Handover Optimization Algorithm Based on T2RFS-FNN.

Area of Science:

• Telecommunications Engineering
• Railway Systems
• Artificial Intelligence

Background:

• Handover failures in fifth-generation mobile communication for railway (5G-R) systems pose significant risks to high-speed rail safety.
• Increasing train speeds necessitate more frequent handovers, making success rate improvement a critical challenge.
• Existing handover algorithms in 5G-R struggle with fixed hysteresis thresholds and limited parameter considerations.

Purpose of the Study:

• To propose an optimized handover algorithm for 5G-R high-speed railway wireless communication.
• To address the limitations of fixed hysteresis thresholds and single-parameter considerations in current handover algorithms.
• To enhance the reliability and safety of high-speed railway communication systems.

Main Methods:

• An interval type-2 feature selection recurrent fuzzy neural network (T2RFS-FNN) was developed.
• The algorithm integrates Reference Signal Receiving Power (RSRP), Reference Signal Receiving Quality (RSRQ), and throughput to optimize the hysteresis threshold.
• A feedforward neural network with interval type-two Gaussian membership functions and a feature selection layer was designed.

Main Results:

• The proposed T2RFS-FNN algorithm effectively optimizes the hysteresis threshold for handover.
• Simulation analysis demonstrated a significant improvement in handover success rate compared to existing algorithms.
• The algorithm successfully reduced the ping-pong handover rate, enhancing communication stability.

Conclusions:

• The developed T2RFS-FNN algorithm offers a robust solution for optimizing handover in 5G-R systems.
• This research provides a theoretical foundation for advancing high-speed train speeds and the transition from GSM-R to 5G-R.
• The findings contribute to the development of safer and more reliable railway communication infrastructure.