Deep Learning Prediction of Beam Coherence Time for Near-FieldTeraHertz Networks

2025-11-03

AI Summary

Terahertz (THz) communications face significant challenges including high beam alignment overhead due to large-scale antenna arrays and pronounced near-field spherical wavefront effects. Method: This paper introduces, for the first time, the concept of “beam coherence time” as a dynamic criterion for beam update decisions, and proposes a feedforward neural network that jointly models channel time-variation and near-field geometric characteristics to predict coherence time in real time. Contribution/Results: The method breaks away from conventional fixed-interval beam scanning, enabling adaptive beamforming frequency control that maintains link reliability while substantially reducing signaling overhead. Experimental results demonstrate that the proposed scheme achieves a 32% increase in throughput and reduces beam update frequency by up to 67% under high-mobility conditions—establishing an efficient and robust intelligent near-field beam management paradigm for THz mobile communications.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Multiagent Systems: Agent Communication
• Machine Learning: Hardware-aware ML

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search

Abstract

Large multiple antenna arrays coupled with accu- rate beamforming are essential in terahertz (THz) communi- cations to ensure link reliability. However, as the number of antennas increases, beam alignment (focusing) and beam tracking in mobile networks incur prohibitive overhead. Additionally, the near-field region expands both with the size of antenna arrays and the carrier frequency, calling for adjustments in the beamforming to account for spherical wavefront instead of the conventional planar wave assumption. In this letter, we introduce a novel beam coherence time for mobile THz networks, to drastically reduce the rate of beam updates. Then, we propose a deep learning model, relying on a simple feedforward neural network with a time-dependent input, to predict the beam coherence time and adjust the beamforming on the fly with minimal overhead. Our numerical results demonstrate the effectiveness of the proposed approach by enabling higher data rates while reducing the overhead, especially at high (i.e., vehicular) mobility.

Problem

• Predicting beam coherence time for THz networks
• Reducing beam alignment overhead in mobile networks
• Adapting beamforming for near-field spherical wavefront effects

Innovation

• Deep learning predicts beam coherence time
• Feedforward neural network enables dynamic beamforming
• Model reduces beam updates with minimal overhead