Skip to content

Latest commit

 

History

History
30 lines (20 loc) · 2.25 KB

2018.04.03-Algorand.md

File metadata and controls

30 lines (20 loc) · 2.25 KB

Algorand: Scaling Byzantine Agreements for Cryptocurrencies

Yossi Gilad, Rotem Hemo, Silvio Micali, Georgios Vlachos, and Nickolai Zeldovich

Summary

Cryptographic currencies are becoming more and more popular. Existing cryptocurrency systems suffer from a trade-off between latency and confidence in a transaction. In current Bitcoin design, the user has to wait for an hour before a transaction gets confirmed. In this paper, the authors design and implement Algorand, a new cryptocurrency that can lower the latency to the order of a minute with a negligible probability of forking. Algorand mainly improves the consensus protocol of Bitcoin. It uses a new Byzantine Agreement protocol with Verifiable Random Functions. Apart from these, Algorand addresses three challenges using techniques like weighted users, consensus by committee, cryptographic sortition, and participant replacement. At last, the authors evaluate the performance of Algorand. The experimental results show that Algorand achieves sub-minute latency and 125x the throughput of Bitcoin. Furthermore, Algorand can scale well to 500000 users.

Contributions

  • This paper presents Algorand, a new cryptocurrency, which largely reduces the latency of transaction confirmation.
  • Algorand uses a new Byzantine Agreement (BA) protocol to reach consensus among users on the next set of transactions.
  • The authors also implement a prototype of Algorand and BA.
  • The authors conduct practical evaluations on 1000 Amazon EC2 VMs and demonstrate the effectiveness of Algorand.

Strengths

  • Algorand has promising scalability. It uses a novel mechanism called Verifiable Random Functions (VRFs) to randomly select users in a private and non-interactive way.
  • Algorand's agreement protocol allows users to reach consensus on a new block with low latency and without the possibility of forks.
  • Algorand addresses three challenges: (1) Sybil attacks, (2) scalability, and (3) denial-of-service attacks.

Weaknesses

  • This paper does not cover incentive mechanism.
  • The authors do not discuss forward security in this paper.
  • Algorand is still challenging in terms of bootstrapping cost.
  • Algorand requires lots of signatures to produce a final certificate for a block, which is a very high threshold.