Yvonne-Anne Pignolet

Thomas Locher, Sebastian Obermeier, Yvonne-Anne Pignolet

The functionality that distributed ledger technology provides, i.e., an immutable and fraud-resistant registry with validation and verification mechanisms, has traditionally been implemented with a trusted third party. Due to the distributed nature of ledger technology, there is a strong recent trend towards using ledgers to implement novel decentralized applications for a wide range of use cases, e.g., in the financial sector and sharing economy. While there can be several arguments for the use of a ledger, the key question is whether it can fully replace any single trusted party in the system as otherwise a (potentially simpler) solution can be built around the trusted party. In this paper, we introduce an abstract view on ledger use cases and present two fundamental criteria that must be met for any use case to be implemented using a ledger-based approach without having to rely on any particular party in the system. Moreover, we evaluate several ledger use cases that have recently received considerable attention according to these criteria, revealing that often participants need to trust each other despite using a distributed ledger. Consequently, the potential of using a ledger as a replacement for a trusted party is limited for these use cases.

Qin Wang, Sandro Schoenborn, Yvonne-Anne Pignolet et al.

Currently, engineers at substation service providers match customer data with the corresponding internally used signal names manually. This paper proposes a machine learning method to automate this process based on substation signal mapping data from a repository of executed projects. To this end, a bagged token classifier is proposed, letting words (tokens) in the customer signal name vote for provider signal names. In our evaluation, the proposed method exhibits better performance in terms of both accuracy and efficiency over standard classifiers.

Seyed-Mohsen Moosavi-Dezfooli, Yvonne-Anne Pignolet, Dacfey Dzung

In this paper, we present an acoustic localization system for multiple devices. In contrast to systems which localise a device relative to one or several anchor points, we focus on the joint localisation of several devices relative to each other. We present a prototype of our system on off-the-shelf smartphones. No user interaction is required, the phones emit acoustic pulses according to a precomputed schedule. Using the elapsed time between two times of arrivals (ETOA) method with sample counting, distances between the devices are estimated. These, possibly incomplete, distances are the input to an efficient and robust multi-dimensional scaling algorithm returning a position for each phone. We evaluated our system in real-world scenarios, achieving error margins of 15 cm in an office environment.