Bayes Centre

Performance Optimization of Distributed Trust System Infrastructure Industry Collaborators

Since blockchain technologies have grown tremendously, their vast success can be attributed to both the research community and the industry.

Although, initially, blockchain was confined to financial sector only, its decentralized and immutable ledger availability has made it popular for non-financial services as well. The ever-increasing size of blockchains like Bitcoin, Ethereum, and so on has led to issues of scalability. Thus, high-scalable consensus becomes a crucial demand of extending the use of blockchain technologies to more industry applications.

To address the issue, the project aims to develop an extra-fast and high-scalable consensus algorithm based on a multilayer parallel BFT framework. The parallel BFT (ParBFT) algorithm allows consensus running parallelly in a group of committees, which ensures the hight throughput and scalability with the growth of peers. Moreover, the ParBFT is designed with an optimisation scheme that can support the efficient portioning of committees and allocation of peers. The optimisation scheme aims to improve the system consensus performance by minimising the failure of committees and peers while maximising the consensus throughput. Fig. 1 presents an overall framework of parallel consensus with an optimisation scheme.

 

Fig. 1 Parallel Consensus Framework with Optimisation Scheme

Fig. 1 Parallel Consensus Framework with Optimisation Scheme

 

The project contributes a high-efficient and scalable consensus algorithm, which can enhance the performance of blockchain systems with the following aspects:

  • Improve consensus scalability of BFT-like protocol to hundreds of thousand-level.
  • Improve system throughput to million-level.
  • Reduce consensus latency to millisecond-level.
  • Enhance consensus security without increasing energy cost.

 

Research Projects