TY - THES A1 - Korb, Andreas Diemo T1 - Evaluating approaches for more efficient UDS protocol scanning N2 - Electronic Control Units (ECUs) in cars are the primary target in an attack on a car network. To assess the security of a particular network, a penetration tester must quickly and preferably automatically gather information about them. This can be accomplished by scanning ECUs with diagnostic protocols; most ECUs support at least one. The most widely supported diagnostic protocol is the Unified Diagnostic Services (UDS) protocol. The idea is to send every possible request and evaluate its response. This brute force approach leads to high run times. In this thesis, three approaches were evaluated to reduce the runtime by avoiding requests that are not expected to produce positive responses, namely reusing information within the same service (approach 1), using probabilities of positive answers based on blocks (approach 2) and skipping unsupported services (approach 3). Data from eleven ECUs was collected for analysis. Simulations based on this data were used to determine which exact behavior of each approach produces the best results. They were implemented for the existing UDS Scanner in the Scapy program. Ultimately, the approaches were found to provide high coverages and request savings leading to new runtimes between 10 % - 30 % of the original runtimes. Nevertheless, complete coverage cannot be guaranteed. Y1 - 2021 UR - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:573-22554 CY - Ingolstadt ER -