Bachelor and Master Theses

To apply for conducting this thesis, please contact the thesis supervisor(s).
Title: Network-Aware Experimental Evaluation of Byzantine Fault Tolerant Consensus Protocols in Distributed Systems
Subject: Computer network engineering, Computer science, Embedded systems, Software engineering, Distributed systems
Level: Basic, Advanced
Description:

Distributed systems rely on consensus protocols to maintain consistency, reliability, and fault tolerance across multiple communicating nodes. In decentralized and distributed architectures, consensus mechanisms must tolerate network delays, packet loss, node failures, and potentially malicious behavior while maintaining acceptable system performance. Byzantine Fault Tolerant (BFT) consensus protocols are particularly important in environments where distributed nodes cannot fully trust each other, but these protocols often introduce significant communication and computational overhead.

This thesis investigates the behavior and performance characteristics of consensus protocols under varying distributed network conditions, with particular focus on Byzantine Fault Tolerant approaches such as Practical Byzantine Fault Tolerance (PBFT) and related state-machine replication mechanisms. The research adopts an experimental and systems-oriented methodology emphasizing communication overhead, scalability, latency sensitivity, and fault tolerance in distributed environments.

The study focuses on how network conditions influence:

  • consensus latency,
  • transaction throughput,
  • message complexity,
  • scalability,
  • fault recovery behavior,
  • resource utilization.

Experiments will evaluate system performance under different:

  • node counts,
  • network topologies,
  • delay configurations,
  • packet loss scenarios,
  • bandwidth limitations,
  • fault injection conditions.

The research does not aim to design new consensus algorithms or implement production-scale blockchain platforms. Instead, the thesis focuses on experimental evaluation and benchmarking of existing Byzantine fault tolerant consensus architectures under controlled network conditions.

The implementation and experimentation environment may utilize distributed systems and blockchain research frameworks such as BFT-SMaRt, Hyperledger Fabric, Mininet, Docker, and Linux-based network emulation tools depending on the selected experimental setup and feasibility considerations. Benchmarking, monitoring, and logging tools such as Prometheus, Grafana, or similar platforms may also be incorporated for performance analysis and system evaluation.

The student is not restricted to a specific implementation framework and may select alternative open-source tools, consensus platforms, or network emulation environments in consultation with supervisors based on the final research direction and experimental requirements.

Performance metrics may include:

  • consensus completion latency,
  • throughput,
  • communication overhead,
  • CPU and memory utilization,
  • network traffic volume,
  • recovery time after node failures,
  • scalability behavior under distributed workloads.

Expected outcomes include:

  • a reproducible benchmarking framework for permissioned blockchain systems,
  • comparative performance analysis of consensus configurations,
  • evaluation of resource overhead in distributed environments,
  • recommendations regarding suitability of consensus mechanisms for distributed traceability infrastructures.
  • The thesis contributes to the fields of distributed systems and blockchain infrastructure by providing experimentally grounded analysis of consensus behavior under realistic transactional workloads.
Start date:
End date:
Prerequisites:

  • Knowledge of distributed systems and computer networks
  • Programming experience in Go, JavaScript, or Python
  • Familiarity with Linux environments and Docker containers
  • Basic understanding of blockchain architectures and consensus mechanisms
  • Experience with performance evaluation or benchmarking is beneficial
IDT supervisors: Hossein Fotouhi
Examiner:
Comments:

The thesis is primarily experimental and systems-oriented. The work focuses on benchmarking existing blockchain infrastructures and consensus configurations rather than implementing new cryptographic protocols or production-scale applications. The project is suitable for students interested in distributed systems, blockchain infrastructure, performance analysis, and edge or embedded computing environments.
Company contact:

Spritju (Dr. Ahmad Karnama)

http://www.spritju.com/

https://www.linkedin.com/in/karnama/