Reverse-engineering of mobile network topologies

Research into future mobile networks, for example in the context of 6G, requires a deep understanding of the structure of today's mobile networks. However, detailed information on the topology of modern mobile network infrastructures is not usually published by network operators.

One way to gain insights into such networks is to reverse engineer the network topology from an end devices perspective. By measuring the connection characteristics of a mobile device - such as latencies, visible network paths or IP addresses - it is possible to derive observable characteristics of the underlying mobile network infrastructure. In combination with the respective position (GPS), assumptions can be made about the underlying topology. For example, network handover points, cluster sizes or other internal structures can be identified.

The bachelor thesis will analyze which metrics or data points are suitable for deriving such topology properties. To this end, suitable software components are to be designed and implemented that collect measurement data on a mobile device and store it in a structured manner so that it can then be analyzed. On this basis, experiments with different mobility patterns (walking, public transport, ...) will be carried out and the collected data will then be analyzed.

A Raspberry Pi 5 with a 5G module (Quectel RM520N) will be used as the measurement platform. Metrics and measurement data are collected using common Linux measurement tools (e.g. ping, traceroute, iperf3, ...) and directly from the 5G modem.