Dr. Benjamin Erb

Welcome! My name is Benjamin Erb and I am a computer scientist primarily working in the field of distributed systems and interdisciplinary projects, especially with researchers from psychology.

I am currently employed as a researcher at the Institute of Distributed Systems, Ulm University. I hold a Diploma degree in Computer Science in Media and a Bachelor degree in Psychology from Ulm University. In 2019, I received my doctoral degree for my work on a novel live graph computing approach that combines concepts of traditional graph computing with features from event-driven architectures.

Main Research Interests

I have a strong interest in the design and implementation of data-intensive architectures, in particular distributed platforms for data processing. This includes processing models and programming abstractions for special data (e.g., evolving graph structures) as well as underlying platform architectures that aim for proper performance and scalability. I take special interest in data-intensive systems that consume stream-based input data and provide temporal capabilities such as retrospection. Here, I am using techniques such as event sourcing to enable history-aware processing.
A second research interest addresses the interaction between data privacy and psychology. On the one hand, this includes the tension field between confidentiality, data protection, and empirical research in psychology. Here, I am exploring potential threats to privacy from public research data sets due to the open science movement, but also technical solutions and mitigations based on privacy-enhancing technologies. On the other hand, I am also interested in psychological aspects of privacy to better understand user behavior when it comes to privacy decisions.

Updates

Apr 15, 2026	Additional publications related to the Trust Assessment Framework concepts have been accepted: Improving Misbehaviour Detection Through Infrastructure Support Without Raising Complexity (IEEE VTC2026-Spring), VeReMi NextGen: A Dataset for Evaluating Misbehavior Detection Systems in VANETs (IEEE VNC 2026), and Disbelief-Favouring Trust Discounting for Adversarial Multi-Hop Trust Assessment using Subjective Logic (FUSION 2026).
Oct 1, 2025	The Horizon Europe CONNECT project was successfully completed. Relevant source code artifacts have been released as open source at github.com/horizon-connect-eu. This includes the Trust Assessment Framework (TAF) prototype, for which I served as lead architect. Latest versions of the TAF can be found here at github.com/vs-uulm/go-taf.
Sep 15, 2025	Several new publications on our ongoing work on the Trust Assessment Framework concepts have been accepted and published: Quantification Methods for Trust in Cooperative Driving (IEEE VNC 2025), Demo: Quantifying Trust in a Trust Assessment Framework (IEEE VNC 2025), Vehicle-to-Everything Trust: Enabling Autonomous Trust Assessment of V2X Data by Vehicles (ACM CSCS 2025), and Hyperparameter Optimization-Based Trust Quantification for Misbehavior Detection Systems. (IEEE ITSC 2025).
Nov 8, 2024	Our paper on PrePaMS – a privacy-preserving participant management system for studies – has been accepted for the Privacy Enhancing Technologies Symposium 2025.1.
Jun 20, 2022	I have been invited as a guest lecturer for the Reykjavík Summer School on Secure and Reliable Distributed Systems 2022 and gave a talk on Reliable Online Event Processing with Event Sourcing.

Latest Publications

Conference Paper

Improving Misbehaviour Detection Through Infrastructure Support Without Raising Complexity

Bassi, Francesca, Zhang, Jiahao, Jemaa, Ines Ben, Kargl, Frank, and Erb, Benjamin

In 2026 IEEE 103rd Vehicular Technology Conference (VTC2026-Spring) 2026

Abstract

Ensuring the semantic correctness of exchanged kinematic data is critical for safety-critical Vehicle-to-Everything (V2X) applications. While onboard Misbehaviour Detection (MBD) mechanisms help address this issue, their effectiveness is inherently limited by the vehicle’s local view. This work investigates whether lightweight, rule-based MBD can be significantly enhanced through infrastructure support without increasing computational complexity. We adopt a Trust Assessment Framework to control the inclusion of V2X data into the vehicle’s Extended Perception Map (EPM), based on trust levels derived from MBD outputs. We compare a standalone setup relying on local evidence only with a federated setup in which the infrastructure aggregates Misbehaviour Reports from multiple vehicles, assesses node trustworthiness, and disseminates this information back to vehicles. Simulation results under kinematic falsification attacks show that the federated setup consistently outperforms the standalone one in filtering altered observations.
Conference Paper

VeReMi NextGen: A Dataset for Evaluating Misbehavior Detection Systems in VANETs

Hermann, Artur, Remmers, Jan Niklas, Eisermann, Dennis, Erb, Benjamin, and Kargl, Frank

In 2026 IEEE Vehicular Networking Conference (VNC) 2026

Abstract
Conference Paper

Disbelief-Favouring Trust Discounting for Adversarial Multi-Hop Trust Assessment using Subjective Logic

Trkulja, Nataša, Erb, Benjamin, and Kargl, Frank

In 2026 29th International Conference on Information Fusion (FUSION) 2026

Abstract
Conference Paper

Hyperparameter Optimization-Based Trust Quantification for Misbehavior Detection Systems

Hermann, Artur, Trkulja, Nataša, Eisermann, Dennis, Erb, Benjamin, and Kargl, Frank

In 2025 IEEE International Conference on Intelligent Transportation Systems 2025

Abstract DOI

Vehicular communication via V2X networks significantly improves road safety, but is vulnerable to data manipulation, which can lead to serious incidents. To address this threat, misbehavior detection systems (MBDs) have been developed to detect such misbehavior. In order to enhance the detection of data manipulation, trust assessment in V2X networks has recently gained increasing attention. Trust assessment takes into account the output of various security mechanisms such as MBDs or Intrusion Detection Systems (IDSs) to detect misbehavior. One particular challenge in trust assessment is the appropriate quantification of the output of these security mechanisms into trust opinions. In this paper, we propose a trust quantification methodology that transforms the output of an MBD into a subjective logic opinion. Furthermore, we apply a hyperparameter optimization approach to determine the optimal parameter set for an MBD. Our evaluation using three MBD variants shows that the optimization approach significantly increased the detection-performance of all MBDs. The MBD variant that used the optimization approach and our proposed trust quantification methodology achieved the best performance, increasing the F1 score by over 13% compared to other state-of-the-art MBD variants analyzed in this work.
Conference Paper

Demo: Quantifying Trust in a Trust Assessment Framework

Hermann, Artur, Trkulja, Nataša, Meißner, Echo, Erb, Benjamin, and Kargl, Frank

In 2025 IEEE Vehicular Networking Conference 2025

Abstract DOI

Vehicular communication via V2X networks increases road safety, but is vulnerable to data manipulation which can lead to serious incidents. Existing security systems, such as misbehavior detection systems, have limitations in detecting and mitigating such threats. To address these challenges, we have implemented a software prototype of a Trust Assessment Framework (TAF) that assesses the trustworthiness of received V2X data by integrating evidence from multiple trust sources. This interactive demonstration illustrates the quantification of trust for a smart traffic light system application. We demonstrate the impact of varying evidence coming from a misbehavior detection system and a security report generator on the trust assessment process. We also showcase internal processing steps within our TAF when receiving new evidence, up to and including the eventual decision making on the trustworthiness of the received V2X data.