Hashgraph Consensus Mechanism

Hashgraph Consensus Mechanism

DEFINITION of the Hashgraph consensus mechanism

Hashgraph is a new type of consensus mechanism that builds consensus using the blockchain concepts of gossip, gossip about gossip and virtual voting. It scores compared to other standard consensus-finding algorithms, such as Proof of Work (PoW), in terms of better speed and efficiency, as it does not send any votes or details over the network, which often results in congestion and delays.

Breaking the Hashgraph consensus mechanism

The Hashgraph consensus uses the Gossip protocol and is used in the Hedera blockchain platform. Blockchain participants using gossip protocol relay information (called gossip) on transactions, and they also chat about gossip. A collaborative history of “gossip events” is maintained while participants continue to add information about their previous gossip to each current gossip message.

Harnessing the Power of Gossip

In technical terms, for a blockchain, gossip is the information relayed by each participant repeatedly to another member chosen at random, and telling them everything they know about the transaction. It can be used to transfer a wide variety of information that needs to be distributed, such as gossip about user identities, transactions, or gossip about blocks in the blockchain.

An example: suppose a social gathering where many friends meet and chat during breaks. Let’s say Peter gives gossip to Paul on a subject during the first break, and Paul tells Pamela. At the second break, Pamela may have relayed this information to others, such as Priscilla, so even though Peter (who initiated the gossip) hasn’t spoken directly to Priscilla, she knows Paul’s gossip. At each break, this gossip is known to roughly double the number of people who knew him during the previous break. The Gossip protocol attempts to use this mechanism to build consensus on the blockchain as more and more people learn about the details in whole or in part.

Hashgraph is a data structure that keeps records of who chatted to whom and in what order, i.e. the hashgraph becomes a collaborative story of gossip events as participants continue to add information about their previous gossip with each current gossip message. It becomes “gossip about gossip” because it is the story of gossip itself. Since each member gets a copy of the Hashgraph, each member can calculate what the other member would know or could have sent.

Suppose Peter and Paul are two participants on the blockchain platform which uses the hash consensus mechanism. When a new transaction is placed on the blockchain, it will start to spread to other members. Peter will be aware of the transaction, but will not send any direct information to Paul about it. Peter will know when Paul learned of the transaction and will know when Priscilla learned that Paul had learned of the transaction. For his part, Paul calculates the information that Peter may have sent based on his own prediction of what Peter may know. This predictive calculation by Paul is based on what and when Peter would have learned of the transaction, according to the history available in the hashgraph.

Since no important information is transmitted across the network and all calculations are done by different participants on their own, this keeps network congestion to a large extent. This is a virtual vote – instead of each member sending their information (votes) directly to the other member, each calculates what others can know. Without a single vote in reality, there are very few network communication costs beyond the transactions themselves. Gossip (details of a transaction) and gossip about gossip (details of transaction details) lead to a reliable hash data structure, allowing to build a sufficient consensus to authenticate a transaction (or for any other consensus requirement)).

The hashgraph algorithm aims to achieve equity, because it has complete asynchrony, no dominant leaders, no alternating and high speed operation without the possibility of faults.

Leave a Comment

Your email address will not be published. Required fields are marked *