Institutsseminar/2025-09-26-3: Unterschied zwischen den Versionen

Version vom 24. Juli 2025, 12:26 Uhr

Termin in Kalender importieren: iCal (Download)

3SIM: Blockchain Simulator for Investigating the Blockchain Trilemma
Vortragende(r)	Davis Riedel
Vortragstyp	Bachelorarbeit
Betreuer(in)	Robert Heinrich
Vortragssprache	Deutsch
Vortragsmodus	in Präsenz
Kurzfassung	-- Robert Heinrich wird remote zugeschaltet -- According to the blockchain trilemma, the degree of decentralization (DoD), scalability, and security of a blockchain system cannot be maximized simultaneously. Software architects must identify Pareto-optimal configurations of blockchain systems to balance the trade-offs between DoD, scalability, and security, in order to meet the non-functional requirements imposed by their use case. I developed 3SIM, a Palladio-based discrete-event simulator for simulating blockchain systems. 3SIM provides a metamodel that enables software-architecture-focused modeling of blockchain systems, and a simulation model, that gathers the data necessary to compute several output metrics, which are suitable to investigate the blockchain trilemma. 3SIM extends SM-SIM, a blockchain simulator for simulating selfish-mining attacks. Experiments were conducted to demonstrate the plausibility and utility of results produced by 3SIM. The experiments help to gain a better understanding of how different blockchain system configurations influence the manifestation of the blockchain trilemma.

Quantifying Probabilistic Uncertainty in View-Based Modeling
Vortragende(r)	Claus Hammann
Vortragstyp	Bachelorarbeit
Betreuer(in)	Nathan Hagel
Vortragssprache	Englisch
Vortragsmodus	in Präsenz
Kurzfassung	Uncertainty poses a known thread to the success of development processes. In order to address uncertainty it first must be quantified and expressed within software models. Current approaches in uncertainty quantification within the software modeling domain do not consider view-based approaches like Vitruvius. This is crucial since the development often includes stakeholders from multiple domains. This thesis proposes a quantification approach based on a modified version of Palladio StoEx, a domain specific language for expressing stochastic expressions (StoEx). Furthermore a new interpreter was developed in order to simplify these expressions. The interpreter and StoEx language was then integrated into the UnCertaGator, an approach to express uncertainty within view based modeling, which is based on the Vitruvius framework. Applying this to a running case study of a brake system shows that it allowed stakeholders to take notice of potential issues early on and address them.