Bitcoin light client

[adambor]

Create bitcoin light client based on the Solana implementation (atomiq-contracts-solana/btc-relay).

Verification

Contract checks blockheader consensus rules:

• correct PoW difficulty target
• possible difficulty adjustments
• previous block hash
• blockhash is lower than target (block's PoW)
• timestamp is > median of last 11 blocks
• timestamp is < current time + 4 hours

Storage

To save on storage costs, the blockheader data is emitted as an Event from the program, and only sha256 fingerprint of that blockheader data is stored on-chain.

Another storage costs saving mechanism used is pruning - only last X block headers are kept stored on-chain in a ring buffer. Where X is the pruning factor.

Transaction verification

As merkle roots of the bitcoin blocks from blockheaders are known, they can be used to verify that any transaction was included in a block by its transaction id and merkle proof. However due to pruning, this also means that transaction verification can only be done for transactions confirmed in the last X blocks. Where X is the pruning factor.

Forks

Should a fork on the bitcoin main chain occur, the contract provides a way for anyone to submit fork blockheaders, and they automatically become the main chain when their chain work is greater than that of a current main chain in the bitcoin relay program.

Possible attack vectors

Fake block headers

A party might start submitting valid bitcoin blockheaders to the bitcoin relay and not on the bitcoin main chain. However as those blockheaders must be valid a non-trivial amount of resources must be expedited on PoW. Cost of such an attack depends on whether there is at least 1 honest party submitting blockheaders to the relay:

• if there are no honest parties, the cost of faking 1 blockheader can be expressed as value of lost bitcoin block reward incurred due to not submitting the blockheader to the main chain, which is currently 3.125 BTC ~ 300k usd
• if there is at least 1 honest party, the adversary party needs to capture at least 51% of the bitcoin hashing power to overrun the honest chain submitted by honest party