Time Series Network Utilization KPI Forecasting Using Advanced AI/ML Models

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Niraj Gadhe, Kirti Bhardwaj, Moulik Jain, Shubhi Sharma, Vinay Saini
• Venue: arXiv: cs.LG
• Published: 2026-07-22
• Collected: 2026-07-23
• URL: http://arxiv.org/abs/2607.19974v1
• PDF: https://arxiv.org/pdf/2607.19974v1
• DOI: UNKNOWN

Plain-Language Brief

This study evaluates various AI/ML models for forecasting bandwidth utilization in networks, aiming to improve capacity planning and reduce costly over-provisioning.

What the Document Says

• The proliferation of data-intensive applications and IoT ecosystems has made proactive resource provisioning important for optimal network performance.
• Network administrators face issues with traditional reactive approaches due to inaccurate traffic fluctuation predictions, leading to costly over-provisioning, unexpected downtime, and degraded quality of service.
• Accurate forecasting of bandwidth utilization is essential for efficient capacity planning.
• The study evaluates several models: seasonal decomposition, Prophet, Random Forest, XGBoost, Support Vector Regression, bidirectional LSTMs, and Convolutional LSTMs.
• All models were tested using a common interface dataset and benchmarked across MAPE, NRMSE, and R-square metrics.
• The research provides insights into the trade-offs between model accuracy and computational efficiency.
• The findings empower engineers, operators, and business owners to choose the optimal forecasting model for their specific infrastructure needs.

Key Concepts Mentioned

• data-intensive applications
• cloud infrastructure
• IoT ecosystems
• network administrators
• capacity constraints
• traditional reactive approaches
• bandwidth utilization
• forecasting
• AI/ML models
• seasonal decomposition
• Prophet
• Random Forest
• XGBoost
• Support Vector Regression
• LSTM
• bidirectional LSTMs
• Convolutional LSTMs
• MAPE
• NRMSE
• R-square

Claims Made by the Document

• Accurate forecasting of bandwidth utilization is essential for efficient capacity planning.
• The study evaluates various AI/ML models to improve this forecasting.

Evidence Used

• The text describes the challenges faced by network administrators due to inaccurate traffic fluctuation predictions and the importance of accurate bandwidth utilization forecasting.
• The text lists the models being evaluated and mentions that they were tested using a common interface dataset.

Methods / Approach

• Evaluating various AI/ML models
• Testing models using a common interface dataset
• Benchmarking models across MAPE, NRMSE, and R-square metrics

Limitations Stated by the Authors

Not stated in source

Limitations Noted by the Agent

Agent interpretation — not from the source.

The study's findings are based on the performance of specific AI/ML models. It is unclear how well these models will perform on other types of networks or datasets.

Open Questions

• What is the specific infrastructure needs that the research aims to address?
• How were the models chosen for evaluation?

Notes for Future Q&A

May include interpretation. Labeled where it does.

The text does not provide information about the specific infrastructure needs the study aims to address or how the models were chosen for evaluation.

Raw Abstract / Excerpt

The rapid proliferation of data-intensive applications, cloud infrastructure, and IoT ecosystems has made proactive resource provisioning critical for maintaining optimal network performance. However, network administrators face a constant battle against capacity constraints, where traditional reactive approaches fail to accurately anticipate traffic fluctuations. This inability to foresee demand leads to costly over-provisioning, unexpected downtime, and degraded quality of service directly impacting operational budgets and business continuity. To achieve efficient capacity planning, accurate forecasting of bandwidth utilization is essential. This study addresses the challenge by evaluating a diverse spectrum of models including seasonal decomposition, Prophet, Random Forest, XGBoost, Support Vector Regression, and advanced deep learning architectures like bidirectional and Convolutional LSTMs - using a common interface dataset benchmarked across MAPE, NRMSE, and R-square metrics. Ultimately, this research delivers actionable insights into the trade-offs between model accuracy and computational efficiency, empowering engineers, operators, and business owners to select the optimal forecasting model for their specific infrastructure needs.