SPECTRE Protocol: DAG Structure to Solve Scalability?
SPECTRE, described as a “fast and scalable distributed ledger protocol,” could be the solution to some of Bitcoin’s woes. Utilizing a Directed Acyclic Graph (DAG) similar to ByteBall, SPECTRE, an acronym for “Serialization of Proof-of-work Events: Confirming Transactions via Recursive Elections,” claims to increase the throughput of the network, confirmation times, and be as resilient, if not more so, against 50 percent attacks.
SPECTRE can make the claims of several blocks created per second due to this DAG structure rather than the typical blockchain seen in cryptocurrencies. Miners can create multiple blocks simultaneously, where reference blocks mined in parallel as well as past blocks.
“Just as with the Bitcoin protocol, the probability of a successful double spend decreases exponentially with each block that is built above the transaction. The difference here is that blocks in SPECTRE are created extremely fast.”
The Bitcoin protocol disregards parallel blocks if not set up on top of the longest chain; however, this utilization of parallel blocks not only increases both confirmation times and security but also decreases mining reward variance as well.
Since nodes no longer need to “reconcile their different worldviews regarding the identity of a selected chain at the time of block creation.” Consequently, blocks come into existence without agreement on the main chain or knowledge of the propagation delay to run a mining node, since transactions become approved after the fact, by anyone with a copy of the block DAG.
The paper, as well as the Medium blog post, discuss the more minute technical details regarding the protocol, such as further analysis of DAG structure as well as how SPECTRE handles conflicting transactions made by blocks mined in parallel.
SPECTRE hopes their efforts here will be implemented in other projects, such as off-chain transaction channels like the lightning network and easily allow cryptocurrencies to scale. If the protocol is implemented and can enable the creation of multiple blocks per second, the transaction per second capabilities of cryptocurrencies based on DAG would far exceed the throughput of major credit card companies such as VISA and American Express.
The solution to Bitcoin may not be by increasing the block size, but rather using a protocol like SPECTRE to expedite the creation of blocks. Scalability would be solved, confirmation times would increase, and the incentives are still there for miners. So far, no cryptocurrencies have announced utilization of SPECTRE nor have the people behind the project created a cryptocurrency to display the benefits.
Examination of whether there is a proof of work scheme that is more desirable than the blockchain is recognized as a central problem in the design of cryptocurrencies, as a recent paper on the potential of DAG structures by Boyen et al. highlights:
“The holy grail would be to design a consensus protocol with is ‘naturally’ low-variance by rewarding miners a small amount for lower-difficulty puzzles [. . . ] It remains an open question if there is an alternative version of the consensus protocol which would enable mining puzzles without near-instantaneous broadcast of all solutions.”