A
Bandwidth-Optimal Group-Injected Data Filtering and ElGamal-Based
Authentication Framework for Secure Wireless Sensor Networks
Priyanka
Soni1*, Dr. Rajeev Yadav2
1 Research
Scholar, Shri Krishna University, Chhatarpur, M.P., India
baghelrashmi805@
2 Professor,
Shri Krishna University, Chhatarpur, M.P., India
Abstract: Wireless sensor networks (WSNS) place a
premium on efficient use of resources and safety. Particularly in multipath sensor
network routing, studies have concentrated on reducing resource consumption
while increasing security. False information is a major problem since it makes
it harder to aggregate and authenticate data, which in turn increases the need
for bandwidth and battery life. In response, we have built a bandwidth-optimal
group-injected data filtering system to improve network performance by
decreasing packet-transferred malinformation. In addition, a framework for
network authentication has been put in place to strengthen protection against
these types of injections. In addition to promoting improved security in
multipath data transmission, this also guarantees safer data transfer between
nodes and efficiently controls resource utilization. The bandwidth efficient
approach employs a data aggregation algorithm to identify malicious information
injections, with the goal of lowering energy consumption and network overheads
in WSN. To improve the security of wireless sensor networks (WSNS) and reduce
the impact of group-injected misinformation, this study presents the BO-GIDF
architectural framework. In order to reduce bandwidth consumption, the
framework finds all of the neighbors who are spreading false information using
a collaborative neighbor selection method that is based on groups. Moreover, it
uses an associative filtering approach to save processing time and enhance
packet throughput through efficient sink identification. The authors also
provide a system called network authenticated based on Elgamal cryptography
(NAEC) to improve network privacy and security by strengthening authentication
against malinformation injection using asymmetric key encryption. By
identifying and isolating hostile actors, the NAEC framework speeds up
processing and guarantees high interoperability security via an implicit
approved certificate rule that prevents malinformation injection on dynamic
pathways.
Keyword: Malinformation, BO-GIDF, NAEC, WSN
1. INTRODUCTION
Wireless Sensor Networks
(WSNs) represent a specialized class of distributed networks composed of a
large number of low-power sensor nodes that cooperatively monitor physical or
environmental conditions such as temperature, pressure, humidity, vibration, or
motion. These sensor nodes are typically deployed in large-scale, often
unattended environments and communicate sensed information to a sink or base
station through multi-hop wireless transmission. Due to their ability to
operate autonomously and provide real-time data from inaccessible or hazardous
environments, WSNs have become an essential enabling technology for
applications including environmental monitoring, healthcare systems, industrial
automation, military surveillance, disaster management, and smart
infrastructure. A defining characteristic of WSNs lies in their severe resource
constraints. Sensor nodes are equipped with limited computational capability,
restricted memory, constrained communication bandwidth, and finite battery
energy that is often non-replaceable once deployed. Unlike conventional
wireless networks, where devices benefit from continuous power supply and high
processing capacity, WSN nodes must carefully balance sensing accuracy,
communication efficiency, and network lifetime. As communication activities
consume the largest proportion of energy in sensor nodes, minimizing
unnecessary transmissions is critical for sustaining long-term network
operation. These constraints profoundly influence network design choices and
significantly complicate the implementation of robust security mechanisms. Another
fundamental attribute of WSNs is their cooperative and data-centric
communication paradigm. Rather than supporting end-user communication, WSNs
focus on collecting, aggregating, and transmitting sensor-generated data to a
central sink. Intermediate nodes participate in routing and aggregation
processes, forwarding data generated by other nodes without having complete
knowledge of the data’s origin. While this collaborative architecture enhances
scalability and efficiency, it also introduces significant security
vulnerabilities, as compromised nodes can actively participate in network
operations and inject malicious content without raising immediate suspicion.
Security concerns in WSNs are considerably more complex than those in
traditional networks. The open wireless medium exposes transmitted data to
interception, replay, and manipulation attacks, while the unattended nature of
sensor deployments makes physical capture and compromise of nodes a realistic
threat. Once compromised, a sensor node becomes an insider attacker capable of
generating syntactically valid and semantically plausible data that can
propagate through the network undetected. Such attacks directly undermine the
integrity and trustworthiness of sensed information, which is particularly dangerous
in mission-critical applications where decisions rely heavily on accurate
sensor data.
Among the various
security threats targeting WSNs, malinformation injection has emerged as one of
the most damaging and difficult to detect. In this context, malinformation
refers to deliberately manipulated or fabricated sensor data introduced by
compromised nodes with the intent to mislead data aggregation, disrupt network
performance, or exhaust critical network resources. Unlike accidental faults or
environmental noise, malinformation is intentional, adaptive, and often
designed to mimic legitimate sensing behavior. As a result, traditional
validation methods based on fixed thresholds or statistical outlier detection
frequently fail to identify such malicious data. The challenge is further
exacerbated in multipath routing environments, which are commonly used in WSNs
to enhance reliability and fault tolerance. While multipath routing allows data
to reach the sink through multiple alternative routes, it also increases the
attack surface by enabling malinformation to propagate simultaneously across
different paths. Consequently, false or manipulated data may be repeatedly
forwarded, processed, and aggregated, leading to excessive bandwidth
consumption, accelerated energy depletion, increased transmission delays, and
premature network failure. In such scenarios, malinformation attacks function
not only as data integrity threats but also as resource exhaustion mechanisms. Conventional
security solutions for WSNs predominantly rely on cryptographic authentication,
probabilistic filtering, or node-level trust evaluation. Although cryptographic
techniques provide confidentiality and authentication, they are often
computationally expensive and insufficient against insider attackers who
possess valid credentials. Similarly, probabilistic and node-centric filtering
approaches tend to perform poorly when malicious nodes operate in coordinated
groups. Coordinated or group-injected malinformation exploits spatial and
temporal correlations among compromised nodes, enabling attackers to evade
detection mechanisms that assume independent or randomly distributed faults.
These limitations highlight the urgent need for security strategies that move
beyond isolated node verification and incorporate collaborative, behavior-aware
analysis.
2. REVIEW LITERATURE
Haibo Wu (2023) This research offers a distributed
set-membership filtering method utilizing a trust dynamic combination technique
for target tracking issues in wireless sensor networks affected by malicious
assaults. The algorithm employs a prediction-correction recursive updating
framework akin to Kalman filtering. It incorporates a clustering fusion phase
of data received from other nodes between the prediction and measurement correction
update phases. This clustering fusion phase utilizes K-means to categorize data
from trusted and untrusted nodes, with the target state being updated through
the amalgamation of the trusted data set, thereby enhancing resilience against
diverse malicious network attacks. Simulation findings indicate that, in
comparison to the conventional distributed set-membership filtering approach,
the suggested technique has superior target tracking efficacy against severe
network assaults, including random attacks, fake data injection, replay
attacks, and hybrid attacks.
Michael Hooper (2016) asserts that commercially available
Wi-Fi-based unmanned aerial vehicles (UAVs) are susceptible to fundamental
security breaches, executable by novice to intermediate hackers. This is
achieved by illustrating that the conventional ARDiscovery Connection procedure
and the Wi-Fi access point utilized in the Parrot Bebop UAV are vulnerable,
enabling a remote assailant to interrupt the UAV's flying capabilities during
operation. We assert that these vulnerabilities are pervasive in
Wi-Fi-dependent Parrot UAVs. Our methodology monitored the standard operation
(i.e., ARDiscovery Connection process via Wi-Fi) of the Parrot Bebop UAV.
Subsequently, we employed a fuzzing technique to ascertain that the Parrot
Bebop UAV is susceptible to fundamental denial of service (DoS) and buffer
overflow attacks during its ARDiscovery Connection process. The exploitation of
these vulnerabilities may lead to the catastrophic and instantaneous incapacitation
of the UAV's rotors during flight. Furthermore, we identified that the Parrot
Bebop UAV is susceptible to a fundamental ARP (Address Resolution Protocol)
Cache Poisoning attack, which may sever the connection with the primary mobile
device user and, in most instances, compel the UAV to land or return to its
origin.
Althaf Marsoof (2022) asserts that online service
providers and governments have progressively depended on Artificial
Intelligence ('AI') to manage internet content. In many areas, the law has
encouraged, if not mandated, service providers to implement mechanisms for
detecting, tracking, and eliminating problematic information, including
terrorist propaganda. As a result, service providers are compelled to employ AI
for the moderation of online material. Nonetheless, content-filtering AI
systems have constraints that impact their precision and clarity. These
constraints provide the potential for valid information to be eliminated while
bad stuff persists online. This conclusion might jeopardize human well-being
and the fulfillment of our human rights. Given these problems, we contend that
the design and implementation of content-filtering AI systems necessitate
regulation.
Hongyang Du (2023) states that Generative AI (GAI)
models are progressing swiftly, including diverse applications such as
intelligent networks and mobile AI-generated content (AIGC) services. Despite
their myriad applications and promise, such models present prospects for new
security issues. This study analyzes the problems and possibilities presented
by Generative Artificial Intelligence (GAI) in the security of intelligent
network AIGC services, including the formulation of security rules and its dual
role as both a "spear" for prospective assaults and a "shield"
inside various defence systems. Initially, we provide a thorough examination of
the GAI ecosystem, emphasizing its applications and the methodologies that
support these innovations, particularly big language and diffusion models.
Subsequently, we examine the dynamic interaction between GAI's offensive and
defensive functions, emphasizing two principal kinds of possible GAI-related
attacks and their corresponding defence tactics inside wireless networks.
Shiyou Xu (2022) The rapid progress in deepfake
generation technology has jeopardized the credibility of digital media, posing
significant concerns to privacy and security through the creation of deepfake
videos. This study hypothesizes that the integration of sophisticated
algorithms with blockchain technology would significantly improve the accuracy
and security of deepfake detection systems. A unique framework for identifying
deepfake combinations was developed by integrating DDO-AGNN with blockchain
technology for federated learning. The current model utilizes a comprehensive
dataset that encompasses several deepfake films, which were standardized by
min-max normalization during pre-processing. The DDO-AGNN method was executed
in Python and enhanced by DDO for superior feature extraction and
classification. Blockchain technology was utilized for federated learning,
guaranteeing privacy-preserving collaborative model training across several
nodes.
3. RESEARCH
METHODOLOGY
This study proposes an
integrated security framework Bandwidth-Optimal Group-Injected Data Filtering
(BO-GIDF), Network Authentication based on ElGamal Cryptography (NAEC). The
BO-GIDF component employs a collaborative group-based neighbor identification
strategy to detect and isolate nodes involved in malinformation injection
before data aggregation occurs. By filtering malicious packets at an early
stage, the framework significantly reduces unnecessary bandwidth usage and
improves data transmission efficiency. The NAEC module strengthens network
authentication through a lightweight asymmetric encryption mechanism based on
ElGamal cryptography. This approach enhances security while minimizing
computational overhead, enabling faster authentication and improved packet
throughput. The design of the Bandwidth-Optimal Group-Injected Data Filtering
(BO-GIDF) framework is guided by four fundamental principles derived from the
identified threat model and network constraints. First, group-based detection
is prioritized over individual node verification. Since malinformation attacks
often exhibit coordinated behavior, BO-GIDF evaluates data consistency at the
neighbor-group level, enabling early identification of collusive malicious
activity that would otherwise evade node-centric detection. To improve the
authentication strategy on sensor networks and prevent Malinformation data
injection in WSN, an effective framework is presented, which is named Network
Authenticated based on Elgamal Cryptography (NAEC). Furthermore, the ElGamal
encryption system is an asymmetric key encryption technique that was developed
to prevent fake data insertion utilizing the NAEC architecture. From this, it
is clear that the encrypted data is effectively blocking any malicious actors
from injecting fake data into data packets used for communication between
sensor nodes. This, in turn, uses the Diffie-Hellman key exchange paradigm to
identify the malevolent actors in the WSN. Finally, in order to maintain the
network's high level of interoperability with improved efficiency, the Implicit
Authorized Certificate Rule was designed.
4. DATA ANALYSIS
4.1 Securing Wireless Sensor Networks
with Bandwidth Optimal Group Injection Data Filtering
This section compares the proposed Bandwidth Optimal
Group Injected Data Filtering (BO-GIDF) framework with existing methods, such
as the Bandwidth Efficient Cooperative Authentication (BECAN) scheme and the
Data Aggregation and Authentication (DAA) protocol. The former was developed by
Rongxing Lu et al. (2012), while the latter was created by Suat Ozdemir and Hasan
Çam (2010). Results analyze the values in tables and graphs, as well as the
following metrics, and are used to measure the experimental parameters using
the suggested BO-GIDF framework.
Quantification of Data Transfer Rate
According to the BO-GIDF paradigm, the amount of data
transported during a certain time period is the definition of bandwidth usage
in WSN. Bits per second (bps) is the unit of measurement for bandwidth
utilization. The mathematical representation of the bandwidth use is shown
below.
𝐵
= 𝑁𝑢𝑚𝑏𝑒𝑟
𝑜𝑓
𝑛𝑜𝑑𝑒𝑠
∗
𝐷𝑎𝑡𝑎
𝑇𝑟𝑎𝑛𝑠𝑚𝑖𝑡𝑡𝑒𝑑
𝑇𝑖𝑚𝑒(𝑚𝑠) (1)
The bandwidth usage, denoted as 'B' in equation (1),
is assessed with regard to the number of nodes and data transmission with
respect to the given time. If the amount of bandwidth used is reduced, then the
approach is deemed more efficient.
Table 1 Data Table for Bandwidth Usage
|
No. of nodes |
Bandwidth Consumption (bps) |
||
|
Proposed BO-GIDF |
Existing BECAN |
Existing DAA |
|
|
10 |
810 |
965 |
1285 |
|
20 |
1585 |
1853 |
2381 |
|
30 |
2335 |
2850 |
3541 |
|
40 |
3045 |
3700 |
4713 |
|
50 |
3952 |
4680 |
5916 |
|
60 |
4732 |
5421 |
7032 |
|
70 |
5532 |
6300 |
8000 |
|
80 |
6200 |
7192 |
9120 |
|
90 |
6987 |
7947 |
10023 |
|
100 |
7643 |
8963 |
10893 |
Table 1 shows
the current systems, including the BECAN scheme by Rongxing Lu et al. (2012)
and the DAA protocol by Suat Ozdemir and Hasan Çam (2010), as well as the
proposed BO-GIDF framework and its relationship to the number of nodes in a
WSN. The experimental goal involves varying the number of nodes from 10 to 100.
Table 1 demonstrates that all methods see an increase in bandwidth use as the
number of nodes increases. In contrast to the status quo, the suggested BO-GIDF
structure drastically cuts down on bandwidth use. The proposed BO-GIDF
framework is used to characterize the bandwidth usage in WSN in Figure 1. The
present system, which includes the BECAN scheme by Rongxing Lu et al. (2012)
and the DAA protocol by Suat Ozdemir and Hasan Çam (2010), is compared with
this. Figure 4.1 clearly shows that the suggested BO-GIDF architecture uses
less bandwidth than the current approaches. This is because the sensor node's
mobility may be tracked using data gathered from nearby nodes via the mobile
compromised node, leading to a decrease in bandwidth usage and improved
efficiency.
Figure 1 Consumption of bandwidth quantified
Hence, according
Rongxing Lu et al. (2012) and the suggested BO-GIDF framework reduces bandwidth
usage by 15% compared to the current BECAN scheme and by 33% compared to the
existing DAA protocol, respectively.
Evaluation of the Time Required to Send a Packet
A packet's transmission time is the total amount of
time it takes to send a packet—header included—over a network at a certain
speed. Milliseconds (ms) are the standard units of measurement for packet transmission
time. A mathematical expression for the time it takes for a packet to be sent
is.
(2)
The time it takes for a packet to be sent in a sensor
network is represented by the symbol '𝑃𝑇𝑡𝑖𝑚𝑒' in equation (2). The method becomes
much more efficient if the time it takes to transmit packets is decreased.
Table 2 Data Table for Time of Packet
Transmission
|
Packet
size (bytes) |
Packet
Transmission Time (ms) |
||
|
Proposed
BO-GIDF |
Existing
BECAN |
Existing
DAA |
|
|
400 |
3.5 |
4.2 |
5.2 |
|
800 |
3.9 |
4.6 |
5.4 |
|
1200 |
4.3 |
5.1 |
5.8 |
|
1600 |
4.5 |
5.3 |
6 |
|
2000 |
4 |
4.8 |
5.5 |
|
2400 |
3.7 |
4.5 |
5.1 |
|
2800 |
3.5 |
4.2 |
4.75 |
|
3200 |
3.6 |
4.4 |
4.9 |
|
3600 |
3.8 |
4.6 |
5.1 |
|
4000 |
4 |
4.8 |
5.3 |
Table 2 displays the time it takes for packets to be
sent in WSN using the proposed BO-GIDF framework, in relation to their size.
This is compared to current systems, such as the BECAN scheme and the DAA
protocol. During the testing process, the packet size is adjusted between 400
and 4000. Table 2 shows that when the packet size increases, the transmission
time increases for all techniques. When compared to other techniques, the
suggested BO-GIDF architecture offers superior performance in terms of
decreasing packet transmission time.
Figure 2 Temporary evaluation of data
packet transfer
Figure 2 is a comparison of the current system,
comprising the BECAN scheme and the DAA protocol by with the packet
transmission time utilizing the proposed BO-GIDF framework in WSN. Figure 2
shows that, in comparison to previous techniques, the suggested BO-GIDF
architecture significantly reduced the time it takes for packets to be sent. In
order to reduce the time-efficient sink detection algorithm's workload, it
compares the frequency of time events in WSNs across nearby nodes at regular
intervals. As a result, the packet is optimized for transmission to the sink
node after passing through many routers. Therefore, the suggested BO-GIDF
framework shortens packet transmission time by 17% compared to the current
BECAN scheme (Rongxing Lu et al., 2012) and by 27% compared to the DAA protocol
(Suat Ozdemir, 2010; Hasan Çam, 2010).
4.2 Elgamal Cryptography Is Utilized
for Network Authentication in Wireless Sensor Networks (Wsn)
The effectiveness
of the Network Authenticated based on ElGamal Cryptography (NAEC) framework is
evaluated by contrasting it with current methods, such as the Robust Data
Aggregation method developed by and the Game theoretical approach. After that,
in order to provide better network result analysis, the values of the tables
and graphs are evaluated using the performance of the suggested NAEC framework.
Quantification
of the Injection of Malinformation Data
In order to
assess the NAEC framework's incorrect data injection rate, the packet delivery
ratio is used. The percentage of the number of data packets that successfully
reach their destination nodes from their source nodes is called the
Malinformation data detection rate. The rate of fraudulent data injection is
expressed as a percentage. Here is the mathematical calculation for the
incorrect data injection rate.
(3)
In equation (3),
the variable '𝐹𝐷𝐼' stands for the Malinformation data injection rate, which is determined
by the quantity of data packets '𝐷𝑃'. A more efficient approach is one with a greater Malinformation
injection rate.
Table 3 Table for the Rate of
Malinformation Data Injection
|
No.
of Data Packets |
Malinformation
Data Injection Rate (%) |
||
|
Proposed
NAEC |
Existing
Game Theoretical Approach |
Existing
Robust Data Aggregation |
|
|
10 |
66.39 |
49.81 |
43.67 |
|
20 |
70.65 |
53.22 |
47.43 |
|
30 |
68.23 |
51.18 |
45.32 |
|
40 |
72.41 |
55.15 |
49.16 |
|
50 |
73.55 |
56.47 |
50.25 |
|
60 |
70.25 |
53.82 |
47.57 |
|
70 |
73.88 |
56.82 |
50.88 |
|
80 |
76.43 |
59.42 |
53.25 |
|
90 |
79.57 |
62.65 |
56.65 |
|
100 |
82.34 |
65.42 |
59.41 |
Comparing the suggested NAEC framework in WSN with
current systems, such as the Game theoretical approach and the Robust Data
Aggregation technique Table 3 illustrates the Malinformation data injection
rate as a function of the number of data packets. For the sake of
experimentation, the number of data packets used as input is adjusted between
10 and 100. According to the statistics in the table, the rate of fake data
injection rose for all techniques as the quantity of data packets grew. In
contrast to other approaches, the suggested NAEC framework significantly
improves the data input rate in the evaluation.
Figure 3 Quantification of the
Injection of Malinformation Data
When compared to current systems, such as the Robust
Data Aggregation technique and the Game theoretical approach Figure 3 shows the
Malinformation data injection rate for the proposed NAEC framework in WSN. In
comparison to current approaches, the Malinformation data injection rate is
substantially increased by the proposed NAEC architecture (figure 3). Because
of the multiplicative ELGamal encryption technique, which helps with the secure
distributed storage and transmission of data packets in WSN, the Malinformation
data injection rate is improved. In the NAEC architecture, the asymmetric key
encryption technique keeps the network system running smoothly while data
packets are sent, which improves the rate of Malinformation data injection.
Consequently, after comparing the new NAEC framework to the current Game
theoretical approach and the existing Robust Data Aggregation technique the
Malinformation data injection rate in WSN is improved by 30% and 46%,
respectively.
Measure of Security
Using the NAEC architecture, WSN security measures the
sum of all data packets delivered to the sink node minus the sum of all data
packets that were not received. Percentage is the unit of measurement for the
security. Here is an assessment of the mathematical model of security:
𝑆
= 𝑃𝑎𝑐𝑘𝑒𝑡𝑠𝑠
− 𝑃𝑎𝑐𝑘𝑒𝑡𝑠𝑛𝑟 (4)
The security that is assessed in relation to the data
packets transmitted ('𝑜𝑎𝑑𝑘𝑒𝑡𝑦𝑦') and those that were not received (𝑃𝑎𝑑𝑘𝑒𝑡𝑦𝑛𝑟) at the sink node is denoted by '𝑆' in equation (4). When the level of security in WSN
is increased, the process becomes much more efficient.
Table 4 Security Tabulation
|
No.
of Data Packets |
Security
(%) |
||
|
Proposed
NAEC |
Existing
Game Theoretical Approach |
Existing
Robust Data Aggregation |
|
|
10 |
65 |
50 |
45 |
|
20 |
68 |
53 |
48 |
|
30 |
71 |
56 |
51 |
|
40 |
73 |
58 |
53 |
|
50 |
73 |
58 |
53 |
|
60 |
76 |
61 |
56 |
|
70 |
79 |
64 |
59 |
|
80 |
81 |
66 |
61 |
|
90 |
83 |
68 |
63 |
|
100 |
85 |
70 |
65 |
The proposed NAEC framework and two existing systems,
the Game theoretical and the Robust Data Aggregation technique are shown in
Table 4, which shows the security dependent on the quantity of data packets in
the network. In order to run the test, the number of data packets is changed
from 10 to 100. The security of all the strategies in table 4 is enhanced as
the quantity of data packets increases. In contrast to these preexisting
approaches, the suggested NAEC framework outperforms them all by increasing the
security rate.
Figure 4 Security Measure
Figure 4 describes the security in WSN for proposed
NAEC framework and is compared with existing system including Game theoretical
approach and Robust Data Aggregation method From figure 6, the proposed NAEC
framework is comparatively improves the security rate when compared to existing
methods. The homomorphic mapping function employs encrypted data with private
key structure which easily isolate the malicious adversaries from data packets
in NAEC framework. Such that, the data packets are avoided during data packet
transfer between the source and destination nodes through the sink which
resulting in improves the security in WSN. Hence, the rate of security is
improved using the proposed NAEC framework in the network by 25% when compared
to existing Game theoretical approach by Nicola Basilico et al. (2014) and 37%
when compared to existing Robust Data Aggregation method by Mohsen Rezvani et
al. (2015) respectively.
CONCLUSION
Primarily, a technique called Localized Bandwidth
Optimal Group Injected Data Filtering (BO-GIDF) is suggested for flooding WSNs
with fabricated data. To minimize bandwidth usage in WSNs and detect bogus data
injection, the BO-GIDF approach introduces a Group based Collaborative
Neighbour Selection mechanism. To improve packet transmission and security
while doing efficient packet forwarding, the suggested BO-GIDF technique
employs a time-efficient sink detection mechanism. By identifying groups of
intentionally manipulated data during data packet transmission in WSN, the
associative filtering method helps to reduce processing time. For optimally
filtering out the misleading data, the suggested BO-GIDF approach employs a
verification mechanism. So, the suggested BO-GIDF approach makes WSN networks
more secure. The next step in improving the
authentication strategy in WSN is to propose the Network Authenticated based on
ElGamal Cryptography (NAEC) approach. The ElGamal encryption system prevents
unauthorized access to the network by using an asymmetric key encryption
technique. By using the suggested NAEC approach, the ElGamal encryption system
improves the privacy level while maintaining the network's security. To detect
hostile actors in various forms and cut them off from the network, the
Diffie-Hellman key exchange paradigm is used. The suggested NAEC technique
reduces processing time by preserving the authentication rules on the sensor
network.In order to keep the network's high level of interoperability secure
and to prevent the insertion of fake data on dynamic pathways, the Implicit
Authorized Certificate Rule was also designed. As a result, WSNs are better
protected against fake data injection using the suggested NAEC approach.
References
1.
Wu,
H., Zhu, H., Li, X., & Amuri, M. J. V. (2023). Trust-based distributed
set-membership filtering for target tracking under network attacks. IEEE
Access, 11, 84468-84474.
2.
Hooper,
M., Tian, Y., Zhou, R., Cao, B., Lauf, A. P., Watkins, L., ... & Alexis, W.
(2016, November). Securing commercial WiFi-based UAVs from common security
attacks. In MILCOM 2016-2016 IEEE Military Communications Conference (pp.
1213-1218). IEEE.
3.
Marsoof,
A., Luco, A., Tan, H., & Joty, S. (2023). Content-filtering AI
systems–limitations, challenges and regulatory approaches. Information
& Communications Technology Law, 32(1), 64-101.
4.
Mah,
P. M., Skalna, I., & Pelech-Pilichowski, T. (2022, May). AI-Based
Facial-Age Detection and IoT for Enhanced Data Security in Social Media.
In Proceedings of the AAAI Symposium Series (Vol. 5, No. 1,
pp. 242-249).
5.
Bahja,
M., & Safdar, G. A. (2020). Unlink the link between COVID-19 and 5G
networks: an NLP and SNA based approach. Ieee Access, 8,
209127-209137.
6.
Du,
H., Niyato, D., Kang, J., Xiong, Z., Lam, K. Y., Fang, Y., & Li, Y. (2023).
Spear or shield: Leveraging generative AI to tackle security threats of
intelligent network services. arXiv preprint arXiv:2306.02384.
7.
Xu,
S., Yang, X., & Xie, Z. (2022). The future of misinformation control:
integrating advanced algorithms and Blockchain for effective Deepfake
detection. Peer-to-Peer Networking and Applications, 18(4),
244.
8.
Kudari,
R., & Koduru, S. R. (2021). The study of Mobile cloud computing: Design,
Uses, and Security in Covid 19 time. Int. J. of Aquatic Science, 12(2),
2115-2124.
9.
Gerts,
D., Shelley, C. D., Parikh, N., Pitts, T., Watson Ross, C., Fairchild, G., ...
& Daughton, A. R. (2021). “Thought I’d share first” and other conspiracy
theory tweets from the COVID-19 infodemic: Exploratory study. JMIR
public health and surveillance, 7(4), e26527.
10.
Janavičiūtė,
A., Liutkevičius, A., & Morkevičius, N. (2022). Toward the
Implementation of Text-Based Web Page Classification and Filtering Solution for
Low-Resource Home Routers Using a Machine Learning Approach. Electronics, 14(16),
3280.
11.
Wani,
M. A., ELAffendi, M., Shakil, K. A., Abuhaimed, I. M., Nayyar, A., Hussain, A.,
& Abd El-Latif, A. A. (2023). Toxic fake news detection and classification
for combating COVID-19 misinformation. IEEE Transactions on
Computational Social Systems, 11(4), 5101-5118.
12.
Rojas,
N., Boettcher, N., Játiva, P. P., & Sánchez, I. (2022). Technical
Evaluation of Android Parental Control Tools and PDNS Filtering Services: The
Chilean Context. IEEE Access.
13.
Raza,
S., Qureshi, R., Lotif, M., Chadha, A., Pandya, D., & Emmanouilidis, C.
(2022). Just as humans need vaccines, so do models: Model immunization to
combat falsehoods. arXiv preprint arXiv:2505.17870.
14.
Maulana,
A., & Langguth, J. (2023, November). Using gnns for misinformation spreader
detection via assortativity-aware node label classification in Twitter
networks. In 2023 Tenth International Conference on Social Networks
Analysis, Management and Security (SNAMS) (pp. 1-8). IEEE.
15.
Yang,
H., Li, T., Yan, J., & Elvira, V. (2021). Hierarchical average fusion with
GM-PHD filters against FDI and DoS attacks. IEEE Signal Processing
Letters, 31, 934-938.