Master Thesis: Network Digital Twin Fidelity and Synthetic Data Generation for Radio Networks

Join our Team

About this opportunity:

We are looking for a Master's student interested in wireless communications, simulation, and machine learning. The thesis will explore how Network Digital Twins (NDTs) can accurately represent real radio networks and efficiently generate synthetic data for AI and analytics applications.

A Network Digital Twin is a simulation-based virtual replica of a radio network used to predict performance, evaluate new features, and support network optimization. Its usefulness depends on:

Fidelity: how accurately it reproduces real network behaviour.

Efficiency: how effectively it generates data for machine learning and analytics.

The scope will be adapted to your interests and background. Possible focus areas include:

Evaluating Digital Twin fidelity by comparing simulated KPIs with real network measurements and investigating calibration methods.

Exploring synthetic data generation for machine learning, including the trade-off between simulation fidelity, cost, and model performance.

Relevant metrics may include RSRP, SINR, throughput, mobility performance, and other radio network KPIs.

You will gain experience in Network Digital Twin modelling and validation, radio network simulation, real-world measurement analysis, machine learning for wireless systems, experimental research, and large-scale datasets. We offer supervision from experienced researchers and engineers, access to simulation tools and real network data, and a collaborative environment in which you can shape the thesis direction.

What you will do:

Depending on the selected focus area, you will:

Configure and run system-level radio network simulations.

Analyse and compare simulation results with real network measurements.

Define and evaluate simulation-fidelity metrics.

Investigate mismatches between simulated and real-world behaviour.

Explore calibration and modelling improvements.

Design and evaluate synthetic data generation approaches.

Train and evaluate machine learning models using synthetic, real, or mixed datasets.

Document and communicate your findings in a Master's thesis.

The skills you bring:

You are enrolled in a Master's programme in Electrical Engineering, Computer Science, Engineering Physics, Data Science, or a related field.

You have programming experience in Python.

You have a background in wireless communications, machine learning, statistics, or another quantitative discipline.

You are interested in simulation, data analysis, and research.

You have strong analytical, problem-solving, written, and verbal communication skills.

The following are considered a plus:

Familiarity with wireless communication systems, including 5G NR.

Experience with simulation tools or modelling environments.

Knowledge of machine learning workflows and data analysis.

Experience with experimental or measurement data.

Familiarity with reproducible research or software development practices.