## Cryptology ePrint Archive: Report 2021/086

On Elapsed Time Consensus Protocols

Mic Bowman and Debajyoti Das and Avradip Mandal and Hart Montgomery

Abstract: Proof of Elapsed Time (PoET) is a Nakamoto-style consensus algorithm where proof of work is replaced by a wait time randomly generated by a trusted execution environment (TEE). PoET was originally developed by Intel engineers and contributed to Hyperledger Sawtooth, but has never been formally defined or analyzed. In particular, PoET enables consensus on a bitcoin-like scale without having to resort to mining. Proof of Luck (PoL), designed by Milutinovic et. al., is a similar (but not identical) protocol that also builds a Nakamoto-style consensus algorithm using a TEE. Like PoET, it also lacks a formal proof.

In this work, we formally define a simplified version of PoET and Proof of Luck, which we call elapsed time (ET) consensus with a trusted timer. We prove the security of our ET consensus protocol with a trusted gimer given an honest majority assumption in a model very similar to the bitcoin backbone model proposed by Garay et al. which we call the elapsed time backbone model. Our model and protocol aims to capture the essence of PoeT and PoL while ignoring some of the more practical difficulties associated with such protocols, such as bootstrapping and setting up the TEE.

The PoET protocol also contains a function called the $z$-test that limits the number of blocks a player can publish in any particular larger set of blocks. Surprisingly, by improving this $z$-test a little bit we can prove the security of our ET consensus protocol without any TEEs with a (slightly stronger) honest majority assumption. This implies that Nakamoto-style consensus with rate limiting and no proofs of work can be used to obtained scalable consensus in a permissioned setting: in other words, bitcoin without proofs of work'' can be made secure without a TEE for private blockchains!

Category / Keywords: cryptographic protocols / Blockchain, Consensus, Trusted Execution Environment (TEE)