Prevention of Cyber Security with the Internet of Things Using Particle Swarm Optimization

Area of Science:

• Computer Science
• Cybersecurity
• Artificial Intelligence

Background:

• The Internet of Things (IoT) connects physical devices, but faces significant security risks from increasing cyberattacks.
• Attacks like user-to-root (U2R) and denial-of-service threaten IoT system integrity and data exchange.
• Effective attack detection is crucial for establishing robust security in IoT environments.

Purpose of the Study:

• To evaluate performance-based Artificial Intelligence (AI) models for predicting and detecting security threats in IoT devices.
• To enhance the accuracy and reliability of attack detection systems for intelligent machinery and residential appliances.

Main Methods:

• Investigated various AI models, including Particle Swarm Optimization (PSO), genetic algorithms, and ant colony optimization.
• Demonstrated the effectiveness of the proposed technique using PSO against common IoT attacks.
• Evaluated performance based on four distinct parameters to quantify detection accuracy.

Main Results:

• The proposed AI-driven method, particularly using PSO, showed superior performance in predicting IoT attacks.
• Particle Swarm Optimization (PSO) achieved an approximate 73% improvement in accuracy compared to existing security systems.
• The AI models effectively identified various attack types, including U2R and denial-of-service.

Conclusions:

• AI models, especially PSO, offer a promising solution for enhancing IoT security through accurate attack prediction.
• The findings highlight the potential of optimized AI algorithms to build more resilient and secure IoT ecosystems.
• Further research into AI for IoT security can lead to advanced defense mechanisms against evolving cyber threats.