🤖 AI Summary
To address critical challenges in high-speed railway (HSR) communications—including rapid channel time-variations, excessive signaling overhead, and high radio link failure (RLF) rates caused by high-speed mobility—this work proposes an AI-driven, joint beam-level and cell-level mobility management framework. We design a compressed sensing–based multi-beam spatial measurement compression scheme to drastically reduce measurement overhead. Furthermore, we develop an LSTM-GNN hybrid model that jointly exploits spatiotemporal features for accurate beam prediction and introduce an AI-enabled proactive handover decision mechanism. Experimental results demonstrate that, under identical measurement overhead, our beam prediction accuracy surpasses conventional downsampling methods; RLF rate is significantly reduced; and beam measurement overhead is cut by 50%, achieving an optimal trade-off between low signaling cost and high reliability.
📝 Abstract
High-speed railway (HSR) communications are pivotal for ensuring rail safety, operations, maintenance, and delivering passenger information services. The high speed of trains creates rapidly time-varying wireless channels, increases the signaling overhead, and reduces the system throughput, making it difficult to meet the growing and stringent needs of HSR applications. In this article, we explore artificial intelligence (AI)-based beam-level and cell-level mobility management suitable for HSR communications. Particularly, we propose a compressed spatial multi-beam measurements scheme via compressive sensing for beam-level mobility management in HSR communications. In comparison to traditional down-sampling spatial beam measurements, this method leads to improved spatial-temporal beam prediction accuracy with the same measurement overhead. Moreover, we propose a novel AI-based proactive handover scheme to predict handover events and reduce radio link failure (RLF) rates in HSR communications. Compared with the traditional event A3-based handover mechanism, the proposed approach significantly reduces the RLF rates which saves 50% beam measurement overhead.