BERTO: an Adaptive BERT-based Network Time Series Predictor with Operator Preferences in Natural Language

2025-12-05

AI Summary

This work addresses the joint optimization of traffic forecasting and energy consumption in cellular networks. We propose a BERT-based, prompt-driven time-series forecasting framework that integrates the Transformer architecture with pre-trained BERT, incorporates natural language prompting techniques, and introduces a custom balanced loss function. Crucially, it enables dynamic adjustment of under-prediction versus over-prediction bias via operator-issued natural language instructions—marking the first such approach to flexibly trade off energy savings against QoS requirements. Experiments on real-world datasets show a 4.13% reduction in MSE; the framework supports continuous, fine-grained control across a 1.4 kW power range and up to 9× QoS variation, significantly enhancing the adaptability and deployment efficiency of intelligent RANs. The core innovation lies in embedding natural language interaction into a closed-loop, multi-objective time-series optimization pipeline, thereby transcending conventional static weighting paradigms.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Natural Language Processing: Learning & Optimization for NLP
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

We introduce BERTO, a BERT-based framework for traffic prediction and energy optimization in cellular networks. Built on transformer architectures, BERTO delivers high prediction accuracy, while its Balancing Loss Function and prompt-based customization allow operators to adjust the trade-off between power savings and performance. Natural language prompts guide the model to manage underprediction and overprediction in accordance with the operator's intent. Experiments on real-world datasets show that BERTO improves upon existing models with a $4.13$% reduction in MSE while introducing the feature of balancing competing objectives of power saving and performance through simple natural language inputs, operating over a flexible range of $1.4$ kW in power and up to $9 imes$ variation in service quality, making it well suited for intelligent RAN deployments.

Problem

• Predicts cellular network traffic and energy usage
• Balances power savings and performance via operator prompts
• Reduces prediction error while managing service quality trade-offs

Innovation

• BERT-based transformer for network time series prediction
• Balancing Loss Function with natural language prompts
• Adjusts power-performance trade-off via operator preferences