Analysis

Safety

The safety of BPX consensus is similar to that of other Nakamoto consensus algorithms like Bitcoin. There is no guaranteed finality, but the more confirmations a transaction has, the safer it is.

A transaction needs a certain number of confirmations for a receiver to assume that it cannot be re-orged, under the < 42.7% (* vdf advantage) colluding assumption. Since farmers can theoretically sign multiple blocks at the same height, more confirmations should be used in BPX than in Bitcoin. However, BPX doesn't require anywhere near as much clock time as Bitcoin for a transaction to be considered safe.

In BPX, there are two main reasons to wait for a certain number of confirmations:

  1. To be confident there won't be a chain re-org. A small re-org is a natural occurrence in blockchains, though rare in BPX.

To be confident that there won't be a chain re-org, you should wait for six beacon blocks to be created (around two minutes after the first confirmation).

  1. Just in case there is a foliage re-org attack. This type of attack would require an attacker to discover the identity of - and successfully bribe - a large and consecutive number of anonymous block winners. This attack would be difficult to pull off, so it is expected to be extremely rare, if it is ever even attempted.

If you want to be nearly certain that even a successful foliage re-org attack won't reverse your transaction, you should wait for 32 beacon blocks to be created (around ten minutes after the first confirmation).

It's worth noting that the 54% requirement only pertains to non-colluding space, rather than honest farming space. Profit-seeking farmers gain very little by deviating from the protocol.

There is the added assumption that at least one fast timelord must be connected to the non-colluding portion of the network, and that the attacker's timelord is not significantly faster.

Liveness

The liveness of the BPX consensus system is one of its greatest strengths. Like Bitcoin, the BPX system continues advancing even when a majority of the space goes offline. Unlike bitcoin, though, the system does not slow down significantly when this happens, since not all blocks are transaction blocks. Therefore transaction throughput does not drop significantly if many participants go offline.

The network will continue to advance even if only one farmer is online, although there will be many empty slots, since a transaction block can only be created if it's below the sub-slot iterations threshold.

Of course, in the event of a long-term network split, the effects are that one chain must be chosen, so there can be large re-orgs in this case. The network will automatically choose the heavier chain, similar to PoW.

Comparison to Nakamoto PoW

("+" means a pro for BPX)

Comparison to Proof of Stake

BPX consensus algorithm could also be used for Proof of Stake, where the space farmers are replaced by stakers who own coins in the system. The benefit would be the ability to slash (delete people's stake), and farmers would have "skin in the game", but there are some concerns if Proof of Stake is used. ("+" means benefit for using space vs stake).

Comparison to BFT consensus algorithms

Proof of Space could also be used as a Sybil-resistant mechanism in order to bootstrap a Byzantine consensus (k-agreement) system. Filecoin, and many Proof of Stake systems use aspects of Byzantine consensus.

The pros and cons of using BPX consensus vs Byzantine consensus, which vary from algorithm to algorithm ("+" means a pro for BPX):


Revision #3
Created 5 June 2023 14:45:48 by Admin
Updated 27 October 2024 10:33:49 by Admin