Q-learning and fuzzy logic multi-tier multi-access edge clustering for 5g v2x communication

Area of Science:

• Telecommunications Engineering
• Computer Networks
• Artificial Intelligence

Background:

• 5th generation (5G) networks are essential for increasing data demands and customer growth, supporting diverse applications like the Internet of Things.
• Existing IEEE 802.11p standards face limitations in dense vehicular environments, including limited coverage, congestion, and connectivity issues, hindering efficient data distribution.
• Cellular Vehicle-to-Everything (C-V2X) communication is crucial for overcoming bandwidth constraints in modern, dense networks.

Purpose of the Study:

• To address the challenges of efficient data distribution in dense vehicular networks within 5G.
• To propose a novel multi-layered multi-access edge clustering protocol for enhanced C-V2X communication.
• To improve network performance by reducing vehicle contention and adapting to dynamic network conditions.

Main Methods:

• Implementation of vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), and vehicle-to-pedestrian (V2P) services.
• Application of multi-layered multi-access edge clustering to reduce vehicle contention.
• Utilization of fuzzy logic and Q-learning for an intelligent multi-hop route selection system.

Main Results:

• The proposed clustering protocol effectively reduces vehicle contention in dense network scenarios.
• The integration of Q-learning allows dynamic adjustment of cluster-head nodes, enhancing adaptability.
• Improved data distribution and network efficiency are achieved in varying bandwidth and vehicle density conditions.

Conclusions:

• The developed protocol offers a robust solution for efficient data distribution in 5G vehicular networks.
• Intelligent routing and adaptive clustering significantly improve network performance and reliability.
• This approach enhances the capabilities of C-V2X services, paving the way for future intelligent transportation systems.