Stealth beats Ethereum to Bonded Proof of Stake
For three weeks, the Stealth cryptocurrency has been running on one of the most efficient blockchain protocols possible, delivering dependable five second transaction times, while at the same time offering an array of features necessary for a modern cryptocurrency. These features include multi-signatures, feeless transactions, and scripted spending conditions.
Stealth is able to achieve this combination of speed and features using a novel blockchain technology that can be described as bonded proof-of-stake, which will be abbreviated as bPoS herein.
To understand bPoS, it is necessary to understand proof-of-stake (PoS). Proof of stake is a blockchain protocol wherein blocks of transactions are validated by entities (individuals or institutions) that have stake in the network. The original PoS protocol allows pseudo-randomly chosen validators to prove ownership of a quantity of stake of almost any value, from small to large. The likelihood that a validator is chosen to win any given lottery is proportional to the validator’s stake. This proportionality makes the protocol fair in that the more stake one has, the more blocks one signs and hence the more block rewards one can claim. This fairness even extends to the projected relative wealth of different participants because fractional ownership of the money supply for any one participant should remain approximately constant.
Although fair, this original PoS protocol has inefficiencies arising from the lottery-like nature of validator selection. Any lottery system is a de facto competition, and competitions are inherently inefficient. An example of this inefficiency comes from communication latency, where two parts of the network may temporarily disagree on the lottery winner. Although traditional PoS can eventually achieve consensus on the winner, this process of resolution takes time and results in long and inconsistent transaction times.
The most straightforward way to eliminate a lottery competition is to have a method to know the full validator set and to assign these validators slots in a queue. This technique is called block scheduling and has been employed by a blockchain protocol called delegated proof-of-stake (dPoS). In dPoS, stakeholders vote on a set of validators, and those with enough votes are put into a randomly ordered queue.
Efficiency in dPoS arises because validator selection (stake-weighted voting) is separated from block validation. dPoS results in consistent transaction times, but has game theoretic problems in that voting weight tends to consolidate with a few large stakeholders (or one very large stakeholder) who can undermine the interests of smaller stakeholders, sometimes with ghastly results. In one egregious example in the Steem cryptocurrency, the majority stakeholder used its voting weight to strip the assets of many smaller stakeholders.
One reason dPoS enables such injustice is that it requires an account-based ledger. Indeed accounts are necessary to tally voting, but accounts also make it easier to target specific stakeholders, as seen with Steem.
The type of ledger used with XST is different from an account based ledger, and is known as a UTXO ledger. Here each transaction is a bearer instrument, similar to an electronic traveler’s cheque. To spend a UTXO (i.e. electronic cheque), user A signs it and writes (1) a new check to user B and (2) another check back to user A as change. Because a user can have virtually unlimited pseudonyms, it becomes uncertain which UTXOs (cheques) are user B’s received payments and which are user A’s change. The ambiguity builds up over time, making it practically impossible to target individual stakeholders with absolute certainty.
The ambiguities inherent to a UTXO ledger seem to require an open validator set, precluding a known queue. One question, therefore, is if there is any way for a UTXO system to have a known validator set that could be assembled into a queue without a lottery. The answer is that it is possible, using non-fungible tokens (NFTs) that confer validation rights. These NFTs are called StealthNodes. In this system, the StealthNodes are shuffled into a queue, and each signs a block with a key pair assigned to the StealthNode by its owner. This key pair is called a “delegate”, but it is not a delegate for stakeholder votes, it is a delegate for the StealthNode owner.
This type of delegation is critically useful because it means the StealthNode owner can outsource its operation without worrying about the operator’s stealing the StealthNode. Additionally, the Stealth blockchain protocol handles any remittances owed to the operator, simplifying ownership considerably. This type of owner-operator relationship is unique to Stealth.
In the Stealth consensus protocol, called Junaeth, StealthNode NFTs are purchased from the blockchain in a transaction that simultaneously destroys the purchasing funds while minting a new StealthNode. This burn-mint process means (1) the purchasing XST (Stealth’s native currency) is completely removed from the money supply, and (2) StealthNodes can not be converted back into XST. Because of the latter consequence, StealthNodes are permanently bonded validators. Stealth is the first blockchain protocol to use bonded validators, and the only blockchain protocol to use permanently bonded validators.
Ethereum also plans bonded validators in its upcoming PoS protocol, but it has yet to be released. Also, Ethereum validators lack some critical features. First, validator owners cannot outsource operations without the risk of theft. Second the validator set is open because validators can be converted back into Ethereum. In other words, Ethereum bonded validation is not permanent, requiring a set of complicated and untested rules to allow entry and exit into the validation pool. Because Stealth allows only entry, the validator set will quickly reach an equilibrium based on opportunity costs and yields. As a result, entry will become economically prohibitive over time, eventually stabilizing the validator set.
In summary, Stealth has beaten Ethereum to bonded PoS block validation. Additionally, Stealth’s type of bonded validation is permanent, making the protocol management of the validation queue simpler and less prone to exploits. Moreover, Stealth offers a number of features that improve the experience and security of validator ownership. Owners can delegate operations without risking their assets. Finally, although it was not discussed above, management of a StealthNode can be separated from ownership, allowing the owner key to be kept in cold storage, offering the highest level of security possible.
— — — — — — — — —
Dr. James Stroud PhD
— — — — — — — — —
