Studying the Role of Synthetic Data for Machine Learning-based Wireless Networks Traffic Forecasting

TL;DR

This paper tackles the data scarcity and privacy concerns in ML-based Wi‑Fi traffic forecasting by introducing a Gauss–Markov AR-based synthetic data generator driven by seed knowledge extracted from real AP traffic. The method anchors synthetic traces to weekly and hourly statistics while injecting controlled stochasticity to expand the training space. Empirical results show that models trained with synthetic data achieve MAE within 10–15 of real-data baselines for the same APs and can outperform real-data models in cross-AP generalization when enough synthetic data is used, highlighting the approach’s potential for privacy-preserving, scalable network analytics. The findings underscore the importance of model choice (CNN vs. LSTM) and task framing for maximizing the benefits of synthetic data in wireless traffic forecasting, with practical implications for proactive network management and data-sharing policies.

Abstract

Synthetic data generation is an appealing tool for augmenting and enriching datasets, playing a crucial role in advancing artificial intelligence (AI) and machine learning (ML). Not only does synthetic data help build robust AI/ML datasets cost-effectively, but it also offers privacy-friendly solutions and bypasses the complexities of storing large data volumes. This paper proposes a novel method to generate synthetic data, based on first-order auto-regressive noise statistics, for large-scale Wi-Fi deployments. The approach operates with minimal real data requirements while producing statistically rich traffic patterns that effectively mimic real Access Point (AP) behavior. Experimental results show that ML models trained on synthetic data achieve Mean Absolute Error (MAE) values within 10 to 15 of those obtained using real data when trained on the same APs, while requiring significantly less training data. Moreover, when generalization is required, synthetic-data-trained models improve prediction accuracy by up to 50 percent compared to real-data-trained baselines, thanks to the enhanced variability and diversity of the generated traces. Overall, the proposed method bridges the gap between synthetic data generation and practical Wi-Fi traffic forecasting, providing a scalable, efficient, and real-time solution for modern wireless networks.

Figures (9)

• Figure 1: Overview of ML-empowered networks with integrated synthetic data generation capabilities.
• Figure 2: Gauss-Markov Noise Model scheme.
• Figure 3: Mean and standard deviation of the Bytes transmitted over one week by a real AP and its synthetic counterpart.
• Figure 4: Distribution of the bytes transmitted by a real AP and its synthetic counterpart.
• Figure 5: Correlation between the real and synthetic data.