Sequence-Based Deep Learning for Handover Optimization in Dense Urban Cellular Network

2025-10-03

AI Summary

In dense urban cellular networks, frequent handovers and severe ping-pong effects degrade performance, while conventional A3/RSRP-based handover criteria lack robustness. Method: This paper proposes a sequence-based deep learning handover prediction framework that fuses multi-dimensional spatiotemporal features. Leveraging real-world multi-operator drive-test data, we design a lightweight GRU model that formulates handover decisions as a time-series prediction task, enabling low-latency edge deployment. Results: Experiments show a 98% reduction in ping-pong handovers, a 46.25% decrease in unnecessary handovers, a 46% increase in average user dwell time, and an inference latency of only 0.91 seconds per decision. To the best of our knowledge, this is the first work to systematically validate the substantial gains of sequential models for mobility management in real-world, multi-operator urban deployments—providing a practical, deployable solution for intelligent handover in high-density networks.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Efficient handover management remains a critical challenge in dense urban cellular networks, where high cell density, user mobility, and diverse service demands increase the likelihood of unnecessary handovers and ping-pong effects. This paper leverages a real-world, multi-operator drive-test dataset of 30,925 labelled records collected within a 2 km area around Sunway City to investigate sequence-based deep learning approaches for handover detection and avoidance. We formulate handover prediction as a sequence problem and evaluate Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), and Transformer architectures under Reference Signal Received Power (RSRP)-only and all-feature settings. The integration of multi-dimensional features significantly enhanced handover performance in dense urban cellular networks. The proposed GRU-based model achieved a remarkable 98% reduction in ping-pong handovers, alongside a 46.25% decrease in unnecessary handovers, outperforming the baseline RSRP-only approach which yielded a 22.19% reduction. Furthermore, the model demonstrated a 46% improvement in Time of Stay (ToS), indicating more stable user connections. With an inference time of just 0.91 seconds, the solution proves highly efficient and well-suited for real-time edge deployment scenarios. Compared to the conventional 3GPP A3 algorithm, these improvements demonstrate significant gains in mobility robustness and user Quality of Experience (QoE) improvement. The dataset is released to foster reproducibility and further research in intelligent mobility management for 5G and beyond.

Problem

• Optimizing handover management in dense urban cellular networks
• Reducing unnecessary handovers and ping-pong effects using deep learning
• Improving mobility robustness and user experience through sequence prediction

Innovation

• Sequence-based deep learning for handover optimization
• GRU model reduces ping-pong handovers by 98%
• Multi-dimensional features enhance urban network performance