M.Sc.Topic: Analyzing Mobile Starlink and Cellular Connectivity for Autonomous Vehicles
Motivation
Low-Earth Orbit (LEO) satellite constellations, such as Starlink, are rapidily growing. While the user terminals were initially only meant for stationary use, Starlink later announced satellite dishes suitable for use in-motion, e.g., on boats, planes, and cars. With that capability, it became an alternative to cellular connectivity (5G), for example useful for connecting autonomous vehicles.
Some studies have already analyzed Starlink’s (and partly 5G’s) performance in-motion, e.g.:
- Starlink and Cellular Connectivity under Mobility: Drive Testing Across the Arctic Circle
- Starlink on the Road: A First Look at Mobile Starlink Performance in Central Europe
The Chair of Connected Mobility is part of the EDGAR team. EDGAR is a research vehicle equipped with 5G modems and—newly—Starlink. The goal of this thesis is to develop a measurement toolchain tailored to EDGAR that collects performance data of the Starlink and 5G links and to analyze the measurement artifacts. The overall goal is to research and analyze connectivity characterstics to understand how capable next generation technologies are when it comes to applications for autonomous vehicles such as teleoperation, cooperative sensing and more.
Structure
The thesis would be split into four conceptual phases with the following tasks:
- Surveying the available measurement methods (and underlying procedures) and tools. What research questions can be answered with which tool and which overarching measurement procedure? A sensible trade-off between test complexity and the number of parameters needs to be found to be able to gather measurement data in a reasonable amount of time.
- Building a measurement toolchain that executes a specific measurement procedure with a suitable subset of the surveyed tools from the previous step. The measurements should be executable with minimal human involvement to make them as robust as possible. Ideally, the toolchain is easily executable on different machines to improve reusability and reproducibility, e.g., by executing the toolchain in one or more Docker containers.
- Planning the measurement procedures and doing the data collection using EDGAR.
- Analyzing the collected measurement data to answer research questions. Depending on decisions made in step 1 (which tools and procedure were used, and which research questions one wants to investigate), you will analyze the data and come up with quantified results and visualizations.
Potential Research Questions
We use EDGAR's Starlink dish and 5G link to measure, analyze, and compare the performance of the two link technologies. EDGAR does track speed, orientation and maybe vibration and can provide camera feeds as auxiliary data that should be used for an in-depth analysis of environmental factors that affect Starlink performance when mobile at vehicular speeds. For example:
- How well does Starlink work while moving? Do sharp turns affect its performance?
- How does the current environment (obstructions, bridges) and speed affect Starlink and 5G? Are there correlations between the performance of the two links?
- i.e., are both LEO satellite connections and 5G cellular connections affected by similar environmental factors.
- In how far is the recorded performance at some specific position fixed? How much performance variety do we observe at the same position, at the same speed?
Required Knowledge
- Linux plumbing skills: Docker, networking (tcpdump, Wireshark, iproute2), bash scripting
- Good knowledge of computer networking concepts
- Data analysis, e.g., with python (sqlite, matplotlib, pandas)
Contact
Please send your application by email to Justus Fries and Hendrik Cech. Please include a grade report and your CV.