Thomas Lorünser

Progress in information and communication technology, together with the demand for digital transformation, continues to produce networks, systems, and platforms that enable collaboration and data-driven workflows at scale and beyond established trust boundaries

The increasing need for privacy in machine learning has driven interest in cryptographic methods that enable computation on sensitive data. Secure multi-party computation and fully-homomorphic encryption are two prominent approaches, but their practical trade-offs are often unclear for real-world use. This work presents a pragmatic benchmark and comparison of these technologies for machine learning inference. Using representative open source frameworks, we evaluate performance across basic operations, distance metrics, regression models, and common deep learning architectures. The evaluation accounts for realistic deployment conditions, including communication overhead and partial plaintext execution. The results show that secure multi-party computation scales well to large and complex models but is strongly affected by network latency, while fully-homomorphic encryption is easier to deploy and performs well for small models and regression tasks. We conclude that the two approaches are complementary rather than competing and provide concrete guidance to help practitioners select the most suitable technology for privacy preserving machine learning in practice.

This work addresses the orchestration of large-scale Quantum Key Distribution Networks (QKDNs) using Software Defined Networking (SDN). Building on ETSI and ITU specifications, common best practices and architectures are outlined. The main task of the SDN Controller is to aggregate technical key performance indicators (KPI) from the network and, based on these, select the optimal path. Multiple path selection algorithms, based on Dijkstra or a maximum-minimum capacity algorithm, with built-in load balancing are presented. The algorithms were tested in simulations and their performances, and tradeoffs, are discussed. Additional critical aspects related to SDN controlled QKDNs are discussed, such as query batching, multi-path selection and group key capabilities. An oblivious multi-party protocol is proposed for relay path selection in a multi-domain scenario, so providers don't have to disclose sensitive information about their QKDN. These contributions aim to enhance scalability, resilience and interoperability in quantum-secure network infrastructures.

This paper explores the feasibility and implications of developing a privacy-preserving, data-driven cloud service for predicting the energy consumption of industrial robots. Using machine learning, we evaluated three neural network architectures—dense, LSTM, and convolutional–LSTM hybrids—to model energy usage based on robot trajectory data. Our results show that models incorporating manually engineered features (angles, velocities, and accelerations) significantly improve prediction accuracy. To ensure secure collaboration in industrial environments where data confidentiality is critical, we integrate privacy-preserving machine learning (ppML) techniques based on secure multi-party computation (SMPC). This allows energy inference to be performed without exposing proprietary model weights or confidential input trajectories. We analyze the performance impact of SMPC on different network types and evaluate two optimization strategies, using public model weights through permutation and evaluating activation functions in plaintext, to reduce inference overhead. The results highlight that network architecture plays a larger role in encrypted inference efficiency than feature dimensionality, with dense networks being the most SMPC-efficient. In addition to model development, we identify and discuss specific stages in the MLOps workflow—particularly model serving and monitoring—that require adaptation to support ppML. These insights are useful for integrating ppML into modern machine learning pipelines.

Commitment schemes (CSs) are essential to many cryptographic protocols and schemes with applications that include privacy-preserving computation on data, privacy-preserving authentication, and, in particular, oblivious transfer protocols. For quantum oblivious transfer (qOT) protocols, unconditionally binding commitment schemes that do not rely on hardness assumptions from structured mathematical problems are required. These additional constraints severely limit the choice of commitment schemes to random oracle-based constructions or Naor’s bit commitment scheme. As these protocols commit to individual bits, the use of such commitment schemes comes at a high bandwidth and computational cost. In this work, we investigate improvements to the efficiency of commitment schemes used in qOT protocols and propose an extension of Naor’s commitment scheme requiring the existence of one-way functions (OWFs) to reduce communication complexity for 2-bit strings. Additionally, we provide an interactive string commitment scheme with preprocessing to enable the fast and efficient computation of commitments.

This paper explores the integration of advanced cryptographic techniques for secure computation in data spaces to enable secure and trusted data sharing, which is essential for the evolving data economy. In addition, the paper examines the role of data intermediaries, as outlined in the EU Data Governance Act, in data spaces and specifically introduces the idea of trustless intermediaries that do not have access to their users' data. Therefore, we exploit the introduced secure computation methods, i.e. Secure Multi-Party Computation (MPC) and Fully Homomorphic Encryption (FHE), and discuss the security benefits. Overall, we identify and address key challenges for integration, focusing on areas such as identity management, policy enforcement, node selection, and access control, and present solutions through real-world use cases, including air traffic management, manufacturing, and secondary data use. Furthermore, through the analysis of practical applications, this work proposes a comprehensive framework for the implementation and standardization of secure computing technologies in dynamic, trustless data environments, paving the way for future research and development of a secure and interoperable data ecosystem.

The federated analysis of sensitive time series has huge potential in various domains, such as healthcare or manufacturing. Yet, to fully unlock this potential, requirements imposed by various stakeholders must be fulfilled, regarding, e.g., efficiency or trust assumptions. While many of these requirements can be addressed by deploying advanced secure computation paradigms such as fully homomorphic encryption, certain aspects require an integration with additional privacy-preserving technologies. In this work, we perform a qualitative requirements elicitation based on selected real-world use cases. We match the derived requirements categories against the features and guarantees provided by available technologies. For each technology, we additionally perform a maturity assessment, including the state of standardization and availability on the market. Furthermore, we provide a decision tree supporting application developers in identifying the most promising technologies available matching their needs. Finally, existing gaps are identified, highlighting research potential to advance the field.