October 13, 2022
A recent SCI Summer Scholars project overseen by Dr. Amy Babay, Assistant Professor with the Department of Computer Science and Department of Informatics and Networked Systems, has researched the efficiency of leader election in the consensus algorithm.
This project, titled “Optimization of Leader Election Based on Consensus Algorithm”, hopes to improve this system by rethinking the election timeout mechanism, which will improve overall web stability by removing errors caused when the system is unable to locate the leading node.
The consensus algorithm is instrumental in decentralized cryptocurrencies, forgiving faults such as packet loss and clock drift by running multiple nodes concurrently. Leader election, a key operation determining this algorithm’s efficiency, involves choosing which node to ‘lead’ a data packet transmission.
The potential impact of this project is substantial, with the global cryptocurrency market expected to reach 4.94 billion dollars by 2030. Improvements to the consensus algorithm will have wide reaching indirect effects, which will impact fields such as cryptography, economics, and even marketing.
Currently, Dr. Babay and researchers are investigating open-source Raft implementations on GitHub and analyzing them through diagnostic tools. By examining the code and referencing the diagnostic results, they hope to make inductive findings on how leader election can be adjusted and optimized. This research, along with analysis of the PreVote Mechanism, will allow SCI Scholars an opportunity to gain a new understanding of leader election. More specifically, researchers believe that timeout valve settings can be altered independently during the leader election process to reduce pointless leader elections and delays brought on by split voting.
Projects like this through SCI’s Summer Scholars program reflect its commitment to innovation in computer science and provides scholars opportunities for collaborative research.