DP-LET: An Efficient Spatio-Temporal Network Traffic Prediction Framework

2025-04-04

AI Summary

To address low prediction accuracy, high computational overhead, and noise sensitivity in spatiotemporal network traffic forecasting, this paper proposes DP-LET—a lightweight and efficient framework. DP-LET introduces a novel three-module collaborative architecture: (1) adaptive filtering and spatial decoupling preprocessing to enhance noise robustness and spatial interpretability; (2) multi-scale Temporal Convolutional Network (TCN) branches for refined local spatiotemporal feature modeling; and (3) a lightweight Transformer encoder to capture long-range temporal dependencies. These modules jointly achieve a unified balance between fine-grained local pattern characterization and global modeling under low computational complexity. Evaluated on real-world cellular network traffic data, DP-LET outperforms state-of-the-art methods by reducing MSE by 31.8% and MAE by 23.1%, while cutting inference latency by 40%. The framework thus significantly advances the trade-off among prediction accuracy, robustness to noise, and computational efficiency.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Mixture of Experts (MoE)
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

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
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Accurately predicting spatio-temporal network traffic is essential for dynamically managing computing resources in modern communication systems and minimizing energy consumption. Although spatio-temporal traffic prediction has received extensive research attention, further improvements in prediction accuracy and computational efficiency remain necessary. In particular, existing decomposition-based methods or hybrid architectures often incur heavy overhead when capturing local and global feature correlations, necessitating novel approaches that optimize accuracy and complexity. In this paper, we propose an efficient spatio-temporal network traffic prediction framework, DP-LET, which consists of a data processing module, a local feature enhancement module, and a Transformer-based prediction module. The data processing module is designed for high-efficiency denoising of network data and spatial decoupling. In contrast, the local feature enhancement module leverages multiple Temporal Convolutional Networks (TCNs) to capture fine-grained local features. Meanwhile, the prediction module utilizes a Transformer encoder to model long-term dependencies and assess feature relevance. A case study on real-world cellular traffic prediction demonstrates the practicality of DP-LET, which maintains low computational complexity while achieving state-of-the-art performance, significantly reducing MSE by 31.8% and MAE by 23.1% compared to baseline models.

Problem

• Improving accuracy and efficiency in spatio-temporal network traffic prediction
• Reducing computational overhead in capturing local and global feature correlations
• Optimizing energy consumption and resource management in communication systems

Innovation

• Efficient denoising and spatial decoupling module
• Multiple TCNs for local feature enhancement
• Transformer encoder for long-term dependency modeling