Traffic Weaver: semi-synthetic time-varying traffic generator based on averaged time series

TL;DR

The paper tackles the lack of accessible real traffic data by proposing Traffic Weaver, a semi-synthetic time-varying traffic generator. It reconstructs high-granularity signals from an averaged input using a modular pipeline that includes oversampling, integral matching, smoothing, repeating, trend, and noise. It provides a Python-based Weaver API, multiple oversampling strategies, a dataset catalog based on Sandvine traffic profiles, and support for continuous outputs via splines. The tool enables robust development and evaluation of traffic-prediction and network-optimization methods under varied, controllable conditions, while promoting reproducibility through open-source distribution and ready-to-use baselines.

Abstract

Traffic Weaver is a Python package developed to generate a semi-synthetic signal (time series) with finer granularity, based on averaged time series, in a manner that, upon averaging, closely matches the original signal provided. The key components utilized to recreate the signal encompass oversampling with a given strategy, stretching to match the integral of the original time series, smoothing, repeating, applying trend, and adding noise. The primary motivation behind Traffic Weaver is to furnish semi-synthetic time-varying traffic in telecommunication networks, facilitating the development and validation of traffic prediction models, as well as aiding in the deployment of network optimization algorithms tailored for time-varying traffic.

Figures (3)

• Figure 1: Software architecture.
• Figure 2: Illustrative example.
• Figure 3: Minimal processing example.