Remote Interference Discrimination Testbed Employing AI Ensemble Algorithms for 6G TDD Networks.

Area of Science:

• Telecommunications Engineering
• Wireless Communication Systems
• Artificial Intelligence in Networking

Background:

• Massive Internet-of-Things (IoT) access is a key scenario for future sixth-generation (6G) networks.
• Low-power IoT devices are susceptible to remote interference from atmospheric ducting in 6G time-division duplex (TDD) systems.
• This interference degrades signal integrity, causing distant signals to disrupt local uplink communications and increasing outage probability.

Purpose of the Study:

• To propose and validate a novel remote interference discrimination testbed for 6G TDD networks.
• To evaluate the performance of various artificial intelligence (AI) algorithms in detecting and discriminating remote interference.
• To compare the effectiveness of different AI approaches using a large, real-world dataset.

Main Methods:

• Development of a dedicated remote interference discrimination testbed.
• Collection and utilization of a substantial dataset comprising 5,520,000 TDD network-side data points from real sensors.
• Validation and comparative analysis of nine distinct AI algorithms for interference discrimination capabilities.

Main Results:

• The proposed testbed successfully supports the comparison of multiple AI algorithms for interference detection.
• AI algorithms demonstrated varying degrees of effectiveness in discriminating remote interference.
• The ensemble algorithm achieved an average accuracy 12% higher than single-model algorithms in interference discrimination.

Conclusions:

• The developed testbed is effective for evaluating AI algorithms in 6G interference scenarios.
• Ensemble AI algorithms offer superior performance for remote interference discrimination compared to single models.
• The findings contribute to mitigating interference challenges for massive IoT access in future 6G networks.