … , 𝑖 = 1, 𝑁 form ergodic In this context, the optimization problem is non-linear
with general constraints, making it a non-convex problem
random processes with finite mathematical expectations
with no standard solutions [9–12, 15, 16, 21–27, 34, 35]. To
𝐸[𝑅 ] and variation coefficients that do not exceed the
solve this, a two-stage optimization approach is proposed.
values of 𝜈 respectively.
133
First, an intermediate optimization problem is introduced, ensuring stable video streams even in challenging
where a solution algorithm is already known. Then, the conditions [3, 7, 8, 12, 15, 16]. This approach enhances both
relationship between the main optimization problem and security and efficiency, meeting modern demands for
the intermediate one is established, followed by a relaxation Bluetooth video streaming.
of constraints to solve the original problem. In the transmission buffer, Bluetooth packets are
This approach involves solving the intermediate prioritized and transmitted using one of two modulation
problem first, and then formulating the final problem with schemes based on their priority. Low-priority packets are
relaxed constraints. A new algorithm is proposed to solve sent at 3 Mbps to reduce errors. The Bluetooth ARQ
this, comparing the performance of existing heuristic (Automatic Repeat Request) mechanism, with adjustable
algorithms for wireless channel resource allocation based flush timeouts, prevents packet delays that could lead to
on the QoE criterion QQQ. The lower bound for adaptive frame skips. The flush timeout is set at 1250 microseconds,
video streams is derived and compared with heuristics for equivalent to two Bluetooth timeslots, after which
non-adaptive streams. Simulation modeling of heuristic handshake packets stop [19–21]. Non-flushable settings
scheduling algorithms for adaptive video transmission was protect against data loss in other streams [14, 19, 26, 27, 34,
performed with a fixed number of users per cell, while non- 46].
adaptive video transmission was simulated assuming the As technologies evolve, secure transmission and data
video stream bit rate viewed by all users equals 𝑅 . confidentiality are increasingly vital. Modeling methods,
In summary, the challenges of securing video streams such as the Gilbert-Elliott ergodic chain and Gilbert’s
over a Bluetooth wireless network with increased data rates Markov chains, improve understanding of wireless channel
using adaptive modulation and a three-zone buffer were errors and support effective protection strategies [30–32].
addressed [7–10, 17, 18]. A two-stage optimization approach Adaptive methods like adaptive frequency hopping (AFH)
was proposed: first, an intermediate optimization problem in Bluetooth ensure stable transmissions by avoiding
with a known solution was introduced, followed by interference.
establishing its relationship to the main problem. Finally, an End-to-end encryption (E2EE) is also becoming more
optimization problem with relaxed constraints was common, offering robust protection against unauthorized
formulated and solved [3, 7, 24–27, 34, 35]. access throughout the transmission process [20–25, 27, 31].
An algorithm was developed to solve this optimization Currently, the modeling of the AWGN (Additive White
problem, and its performance was compared with existing Gaussian Noise) channel with a bit error rate (BER) of 10
wireless channel resource allocation heuristics. The study at higher data transfer rates of 3 Mbps and an 𝐸 /𝑁 (energy
derived a bound for adaptive video streams and compared it per bit to noise power spectral density ratio of 16 dB is
with non-adaptive heuristics. Simulation modeling of widely used. This method allows for analyzing the response
heuristic scheduling algorithms for both adaptive and non- of various protection schemes to different channel
adaptive video was conducted, assuming a fixed number of conditions. The Gilbert-Elliott ergodic chain with two states
users per cell. For non-adaptive video, the modeling in discrete time and Gilbert’s Markov chain are applied to
assumed all users viewed video streams at the average model the error characteristics of a wireless channel. Even
bitrate 𝑅 . with the use of modern technologies such as adaptive
The study tackled the challenges of securing video frequency hopping (AFH) in Bluetooth version 5.2, the
streaming over Bluetooth networks using adaptive Gilbert-Elliott model remains an effective tool for channel
modulation with a three-zone buffer. A nonlinear analysis, considering its limitations and impact on
optimization problem was formulated and solved using a audio/video applications. Studying the average durations of
two-stage approach [12–17, 19]. The proposed algorithm good and bad channel states 𝑇 and 𝑇 , helps to understand
effectively enhanced video stream security. A comparative the system’s behavior in real-world operating conditions,
analysis of resource allocation methods demonstrated the while setting the timeout at the appropriate level allows for
benefits of adaptive modulation over non-adaptive effective data flow management to ensure the security and
approaches. As a result, the developed radio channel reliability of information transmission. The average
resource allocation algorithms proved effective in ensuring duration of a good state in seconds, and the duration of a
both security and efficiency for video streaming over bad state is 𝑇 = 0,25 𝑠𝑒𝑐𝑜𝑛𝑑𝑠. In Bluetooth time slot units,
Bluetooth networks with increased data rates. where each time slot lasts 625 microseconds, 𝑇 = 3200,
Integrated Priority and Information Protection Strategy and 𝑇 = 400. Consequently, if the current state is good (g),
for Multi-Level Interaction Among Users: the probability that the next state will also be good (𝑔),
The integrated strategy for managing priorities and (𝑃𝑔𝑔) is 0,9996875, while the probability that the next state
information protection in Bluetooth wireless networks will be bad (𝑏), (𝑃𝑏𝑏) given the current state is 0.9975.
combines advanced traffic management and security 1 1
technologies. It includes a multi-level interaction model that 𝑇 = , 𝑇 = (24)
1 − 𝑃𝑔𝑔 1 − 𝑃𝑏𝑏
dynamically adjusts bandwidth and priorities based on data
At a speed of 3 Mbps, the bit error rate (BER) in the good
type and importance, ensuring secure transmission through
state is 10 , and in the bad state, it is—10 . The two SNR
priority buffering policies that consider transmission
states of 16.00 and 14.70 dB correspond to varying
context and confidentiality [20–23].
transmission conditions. The first SNR value is ideal for
Recent innovations, such as AI integration, enable real-
comparison with a single-stationary model, while the
time traffic analysis and adaptive decision-making. Machine
second is optimal for 2 Mbps transmission speeds. As SNR
learning optimizes transmission priorities and bandwidth,
worsens below 10 dB, only basic rate-protected packets
134
remain viable. Adaptive User Priority (UP) schemes are dynamically adjusted as data and buffer usage change. The
effective for Enhanced Data Rate (EDR) modes, which build boundary between protected and unprotected P-frames
on earlier concepts like IBM’s BlueHoc and Blueware, with shifts according to historical encoding ratios, adapting in
specifications covering baseband, L2CAP, and multi-slot real-time.
packet transmission [28–33, 44, 45]. Clock drift is considered In zone 3, when the buffer is full, B-frames and P-frames
in timing and scheduling, although real-world lose protection, but I-frames retain it, like in zone 1, using
implementations may experience slower EDR mode random number generation and fraction comparison. The
switching. UP policy is linear in zones 1 and 3 but remains nonlinear
Changes in priority buffering policies align with for P-frames, balancing content importance with buffer fill
advanced traffic management and video security measures. levels. This ensures that P-frame output adjusts toward a
Adaptive buffers improve data prioritization by considering linear mode, compensating for buffer saturation.
factors like data importance, confidentiality, quality of As security and efficiency demands in video
service, and security requirements. Critical data, such as transmission grow, improving buffer priority policies
confidential information, is given higher priority to ensure become crucial [12, 22–24, 32, 37, 44–46]. Modern methods
faster and more secure transmission [20, 23, 28, 29, 31, 32, safeguard data while ensuring quality transmission.
44, 45]. Adaptive buffers, combined with modulation technology,
The integration of adaptive modulation with a three- allow dynamic responses to network conditions, enhancing
zone buffer allows dynamic adjustments to signal changes security and performance in wireless communication.
and transmission quality, optimizing bandwidth and Dynamic frame content regulation for enhancing video
reducing the risk of buffer overflow as data volumes and stream protection efficiency:
noise levels fluctuate. Modern priority buffering policies Dynamic frame content regulation is a key approach for
have become essential for securing video streams over optimizing cybersecurity and information protection in
Bluetooth networks, balancing transmission efficiency, data video sequences by managing frame sizes based on
protection, and user satisfaction. parameters like brightness, motion, and texture [23–27, 30–
These policies prioritize data based on buffer fill levels. 32, 34, 35, 44, 45]. Real-time adjustments, such as resolution
In zone 1, where the buffer is less full, protection strategies changes or compression, are made to ensure efficient video
like random number generation are used [8–10]. In zone 3, transmission.
when the buffer is nearly full, urgent data is transmitted Modern systems analyze brightness, motion, texture,
immediately, with reduced protection to maintain speed. and context, allowing quick responses to video stream
This approach optimizes security while managing data flow alterations. Advanced algorithms adapt transmission
and resource use [1–6, 11, 17, 18]. Given the rise of cyber parameters, reducing resolution or applying compression
threats, such policies are crucial for ensuring strong data during high motion to decrease data volume while
protection in dynamic network environments. maintaining essential details. This improves network
In zone 1 of the buffer, all Bluetooth packets of type I- efficiency and prevents unauthorized access [13].
or P-frames are automatically protected by sending them at Dynamic regulation is especially important in Bluetooth
a lower data transmission rate. B-frame packets are only wireless networks using adaptive modulation and a three-
protected in zone 1 if they pass the following test. A zone buffer, optimizing bandwidth and reducing data
comparison is made between a uniformly distributed leakage during intense video changes to balance protection
random number generated in the interval [0,1] and the and efficiency [2, 7, 9]. Adjusting the ratio between buffer
fraction 𝑓, which determines the buffer’s occupancy with zones enhances system resilience. Spatial content affects I-
packets relative to the zone’s bandwidth (Fig. 2). If the frame size and transmission speed, while temporal content
random number is greater than 𝑓, the B-frame packet is also impacts B- and P-frame processing, crucial for
transmitted at a reduced transmission speed, indicating its confidentiality. Research [36] evaluates spatial and temporal
protection. This test is implemented so that the number of information using brightness filtering and the Sobel
protected B-frame packets increases linearly with the algorithm to strengthen video data protection (Fig. 3).
buffer’s fill level in zone 1.
Figure 2: Temporal Change of Spatial Information
Source: Aggregated from sources [8–13, 15, 16, 22–27, 34, 35]
Figure 3: Temporal information changes over time for the
As the buffer fills and packets enter zone 2, a distinct same sequence
priority policy applies for P-frames. In this zone, I-frames Source: Aggregated from sources [11, 15, 16, 22–27, 34, 35]
remain protected, while B-frames lose their protection. P-
The analysis of the temporal dimension of video data
frames are partially protected based on the ratio of
involves calculating brightness differences between frames
internally encoded macroblocks, with protection levels
135
and computing frame-by-frame standard deviation (SD), Reality, 4th Visual Information Engineering (2007)
which helps assess changes in spatial and temporal 1245–1300.
information. This highlights the need for dynamic buffer [3] W. Wilkowska, et al., Analyzing Technology
zone adjustments to ensure efficient data storage and Acceptance and Perception of Privacy in Ambient
processing, particularly for cybersecurity and data Assisted Living for using Sensor-based Technologies,
protection. PLoS ONE, 17(7) (2022).
Dynamic frame regulation is crucial for optimizing data [4] S. Ye, R. S. Blum, L. J. Cimini Jr., Adaptive Modulation
transmission and security over Bluetooth connections. for Variable-Rate OFDM Systems with Imperfect
Bluetooth frames made up of asynchronous connectionless Channel Information, in: 55th IEEE Vehicular
(ACL) packets, occupy multiple time slots, and their size Technology Conference, 2 (2002) 767–771.
impacts payload. Continuous monitoring of content [5] M. J. Haenssgen, After Access: Inclusion,
characteristics such as brightness, motion, and texture Development, and a More Mobile Internet, Journal of
allows for dynamic adjustments in frame sizes, improving Human Development and Capabilities, 18(1) (2017).
transmission efficiency and security [7, 8, 14, 18, 25, 46]. [6] S. W. Campbell, From Frontier to Field: Old and New
Packet size quantization often increases Bluetooth Theoretical Directions in Mobile Communication
packet sizes, reducing bandwidth efficiency. Dynamic Studies, Communication Theory, 29(1) (2019) 46–65.
adjustments ensure that when packet sizes don’t match doi: 10.1093/ct/qty021.
content, smaller transmission schemes (e.g., switching from [7] A. Iyer, U. B. Desai, A Comparative Study of Video
3DH5 to 3DH3 or 3DH1) can improve efficiency [3–5, 18, 28, Transfer over Bluetooth and 802.11 Wireless MAC,
34–36]. A key challenge is managing partially filled packets, IEEE Wireless Communications and Networking
which wastes bandwidth. Research suggests forming filled Conference, 3 (2003) 2053–2057.
Bluetooth packets to enhance performance, though this [8] R. Prasad, Book Review: After Access: Inclusion,
could affect noise immunity [1, 2, 26, 28]. Development and a More Mobile Internet,
The CQDDR model adjusts packet types based on Communication and the Public, 2(1) (2017).
channel quality but overlooks content and network [9] C. H. Chia, M. S. Beg, Realizing MPEG-4 Video
congestion. An improved scheme using a three-zone buffer Transmission over Wireless Bluetooth Link via HCI,
effectively prioritizes data transmission by considering IEEE Transactions on Consumer Electronics, 49(4)
these factors [20, 25, 34]. Additionally, Hamming code error (2003) 1028–1034.
correction (DAEC and SEC) enhances transmission [10] O. Kryvoruchko, Y. Kostiuk, A. Desiatko,
reliability by correcting errors, improving both bandwidth Systematization of Signs of Unauthorized Access to
utilization and data protection. Corporate Information based on Application of
Cryptographic Protection Methods, Ukrainian
5. Conclusions Scientific Journal of Information Security, 30(1) (2024)
140–149.
Ensuring secure video streaming over Bluetooth networks [11] K. A. Pearce, Rice Ronald. Digital Divides. From
with enhanced data rates requires adaptive modulation and Access to Activities: Comparing Mobile and Personal
a three-zone buffer. To meet confidentiality, integrity, and Computer Internet Users, J. Commun. 63(4) (2013)
availability requirements, strategies combining data 721–744. doi: 10.1111/jcom.12045.
protection and adaptive resource allocation are essential. [12] R. Razavi, M. Fleury, M. Ghanbari. Power-
Buffer priority policies adjust protection based on buffer fill Constrained Fuzzy Logic Control of Video Streaming
levels, safeguarding I-frames in critical zones and over a Wireless Interconnect, EURASIP Journal on
optimizing resource use. Advances in Signal Processing (2008). doi:
As Bluetooth traffic increases, adaptive buffers, and 10.1155/2008/560749.
modulation maintain video quality. Adjustments based on [13] C. Scheiter, et al., A System for QOS-enabled MPEG-4
frame brightness and buffer size improve stream Video Transmission over Bluetooth for Mobile
management, while Hamming codes enhance reliability. A Applications, International Conference on Multimedia
two-stage optimization, using cryptography and data and Expo (ICME’03), 1 (2003) 789–792. doi:
analysis, ensures secure and efficient streaming, meeting 10.1109/ICME.2003.1221036.
modern cybersecurity standards. [14] S. Tahir, et al., Hybrid Congestion Sharing and Route
Overall, secure video streaming over Bluetooth Repairing Protocol for Bluetooth Networks, Wseas
networks demands an integrated approach that combines Transactions On Computers 20 (2021) 49–55. doi:
encryption, authentication, and quality management to 10.37394/23205.2021.20.6.
ensure data protection. [15] G. R. Reddy, et al., An Efficient Algorithm for
Scheduling in Bluetooth Piconets and Scatternets,
References Wireless Networks, 16(7) (2009) 1799–1816. doi:
[1] W. Wang, et al., Mobile Node Design of Indoor 10.1007/s11276-009-0229-3.
Positioning System based on Bluetooth and LoRa [16] T. Dave, U. Pandya, Simultaneous Monitoring of
Network, Journal of Physics: Conference Series, Motion ECG of Two Subjects using Bluetooth Piconet
1738(1) (2021). and Baseline Drift, Biomedical Engineering Letters,
[2] R. Razavi, M. Fleury, M. Ghanbari, Low-Delay Video 8(4) (2018) 365–371. doi: 10.1007/s13534-018-0081-4.
Control in a Personal Area Network for Augmented
136
[17] Core Specification of the Bluetooth System, Version [32] A. E. Khalil, et al., Efficient Speaker Identification
2.1+EDR (2007). URL: http://www.Bluetooth.com from Speech Transmitted over Bluetooth Networks,
[18] Q. Li, M. van der Schaar, Providing adaptive QoS to Int. J. Speech Technol. 17(4) (2014).
layered video over wireless local area networks [33] V. Lakhno, et al., Methodology for Placing
through realtime retry limit adaptation, IEEE Components of a Video Surveillance System for Smart
Transactions on Multimedia, 6(2) (2004) 278–290. City Based on a Composite Cost Optimization Model,
[19] Z.-K. Chen, An Adaptive FEC to Protect RoHC and Software Engineering Perspectives in Systems, 501
UDP-Lite H.264 Video Critical Data, Journal of (2022). doi: 10.1007/978-3-031-09070-7_2.
Zhejiang University—Science A, 7(5) (2006). doi: [34] A. B. Nahas, et al., BlueFlood: Concurrent
10.1631/jzus.2006.A0910 Transmissions for Multi-Hop Bluetooth 5—Modeling
[20] C.-H. Yang, S.-J. Lee, Virtual Scatternet Formation for and Evaluation, Comput. Res. Repository, 2(4(22))
Supporting Multicast in Bluetooth Networks, Int. J. (2021) 1–30. doi: 10.1145/346275.
New Technol. Res. (2022). [35] T. Chi, M. Chen, A Frequency Hopping Method for
[21] L.-J. Chen, et al., Audio Streaming over Bluetooth: an Spatial RFID/WiFi/Bluetooth Scheduling in
Adaptive ARQ Timeout Approach, in: 24th Agricultural IoT, Wireless Networks, 25(2) (2017) 805–
International Conference on Distributed Computing 817. doi: 10.1007/s11276-017-1593-z.
Systems, 24 (2004) 196–201. [36] K. P. Rajesh, Fuzzy Logic Controller for Wireless
[22] C. Ru, et al., A New UEP Scheme for Robust Video Video Transmission, J. Comput. Sci. 7(7) (2011) 1119–
Transmission in MIMO System, China 1127. doi: 10.3844/jcssp.2011.1119.1127.
Communications, 4(5) (2006) 102–108. [37] V. Sokolov, et al., Method for Increasing the Various
[23] S. Li, Y. Lou, B. Liu, Bluetooth Aided Mobile Phone Sources Data Consistency for IoT Sensors, in: IEEE 9th
Localization, ACM Transactions on Embedded International Conference on Problems of
Computing Systems (TECS), 13(4) (2014) 1–15. doi: Infocommunications, Science and Technology (2023)
10.1145/2560018. 522–526. doi: 10.1109/PICST57299.2022.10238518.
[24] L.-J. Chen, H.-H. Hung, A Two-State Markov-Based [38] V. Sokolov, P. Skladannyi, N. Mazur, Wi-Fi Repeater
Wireless Error Model for Bluetooth Networks, Influence on Wireless Access, in: IEEE 5th
Wireless Personal Communications, 58(4) (2009) 657– International Conference on Advanced Information
668. doi: 10.1007/s11277-009-9899-5. and Communication Technologies (2023) 33–36. doi:
[25] R. Razavi, M. Fleury, M. Ghanbari, Deadline-Aware 10.1109/AICT61584.2023.10452421.
Video Delivery in a Disrupted Bluetooth Network, [39] V. Sokolov, P. Skladannyi, V. Astapenya, Wi-Fi
IEEE Sarnoff Symposium (2007). doi: Interference Resistance to Jamming Attack, in: IEEE
10.1109/SARNOF.2007.4567352. 5th International Conference on Advanced
[26] Y. Kostiuk, Y. Konstantinov, Improved Security Information and Communication Technologies (2023)
Methods in 4G Networks to Provide Effective 1–4. doi: 10.1109/AICT61584.2023.10452687.
Protection Against Data Transmission Attacks, (Series [40] V. Sokolov, P. Skladannyi, N. Korshun, ZigBee
“Pedagogy”, Series “Law”, Series “Economics”, Series Network Resistance to Jamming Attacks, in: IEEE 6th
“Physical and Mathematical Sciences”, Series International Conference on Information and
“Technology”): Journal, 6(34) (2024). Telecommunication Technologies and Radio
[27] S. Tahir, A Self-organizing Location and Mobility- Electronics (2023) 161–165. doi: 10.1109/
Aware Route Optimization Protocol for Bluetooth UkrMiCo61577.2023.10380360.
Wireless, Int. J. Adv. Comput. Sci. Appl. (2016). doi: [41] V. Sokolov, P. Skladannyi, A. Platonenko, Jump-Stay
10.14569/IJACSA.2016.070631. Jamming Attack on Wi-Fi Systems, in: IEEE 18th
[28] C.-M. Chen, C.-W. Lin, Y.-C. Chen, Packet Scheduling International Conference on Computer Science and
for Video Streaming over Wireless with Content- Information Technologies (2023) 1–5. doi:
Aware Packet Retry Limit, in: 8th IEEE Workshop on 10.1109/CSIT61576.2023.10324031.
Multimedia Signal Processing (MMSP’06) (2006) 409– [42] V. Sokolov, P. Skladannyi, V. Astapenya, Bluetooth
414. Low-Energy Beacon Resistance to Jamming Attack,
[29] N. Golmie, N. Chevrollier, O. Rebala, Bluetooth and in: IEEE 13th International Conference on Electronics
WLAN Coexistence: Challenges and Solutions, IEEE and Information Technologies (2023) 270–274. doi:
Wireless Communications, 10(6) (2003) 22–29. doi: 10.1109/ELIT61488.2023.10310815.
10.1109/MWC.2003.1265849 [43] V. Sokolov, P. Skladannyi, A. Platonenko, Video
[30] R. Razavi, et al., An Efficient Packetization Scheme for Channel Suppression Method of Unmanned Aerial
Bluetooth Video Transmission, Electronic Letters, Vehicles, in: IEEE 41st International Conference on
42(20) (2006) 1143–1145. Electronics and Nanotechnology (2022) 473–477. doi:
[31] J. H. Yoon, S.-B. Lee, S.-C. Park, Packet and 10.1109/ELNANO54667.2022.9927105.
Modulation Type Selection Scheme based on Channel [44] S. K. Mohsin, M. A. Mohammed, H. M. Yassien,
Quality Estimation for Bluetooth Evolution Systems, Developing of Bluetooth Mesh Flooding between
IEEE Wireless Communications and Networking Source-Destination Linking of Nodes in Wireless
Conference (WCNC ’04), 2 (2004) 1014–1017. Sensor Networks, Eastern-European Journal of
Enterprise Technologies 6(9(114)) (2021). :
10.15587/1729-4061.2021.248978.
137
[45] R. Razavi, M. Fleury, M. Ghanbari, Detecting
Congestion within a Bluetooth Piconet: Video
Streaming Response, London Communications
Symposium (2006) 181–184.
[46] S. Rzaieva, et al., Methods of Modeling Database
System Security, in: Cybersecurity Providing in
Information and Telecommunication Systems, vol.
3654 (2024) 384–390.
138