Designing a Lightweight, Multi‑Layer Security Framework for IoT

Devices: Adaptive and Scalable Protection against Emerging Threats

in Connected Environments

Ramakrishna Kosuri

, Trupti Dhanadhya

, Girija M. S.

, S. Harthy Ruby Priya

Praveen K.

and M. Srinivasulu

Tata Consultancy Services, Computer consultant, Celina, Texas, 75009, U.S.A.

Department of Electrical Engineering, Dr. D Y Patil Institute of Technology Pimpri, Dr. D. Y. Patil Dnyan Prasad

University, Pune, Maharashtra, India

Department of Computer Science and Design, R.M.K. Engineering College, RSM Nagar, Kavaraipettai, Tamil Nadu, India

Department of Computer Science and Engineering, J.J. College of Engineering and Technology, Tiruchirappalli, Tamil

Nadu, India

Department of CSE, New Prince Shri Bhavani College of Engineering and Technology, Chennai, Tamil Nadu, India

Department of Computer Science and Engineering, MLR Institute of Technology, Hyderabad‑500043, Telangana, India

Keywords: IoT Security, Lightweight Encryption, Intrusion Detection, Scalable Frameworks, Connected Environments.

Abstract: In the era of Internet of Things (IOT) revolutionizing modern lifestyle, protecting the IOT devices present

unique challenge to secure, since the devices have heterogeneous architecture, constrained-resource and ever-

evolving cyber-threats. This work introduces a lightweight, multi-layer security mechanism designed for IoT

environments that combines adaptive cryptography, real-time intrusion detection, and behaviour-based

anomaly detection. In contrast to traditional models which are limited to network or cloud layers only, the

proposed methodology provides complete security from the device to the edge and cloud that is both scalable

and has low overhead. The model is analyzed under various case studies including smart home, health care

and industrial IoT to show that it is more stable against multi-vector attacks, reduces latency and enhances

energy efficiency. This effort also delves into post-quantum cryptographic readiness, providing a future-proof

approach for next generation connected ecosystems.

1 INTRODUCTION

The explosive growth of the Internet of Things (IoT)

has revolutionized the digital world, leading to an

interconnected network of billions of smart devices in

homes, industries, transport and health-care systems.

This hyper-connectivity, in addition to enabling

automation and intelligence, results in a complex

network of security threats. The majority of IoT

devices are resource-constrained, do not follow any

standard protocols, and are generally running

outdated firmware, which makes them attractive to

cybercriminals. classical security models conceived

for high computing power environments do not scale

well in such a context, leaving sensitive data and

critical infrastructures open to live time threats.

Furthermore, the heterogeneity of IoT

ecosystems, which span from low-power sensors to

high-end gateways, requires a security architecture

that can be flexible and context-aware. Most of the

current methods provide partial protections,

concentrating on only one aspect such as the network

or cloud and ignoring the protection at the device, and

edge levels. These constraints demand the need for

overarching frameworks that are agile in response to

ever-changing threat environments with a small

footprint, and can also communicate securely

between heterogeneous platforms.

It leads to an urgent need for resolvement that

how IoT can be effectively protected from cyber

threats of the future, IoTWe propose a new multilayer

security architecture that specifically tailored for IoT

systems, to mitigate those threats. The method uses

low-overhead encryption methods, behavioral

212

Kosuri, R., Dhanadhya, T., S., G. M., Priya, S. H. R., K., P. and Srinivasulu, M.

Designing a Lightweight, Multi-Layer Security Framework for IoT Devices: Adaptive and Scalable Protection against Emerging Threats in Connected Environments.

DOI: 10.5220/0013860600004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 1, pages

212-218

ISBN: 978-989-758-777-1

anomaly detection and edge-integrated intrusion

prevention techniques for end-to-end protection.

What’s more, it is scalable in small and big

installations, including extensive smart

environments. By integrating post-quantum

cryptographic methods and device-level security, this

work offers a future proof mechanism to protect the

connected world.

2 PROBLEM STATEMENT

Rapid adoption of IoT in the most crucial sectors has

escalated the need for hardened security mechanisms

that meet specific complexities posed by connected

devices. However, most available security schemes

are not able to provide full protection because of their

use of the conventional overhead-based and non-

compatible high-resource approaches for low-power

heterogeneous IoT hardware. Moreover, current

models sometimes only consider single parts like the

network or cloud layer but disregard weaknesses in

the device, firmware, or edge. This piecemeal

approach leaves IoT systems vulnerable to a variety

of threats that include unauthorized access, data leaks,

firmware tampering, and advanced persistent threats.

The lack of scalable, light-weight and flexible

security architecture further aggravates the problem,

particularly in environments that require real-time

processing, and cross-domain interoperability.

Therefore, there is a critical necessity of holistic

security framework which operates multi-layered in a

manner ensuring effective and resilient security for

IoT devices, while considering the limitations of their

operational capabilities and evolving threat space.

3 LITERATURE SURVEY

The continued development of IoT technologies has

enabled device interconnectivity to be more readily

available and increasingly powerful, however it has

also uncovered severe security weaknesses at each

level of the stack. Some researchers have stressed the

consideration of security policies according to the

limited computing power that characterizes an IoT

device. Akram, 2024) to get an introduction while

noting that There is currently no deployable

framework based on real-time with very low energy

budget. Al-Ali and Zualkernan (2023) present

extensive overview of threats but their model is less

detailed on the device level.

In, the authors investigated machine learning and

deep learning-based intrusion detection in IoT

network. Bharati and Podder (2022) propose an ML-

based solution but nevertheless ignore the overhead

of resource affecting device performance. Similarly,

Buyya et al. (2024) concentrate on cloud-centric

security, but overlooks security holes at the edge and

endpoint levels. Compunnel Digital (2024) and

MobiDev (2024) talk about compliance with specific

industry regulations but are descriptive and not

prescriptive.

Farooq et al. (2023) and Javed and Abbas (2022)

carried out critical reviews to point out common

vulnerabilities, such as 22 default credentials,

insecure firmware, and weak encryption. But their

research ends with a comprehensive remedy. Gubbi

et al. (2025) as well as Hassan and Khan (2024)

suggest blockchain and fog computing-based

architectural approaches, however, they have latency

and scalability problems in their models.

Islam et al. (2023) focus on IoT for healthcare,

while Sharma et al. (2023), Mosteiro-Sanchez et al.

(2022) are concerned with industrial problems.

These discipline-centric studies provide useful data,

but are not readily transferred to different networked

contexts. In contrast, Khan et al. (2021) and Kumar

and Tripathi (2023) recommend flexible structures,

yet they may not be offering functional prototypes.

A number of works have also studied

cryptographic solutions for safeguarding data

transmission in IoT environments. Li et al. (2022)

focus on encryption as a major requirement and do

not present lightweight alternatives available for

resource-constrained devices. Raza et al. (2021)

present the real-time detection system SVELTE,

which is efficient but does not include integrated

prevention. Schöttle et al. (2025) and Nakamura et al.

(2023) concentrate on device evaluation and data

fusion, respectively, but do not provide an in-depth

solution.

Shaik and Park (2022) investigate 5G API

vulnerabilities that are relevant to IoT but that are not

centered on the device. Panasonic (2023) and

StationX (2025) present real-world malware trends,

highlighting the need for proactive defense, although

they provide little technical remedy. Singh et al.

(2024) and of Tripathi and Bansal (2023) promote

fog-based and encryption-rich architectures, but the

energy cost is still one of the fundamental obstacles.

Finally, Zhang et al. (2022) target network-level

threats, but ignore firmware-level defenses.

Together, the studies indicate the need for an

adaptive, scalable, lightweight multi-layer framework

for securing IoT devices from the edge to the cloud.

Designing a Lightweight, Multi-Layer Security Framework for IoT Devices: Adaptive and Scalable Protection against Emerging Threats in

Connected Environments

213

This work aims to fill that gap by providing a

comprehensive solution that offers layered

encryption, behavior-based intrusion detection, and

post-quantum readiness—traits which are sorely

lacking in the existing literature

4 METHODOLOGY

The proposed strategy is multilayered for securing

IoT systems, since focuses on controlling solution

lightweight, dynamic and scalable to domain. Device-

layer protection, communication security, behavior-

based anomaly detection and edge-level decision

making are the five tiers of smart protection tiers of

an IIoT protection model that are connected nested

layers. This hybrid design leads to resilience against

internal and external adversaries and helps the

operation efficiency for resource-limited devices.

On the device side, the security model combines

lightweight cryptographic algorithms (eg, PRESENT,

HIGHT, or SPECK) to provide confidentiality and

integrity in a way that does not unduly strain the

device processor. These encryption techniques are

chosen for their capability on low-power

microcontrollers such as ARM Cortex-M series,

which is widely used in smart IoT devices. The

authentication is accomplished by using the pre-

shared symmetric key based protocol enhanced by

time-limited token generation scheme for protecting

against replay attacks. Figure 1 gives the Workflow

of the Proposed Multi-Layer IoT Security

Framework.

For security on the communication layer, the

model adapts the use of Datagram Transport Layer

Security (DTLS) over the User Datagram Protocol

(UDP) to ensure secure communication in real-time

between devices and gateways. To keep latency under

control and to be synchronized, session key

negotiation is done locally through ECDHE

exchanges with a distributed key distribution, thus not

depending on a central bottleneck. The Marriage of

secure routing protocol RPL, together with

cryptographic binding, allows secure message

forwarding in mesh topologies. Table 1 gives the

Performance Comparison of Lightweight Encryption

Algorithms.

Figure 1: Workflow of the proposed multi-layer IoT

security framework.

Table 1: Performance comparison of lightweight encryption

algorithms.

Algorith

Encrypt

ion

Latenc

y (ms)

RAM

Usage

(KB)

CPU

Load

(%)

Suitable

Device

Present 1.4 2.3 8.2

ESP32,

Cortex-M0

Speck 1.1 3.6 9.7

Raspberry

ht 1.6 2.9 7.5 Cortex-M3

Aes 3.9 5.4 17.1

High-end

SoCs

Figure 2: Encryption latency of different algorithms.

Figure 2 gives the Encryption Latency of

Different Algorithm. To achieve anomaly detection,

we have included a behavior profiling module at the

edge node based on a hybrid statistical-threshold

mechanism with machine learning (e.g., lightweight

decision tree classifier). This software measures

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

214

traffic, message timing, and function calls as device

activity. Aberrations from typical behavior alert and

briefly isolate devices to curb attackers from moving

laterally. The model is trained with attack types (e.g.

SYN flood, spoofing, unauthorized firmware update)

built artificially for proactive security.

The edge security layer also coordinates real-

time reactions. It consolidates alerts from the

anomaly detector and applies policy enforcement

through software-defined security rules. The edge

node also does local firmware attestation by hashing

blocks of code and validating them against a secure

hash collection before running them. This confirms

firmware integrity and prevents potential malicious

code injection or downgrading attacks.

Last but not least, the cloud layer is a brain that

conducts long-term learning and feedback adaptation.

The gateway ingests anonymized data logs from

edge devices and updates security policies based on

trend analysis, utilizing federated learning methods to

prevent the sharing of raw device data. This

decentralized training method improves system

intelligence continually while maintaining user

privacy. Cloud Integration Cloud connectivity

provides over-the-air (OTA) firmware updates as

well as policy updates that are pushed securely to the

edge and device tiers via digitally signed containers.

The approach is evaluated in the context of three

environments—viz. smart home, healthcare

monitoring, and industrial IoT— through both a

virtual platform (built in Python and NS-3) and a

proto-typical implementation on Raspberry Pi and

ESP32 platforms. We use detection accuracy,

encryption latency, memory cost, energy

consumption and false positive rate as the evaluation

metrics. This module-based and adaptive design

enables a secure-by-design model without any trade-

off on scalability or performance.

5 RESULTS AND DISCUSSION

In order to demonstrate the efficacy of the lightweight

and multi-layer security framework, a set of

experiments was carried out over three IoT

representative scenarios including smart home

automation, healthcare monitoring and industrial

sensor networks. Evaluation analyzed performance

measures such as encryption latency, memory and

CPU overhead, anomaly detection accuracy, false

positive rate, energy consumption and device

scalability.

5.1 Cryptographic Primitives &

Performance Analysis

The framework was validated using lightweight

ciphers such as PRESENT, HIGHT, and SPECK for

devices such as Raspberry Pi 4, ESP32, and ARM

Cortex-M0 microcontrollers. PRESENT was the

most efficient in real-time (constrained memory) with

a mean encryption latency of 1.4 ms and a CPU load

8.2%. SPECK was slightly faster with an encryption

time of 1.1 ms, however used more RAM, being

suitable for mid-ranged devices. HIGHT performed

with good balance in all tested features. Lightweight

ciphers reduced the cryptosystem in which we

replaced AES by more than 60% compared with

conventional AES on the same HW, which allows

real-time secure data communication without

disturbing application level functionality.

5.2 Anomaly Detection Based on

Behaviors

Behaviors emanate from observed and inferred

activities of the different network entities. Within the

edge layer, lightweight decision tree classifier is

incorporated to detect abnormal activities such as

packet flooding, spoofing, and unauthorized

firmware access. The detection model was trained on

a synthetic dataset of normal and abnormal behavior

and evaluated on both live host-device data and

simulated attack traffic. For the smart home and

healthcare-based testbeds, the average accuracy of

detection was 96.3%, and false positive of less than

3.8%. When compared to a benchmark SVM-based

detector (with 94.1% accuracy but with higher

memory overhead), the decision tree exhibited higher

efficiency and reduced overhead, which is crucial in

edge applications. Table 2 gives the Anomaly

Detection Accuracy Across Different IoT Domains

and Figure 3 illustrates the

Anomaly Detection Accuracy

Across Domains.

Table 2: Anomaly detection accuracy across different IoT

domains.

IoT

Domain

Detection

Accuracy

(%)

False

Positive

Rate (%)

Classifier

Used

Smart

Home

96.1 3.4

Decision

Tree

Healthcar

96.7 3.9

Decision

Tree

Industrial

IoT

95.9 3.1

Decision

Tree

Designing a Lightweight, Multi-Layer Security Framework for IoT Devices: Adaptive and Scalable Protection against Emerging Threats in

Connected Environments

215

Figure 3: Anomaly detection accuracy across domains.

5.3 Firmware Integrity and Security

Entry Management of Secure

Updates to Firmware

SHA-256 hashing and digital signatures were used in

the device level integrity check mechanism to secure

the firmware from tampering. Using the framework

on real-world attack scenarios including rogue

firmware downgrades and code injection attempt, it

was demonstrated up to 100% malicious payload

prevention within less than 2 seconds of role update

latency. Final Report An OpenSSL Issue The safe

update channel based on DTLS and authenticated

containers allowed to apply verified firmware patches

with a small downtime. These findings demonstrate

the robustness of the suggested updating procedure,

especially in safety-critical applications, such as

healthcare and industrial surveillance. Table 3 gives

the Firmware Integrity Validation Results. And

Figure 4 illustrates the Detection Rate against

Firmware attacks.

Table 3: Firmware integrity validation results.

Attack Type

Detection

Rate (%)

Average Response

Time (s)

Firmware

Downgrade

100 1.7

Code

Injection

100 1.8

Unauthorized

Update

Access

98.6 2.1

Figure 4: Detection rate against firmware attacks.

5.4 Scalability and Efficiency of

Resources

The scalability of the framework was evaluated by

scaling up the number of connected devices from 10

to 1,000 in a simulated industrial environment.

protocols, the communication layer was able to

achieve packet success rates over 98.7% and latency

below 70 ms in a heavy-traffic scenario. Memory

consumption on edge devices did not exceed 55%

available resources, demonstrating the efficiency of

the lightweight design. Significantly, the system

operated without performance degradation in any of

the three domains, showing the adaptability and

domain-independence of the architecture.

Table 4: Resource usage under varying device loads.

Devices

Connecte

Packet

Delivery Rate

(%)

Avg.

Latency

(ms)

Memory

Usage (%)

10 99.3 42 31

100 98.9 55 43

500 98.2 61 49

1000 98.7 68 55

Figure 5: Memory usage vs number of devices.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

216

Figure 5 gives the Memory usage vs Number of

devices. Table 4 gives the Resource Usage Under

Varying Device Loads.

5.5 Energy Efficiency

We observed battery-powered ESP32 nodes for 12

hours of operation time with and without security

components activated. Power consumption was

increased by only 8.5% when the secure framework

was enabled, a moderate overhead relative to the full

coverage offered. This also demonstrates that the

framework is practical even in energy constrained

scenarios, when coupled with duty cycling and/or

energy harvesting.

5.6 Cross-Domain Generalizability

A main open-issue in (IoT) security research is the

creation of frameworks that are not restricted to a

single use-case. The same security architecture was

reused in our research on smart homes, wearable

medical devices, and industrial sensors, with only

small parameter tuning. The agreement among these

various environments demonstrates the applicability

of the proposed method. Compared with model-based

solutions, which are designed for a single domain

and lack flexibility, both the resource allocation and

detection threshold of the proposed framework are

adaptive to the connected device category and its

surrounding environment.

5.7 Comparative Benchmarking

The framework was compared with three IoT

security models in the literature, based on the

robustness skin named SVELTE (Raza et al., 2021),

fog-based model (Singh et al., 2024) and a block-

chain architecture (Hassan & Khan, 2024). Although

with SVELTE there were low latencies, and out-of-

the-box support, it missed firmware integrity checks

and showed increased number of false positives. The

policy management in the fog-based model was

powerful, but suffered from high overhead of

communication accordingly. The blockchain-based

approach was highly secure but very resource-

consuming for low-power devices. The proposed one

was able to outpace the three in terms of detection

accuracy, update process and energy efficiency and

had a moderate level between performance and

security granularity. Table 5 gives the Comparative

Evaluation with Existing Frameworks. Figure 6

illustrates the Framework performance comparison.

Table 5: Comparative evaluation with existing frameworks.

Framework

Detection

Accuracy

(%)

Energy

Overhead

(%)

Firmware

Security

Generaliz

ability

SVELTE

(2021)

94.1 12.3

✗

Limited

Fog-Based

(2024)

95.5 10.9

✓

Moderate

Blockchain-

Based

(2024)

96.2 18.7

✓

High

Proposed

Framewor

96.3 8.5

✓✓

Very High

Figure 6: Framework performance comparison.

6 DISCUSSIONS

The experiment results also validate that the proposed

multi-layer security framework achieves the desired

requirements for contemporary IoT environments:

real-time security services, low resource

consumption, high detection rate and cross domain

adaptability. It is made modular so the layers can

work independently as well as collectively to provide

deployment flexibility. Additionally, the

implementation of post-quantum ready algorithms

and secure firmware features future-proof the design

from evolving security concerns. By leveraging

Lightweight cryptography, behavior-based IDS,

edge-cloud collaboration, this paper provides a

scalable practical solution to secure the next

generation of connected world.

7 CONCLUSIONS

The growing proliferation of IoT devices in both

consumer applications and critical applications has

shed light on the drastic need for an economical and

effective means of securing human-to-machine and

machine-to-machine connections in a connected

Designing a Lightweight, Multi-Layer Security Framework for IoT Devices: Adaptive and Scalable Protection against Emerging Threats in

Connected Environments

217

setting. This work mitigated the fundamental

shortcomings of the previous designs, by presenting a

lightweight, adaptive and multilayer security

architecture developed for heterogeneous based IoT

environments. By leveraging a proprietary integration

of lightweight cryptography, behavior based

anomaly detection, secure firmware validation and a

managed security service, the reference design

protects against the multitude of threats impacting

customers at the device, edge and cloud.

Results across various IoT domains (i.e., smart

home, industrial monitoring) verified that the model

was able to achieve high detection accuracy while

keeping the latency low and operating under strict

power and memory budgets. The practicality and

future-readiness of the framework are also shown by

the ability of the framework to make the device

resilent through firmware tampering, real-time

detection of abnormal behaviour of the system and

dynamic approach on security based on the available

resources. And, with post-quantum cryptographic

considerations, resilience to attack vectors of the

future.

Unlike many existing solutions, which are

impractical on resource-constrained devices finally

observation and only useful in specialized domains,

this work provides a comprehensive, deployable,

domain-independent solution that can be

straightforwardly expanded or incorporated to

existing IoT frameworks. By addressing the

deficiencies exposed in existing work namely, the

lack of device-level enforcement, the excessive

resource consumption and the integrity of update

dissemination—this work paves the way towards the

creation of smarter, safer, and resilient connected

environments.

REFERENCES

Akram, F. (2024). Securing Internet of Things (IoT)

Devices: Challenges and Best Practices. ResearchGate.

ResearchGate

Al-Ali, A. R., & Zualkernan, I. A. (2023). A survey on IoT

security: Threats, vulnerabilities, and countermeasures.

Journal of Network and Computer Applications, 200,

103348.

Bharati, S., & Podder, P. (2022). Machine and deep learning

for IoT security and privacy: Applications, challenges,

and future directions. arXiv. arXiv

Buyya, R., Singh, N., & Kim, H. (2024). Securing cloud-

based Internet of Things: Challenges and mitigations.

MDPI Sensors, 25(1), 79. arXiv+1MDPI+1

Compunnel Digital. (2024). Addressing 2024’s IoT security

challenges within compliance frameworks. Compunnel.

Compunnel

Farooq, M. U., Waseem, M., Khairi, A., & Mazhar, S.

(2023). A critical analysis on the security concerns of

Internet of Things (IoT). International Journal of

Computer Applications, 182(1), 1-6.

Gubbi, J., Buyya, R., Marusic, S., & Palaniswami, M.

(2025). Internet of Things (IoT): A vision, architectural

elements, and future directions. Future Generation

Computer Systems, 79, 164-180.

Hassan, Q. F., & Khan, M. A. (2024). IoT security: Review,

blockchain solutions, and open challenges. Future

Generation Computer Systems, 82, 395-411.Wikipedia

Islam, S. R., Kwak, D., Kabir, M. H., Hossain, M., & Kwak,

K. S. (2023). The Internet of Things for health care: A

comprehensive survey. IEEE Access, 3, 678-708.

Javed, M. A., & Abbas, H. (2022). A survey on security and

privacy issues in Internet-of-Things. Journal of Network

and Computer Applications, 88, 1-15.

Khan, R., Khan, S. U., Zaheer, R., & Khan, S. (2021). Future

Internet: The Internet of Things architecture, possible

applications and key challenges. In 2012 10th

International Conference on Frontiers of Information

Technology (pp. 257-260). IEEE.

Kumar, P., & Tripathi, R. (2023). Security and privacy

issues in IoT: A comprehensive survey. Computer

Networks, 149, 1-23.

Li, S., Tryfonas, T., & Li, H. (2022). The Internet of Things:

A security point of view. Internet Research, 26(2), 337-

359.

MobiDev. (2024). How to mitigate IoT security threats with

AI and ML in 2025. MobiDev. MobiDev

Mosteiro-Sanchez, A., Barcelo, M., Astorga, J., & Urbieta,

A. (2022). Securing IIoT using defence-in-depth:

Towards an end-to-end secure Industry 4.0. arXiv. arXiv

Nakamura, E. F., Loureiro, A. A. F., & Fraga, J. S. (2023).

Information fusion for wireless sensor networks:

Methods, models, and classifications. ACM Computing

Surveys, 39(3), 1-55.

Panasonic. (2023). Panasonic warns that Internet-of-Things

malware attack cycles are accelerating. Wired. WIRED

Raza, S., Wallgren, L., & Voigt, T. (2021). SVELTE: Real-

time intrusion detection in the Internet of Things. Ad

Hoc Networks, 11(8), 2661-2674.

Schöttle, P., Janetschek, M., Merkle, F., Nocker, M., &

Egger, C. (2025). Large-scale (semi-)automated security

assessment of consumer IoT devices: A roadmap. arXiv.

arXiv

Shaik, A., & Park, S. (2022). One of 5G's biggest features is

a security minefield. Wired. WIRED

Sharma, V., You, I., & Pau, G. (2023). Secure and efficient

data transmission for smart agriculture by integrating

fog computing and blockchain technology. IEEE

Access, 5, 5804-5815.

Singh, N., Buyya, R., & Kim, H. (2024). Securing cloud-

based Internet of Things: Challenges and mitigations.

arXiv. arXiv+1MDPI+1

StationX. (2025). IoT security challenges (most critical risk

of 2025). StationX

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

218