A Location-Aware Hybrid Deep Learning Framework for Dynamic Near-Far Field Channel Estimation in Low-Altitude UAV Communications

2025-10-23

AI Summary

This work addresses the challenging problem of dynamic near-field/far-field hybrid channel estimation in low-altitude UAV communications—caused by high mobility, large-scale antenna arrays, and spherical wavefront propagation. We propose a hybrid deep learning framework that explicitly incorporates real-time geometric position priors of both transmitter and receiver. Our approach innovatively embeds 3D spatial coordinates into the model architecture and jointly leverages CNNs for spatial feature extraction, BiLSTMs for temporal dynamics modeling, and multi-head self-attention to emphasize dominant channel components, thereby unifying the characterization of non-stationary near-/far-field channel variations. Compared to state-of-the-art methods, our framework achieves an average normalized mean square error reduction of ≥30.25%, significantly improving estimation accuracy, robustness against mobility-induced distortions, and generalization capability across complex air-to-ground scenarios.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Deep Learning Algorithms
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

In low altitude UAV communications, accurate channel estimation remains challenging due to the dynamic nature of air to ground links, exacerbated by high node mobility and the use of large scale antenna arrays, which introduce hybrid near and far field propagation conditions. While conventional estimation methods rely on far field assumptions, they fail to capture the intricate channel variations in near-field scenarios and overlook valuable geometric priors such as real-time transceiver positions. To overcome these limitations, this paper introduces a unified channel estimation framework based on a location aware hybrid deep learning architecture. The proposed model synergistically combines convolutional neural networks (CNNs) for spatial feature extraction, bidirectional long short term memory (BiLSTM) networks for modeling temporal evolution, and a multihead self attention mechanism to enhance focus on discriminative channel components. Furthermore, real-time transmitter and receiver locations are embedded as geometric priors, improving sensitivity to distance under near field spherical wavefronts and boosting model generalization. Extensive simulations validate the effectiveness of the proposed approach, showing that it outperforms existing benchmarks by a significant margin, achieving at least a 30.25% reduction in normalized mean square error (NMSE) on average.

Problem

• Estimating dynamic near-far field channels in UAV communications
• Overcoming limitations of far-field assumptions in channel estimation
• Integrating geometric priors and deep learning for accurate modeling

Innovation

• Location-aware hybrid deep learning for UAV channel estimation
• Combines CNNs, BiLSTM, and self-attention for spatiotemporal features
• Embeds real-time geometric priors to enhance near-field sensitivity