Related Experiment Video
Updated: Jun 28, 2026
03:14
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
1.3K
Enhancement of cryptography algorithms for security of cloud-based IoT with machine learning models
Mohammed Ali Qasem1, Bokare Madhav Motiram2, Suryakant Thorat3
1School of Computational Sciences, Swami Ramanand Teerth Marathwada University, Nanded, 431606, Maharashtra, India.
Scientific Reports
|March 27, 2026
Summary
This study introduces a hybrid security framework for cloud-based Internet of Things (IoT) systems, balancing encryption efficiency with machine learning for threat detection. The research highlights hybrid cryptography and ensemble models as key to securing IoT data transmission.
Area of Science:
- Computer Science
- Cybersecurity
- Data Science
Background:
- Cloud-based Internet of Things (IoT) systems face escalating security risks due to sensitive data transmission from resource-limited devices.
- Traditional cryptographic methods present significant computational and memory burdens, hindering efficient IoT security.
- There is a pressing need for security solutions that offer robust data protection, optimize resource usage, and enable intelligent threat detection.
Purpose of the Study:
- To propose and evaluate an integrated security framework for cloud-based IoT data transmission.
- To assess the resource efficiency of lightweight and hybrid cryptographic algorithms (XOR, ChaCha20, AES, AES-RSA).
- To investigate the performance of machine learning models (Random Forest, XGBoost, CatBoost, ensemble classifiers) for intrusion detection in IoT environments.
Main Methods:
- Simulated secure data transmission using the MQTT protocol.
- Evaluated encryption techniques based on memory consumption, CPU usage, and Overall Resource Consumption Score (ORCS).
- Employed machine learning classifiers for intrusion detection on MQTTEEB-D and CIC IoT 2023 datasets.
Main Results:
- The hybrid AES-RSA scheme demonstrated high resource efficiency (ORCS of 0.56 on MQTTEEB-D, 0.5425 on CIC IoT 2023) with low memory usage.
- Ensemble classifiers achieved high intrusion detection accuracy (92.68% on MQTTEEB-D, 81.09% on CIC IoT 2023).
- The integrated framework effectively balanced security and efficiency for IoT data transmission.
Conclusions:
- Hybrid cryptographic algorithms offer a practical solution for securing cloud-based IoT systems by optimizing resource utilization.
- Ensemble machine learning models significantly improve the accuracy and effectiveness of intrusion detection in IoT networks.
- The proposed framework provides a viable approach to enhance the security and reliability of cloud-connected IoT devices.
