A new study by a team of research scientists claims to have cracked Bitcoin’s scalability problem. And it does it by eliminating the need for consensus among systems to confirm a transaction, the researchers claim.
The study is detailed in a paper by researchers at the Federal Polytechnic School of Lausanne in Switzerland. The paper, which recently won the Best Paper award at the International Symposium on Distributed Computing in Budapest, describes a very different approach to validating a transaction in cryptocurrencies such as Bitcoin.
Up until now, consensus among systems has been considered a necessity to solve Bitcoin’s double-spending problem, a situation in which the possibility of a transaction being duplicated in a ledger is eliminated. Most approaches to tackle this problem involve achieving a quorum or agreement among all systems comprising a blockchain.
But a quorum has several drawbacks. It is expensive in terms of resources and consumes massive amounts of energy. Quorum can also be time-consuming, if a backlog of transactions piles up in a blockchain or if it relies on consensus among systems spread across multiple geographical regions.
The new study, however, suggests that a quorum is not necessary to validate a transaction. Instead, it aims for an agreement about the transaction from a random sampling of systems within a network.
Not another Bitcoin fork
Rachid Guerraoui, lead author of the study, stresses that the algorithm proposed in their paper is not a variant of existing consensus algorithms, such as proof of work or proof of stake.
“Proof of work is typically used to decide who decides the consensus value,” he told Decrypt, adding that their proposed algorithm also does not run nodes to elect leaders. Instead, it uses a gossip protocol, the same one that is used to communicate nonces or headers of different blocks in Bitcoin, to spread information about a transaction.
A small group initially “confirms” the transaction and communicates the transaction’s details to another, larger group which propagates it further to other groups and so on.
To accomplish this task, the system uses Contagion, a probabilistic algorithm that mimics the spreading of a contagious disease in a population. It consists of three sub-protocols—Murmur, Sieve, and Threshold. The three protocols are responsible for ensuring validity, totality, and consistency of a transaction and for sending the original sampled process to a bunch of randomly picked systems within the network. The size of the randomly sampled systems ensures whether the transaction is valid or not.
Guerraoui said that the size should be logarithmic with respect to the overall system size, meaning it should be large enough to ensure that hackers are not able to penetrate the system. However, it should also be smaller than a quorum, ensuring that it is no more than a representation of the overall network.
Besides reducing the amount of time and resources required to confirm a transaction, the Contagion algorithm also uses minimal energy. “The energy [used to propagate and validate a transaction] is that of sending messages on the Internet,” Guerraoui explained.
The team behind the paper plans to open source the protocol with the help of E.U. funding to disseminate the study’s findings. “People can then use our protocol to build cryptocurrencies that are ‘cheap’ to run,” said Guerraoui.
That might lead to more coins in an ecosystem already teeming with thousands of cryptocurrencies.