Benchmarking Quantum and Classical Sequential Models for Urban Telecommunication Forecasting

2025-08-06

AI Summary

This study addresses univariate SMS-in traffic forecasting on the Milan Telecom dataset, a critical urban communication flow prediction task. Method: We systematically evaluate five sequence modeling paradigms—classical LSTM and four quantum-inspired models (QLSTM, QASA, QRWKV, QFWP)—across varying input lengths (4–64). Contribution/Results: Contrary to the “quantum advantage” assumption, quantum-enhanced models do not universally outperform classical LSTM; their superiority is highly contingent on task characteristics, architectural design, and sequence length. Notably, models exhibit markedly divergent sensitivity to input length, exposing fundamental trade-offs among model capacity, parameter efficiency, and long-horizon temporal modeling capability. These findings empirically challenge the oversimplified notion of inherent quantum superiority and provide methodological insights and evidence-based guidance for rational architecture selection and structural optimization of quantum-inspired models in large-scale urban traffic forecasting.

Technology Category

• Machine Learning: Quantum Machine Learning
• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

In this study, we evaluate the performance of classical and quantum-inspired sequential models in forecasting univariate time series of incoming SMS activity (SMS-in) using the Milan Telecommunication Activity Dataset. Due to data completeness limitations, we focus exclusively on the SMS-in signal for each spatial grid cell. We compare five models, LSTM (baseline), Quantum LSTM (QLSTM), Quantum Adaptive Self-Attention (QASA), Quantum Receptance Weighted Key-Value (QRWKV), and Quantum Fast Weight Programmers (QFWP), under varying input sequence lengths (4, 8, 12, 16, 32 and 64). All models are trained to predict the next 10-minute SMS-in value based solely on historical values within a given sequence window. Our findings indicate that different models exhibit varying sensitivities to sequence length, suggesting that quantum enhancements are not universally advantageous. Rather, the effectiveness of quantum modules is highly dependent on the specific task and architectural design, reflecting inherent trade-offs among model size, parameterization strategies, and temporal modeling capabilities.

Problem

• Compare quantum and classical models for SMS forecasting
• Evaluate model sensitivity to input sequence length
• Assess quantum enhancements' task-dependent effectiveness

Innovation

• Quantum LSTM for SMS activity forecasting
• Quantum Adaptive Self-Attention modules
• Sequence length sensitivity analysis