The developers and supporters of Bitcoin and Ethereum are aware of the weakness of their ecosystems and are trying to eliminate it through various updates. As for Bitcoin, here the main challengers for solving the problem of scalability are the technologies of Lightning Network and RootStock. In the case of Ethereum, at the top of the list are Sharding, Plasma and Caspar technologies.
Lightning Network or Sharding is a kind of add-on over the main blockchain network based on the following idea: not all network participants need to know all the information about transaction history in order to maintain effective network synchronization. This idea is the basis of the directed acyclic graph (DAG) approach.
The blockchain is based on a "vertical architecture", while the DAG operates on a "horizontal" scheme. In a blockchain network,
transactions are grouped into new blocks, which are then added to the blockchain chain. In a “horizontal” DAG network, transactions are directly linked to other transactions, grouping them into blocks.
How DAG works
Directed. When a payment is made on the DAG network, it creates a connection with two (or more) last payments made on the network before it checks the balance and authenticity of older payments, and if everything is fine, the new payment confirms them.
Since new payments always check old ones, it is believed that the older the payment, the more “legitimate” it is and the harder to be cancelled. This is what is meant by the word “Directed” in the abbreviation DAG.
Acyclic. Literally means "not in a cycle." In the case of networks where new payments confirm old ones, this means that an old, already verified payment never checks a new one. That is, verification always goes ahead and never creates a similarity of at least some cycle.
. The name of the payment and connection schedule.
In the graph below: parallel lines are the history of records of nodes (users), circles are events (transactions), lines between events are connections from new events to old ones.
Since the DAG structure is similar to the web of verification, it is often called a tangle. And although this is not a blockchain,
Tangle has the same properties as a decentralized blockchain: a distribution registry based on a peer-to-peer network. Consequently, Tangle, like the blockchain, is a mechanism for validating distributed decision making.
Tangle Network Principles
Tangle (IOTA cryptocurrency) is created by linking individual transactions into a single peer-to-peer network. Relationships between transactions are formed on the basis of a simple rule: for a new transaction to be confirmed, it must confirm two older transactions, each of which confirmed two other even older transactions.
In the picture: green squares - confirmed transactions; red - unconfirmed transactions that confirmed two older transactions; gray - unconfirmed transactions that must confirm two older transactions (they are also called “tips”).
The connection between new and old transactions is established using the Monte Carlo method according to the Markov chain scheme (MCMC), which randomizes the transaction selection process for confirmation and ensures that participants will not verify their own transactions.
Unlike the Bitcoin or Ethereum blockchain, where transactions are confirmed by miners in exchange for a reward: a reward for creating a block + a percentage of a transaction is a commission. In Tangle, the function of processing and approving payments is assigned to all active members of the network. Every user who wants to make a transaction confirms two other new transactions and indirectly confirms the entire transaction history.
Thus, the Tangle network is free from the need to pay miners to confirm transactions, as a result of which there are no transaction fees on the network. And since there is no need to create blocks, transactions are confirmed very quickly: the more active participants, the greater the network bandwidth.
To avoid double costs, Tangle (cryptocurrency IOTA) uses a concept called “Illustrated Introduction”. Its essence is that the system counts all transactions of the user and does not allow him to spend the same coin twice. If the user has two unconfirmed transactions, the amount of which exceeds his balance, the system selects a chain with more “tips”. Chain with a lower rating "cuts off".
The impact of the number of users on the bandwidth and Tangle network bandwidth (IOTA). Source of the image
The combination of DAG and Illustrated Introduction provides Tangle-based projects with a large potential for scaling, because unlike the blockchain, where are limitations on network bandwidth,
the power of the Tangle network increases with the number of active users. At the moment, the speed of the largest networks Tangle, IOTA and Nano is 1000 and 7000 TPS, accordingly.
Hashgraph: the latest generation of DLT
Hashgraph is another way to build a distribution ledger (DLT). It was developed by Leemon Baird in mid-2016, focusing on the needs of the private corporate sector.
Intellectual property on Hashgraph is owned by Swirlds, so the software for creating private DLTs is called Swirlds Hashgraph.
Advantages of Swirlds Hashgraph:
- speed - up to 100,000 transactions per second;
- immutability of transaction history;
- resistance to DDoS attacks;
- disposable DLT.
Swirlds Hashgraph is not a blockchain, so cryptocurrency HYIP has bypassed technology. But it was noticed by banks: a consortium of 6,000 banks in North America, formed by the Union National Association (CUNA) and the Mountain West Credit Union Association (MWCUA), chose Swirlds Hashgraph to create a corporate network common to all.
Thanks to this success, Swirlds decided to launch the Hedera Hashgraph Platform. In fact, this is an analogue of the Ethereum network, only without the traditional blockchain and with a different management model. At the top of the food chain of the platform is the Hedera Governing Council, which has 39 members (private and legal entities), and they will determine the path of development of the project. Swirlds management picks board members. Five council members are already selected: T-Labs, DLA Piper, Nomura Holdings, Swisscom Blockchain and Luiza magazine.
Thus, the management of Hedera Hashgraph will be carried out according to the type of a multinational company (consortium), which more closely resembles the Visa management model, rather than Bitcoin and Ethereum. The same applies to intellectual property: hard forks in the Hedera Hashgraph network are impossible, because the software is protected by copyright.
How Hedera Hashgraph Works
As with Tangle, Hedera Hashgraph is not a blockchain. Unlike the blockchain, where information is stored in blocks, Hedera Hashgraph information is stored in hashes (hence the name “hashgraph”), which describe certain “events”.
Within each event, transaction records that are known to one participant and supposedly unknown to other participants: a timestamp, two different parent hashes, and one or more transactions. Like Tangle, Hashgraph
creates an extensive pattern from events (or transactions), where transactions are arranged in chronological order, so that their history can be traced.
The operation of the platform and the exchange of hashes between participants takes place via the Gossip protocol:
when an event occurs, the node transmits data to it to two other random nodes, which transmit them to two other nodes (in the total already four), and so on. This leads to an exponential spread of information throughout the network.
However, simply distributing information is not enough to reach a consensus on the total amount of information on the network. To do this, each participant in the system (node) must know the transaction history, that is, the exact chronology of all individual transactions (transaction timestamps).
In Hashgraph, the “Gossip about Gossip” consensus algorithm is responsible for this: each node on the network shares all of its information about which node, when and with whom it communicated or, technically speaking, each device shares its own data on the hash graph (exact order of all transactions ever performed on the Hedera Hashgraph network).
Since each node always has a current hash graph, it always knows the entire transaction history. This circumstance makes it possible for Hashgraph to be called “virtual voting”: since all nodes have a copy of transaction history and information about who received the information at a given time, each node can calculate how the other nodes behave.
That is, each node knows in advance the decision of the other without making an active, effective decision (vote). This “vote without a vote” and allows you to reach a network of consensus. In a short time and without the need for a resource-intensive procedure for coordinating nodes among themselves.
Mathematical proof of stability
Interestingly, the consensus algorithm used in the Hedera Hashgraph network has been used in practice for more than 35 years (in a slightly different form). And it is so well made that it has a “mathematically confirmed level of security.” That is, according to mathematical calculations, it is very difficult (if not impossible) to outwit.
That is why the developers of Hedera Hashgraph cite mathematical proofs of reliability, saying that their network is the only DLT technology with asynchronous Byzantine error tolerance (A-BFT, or aBFT). This means that while less than one third of the network participants do not want to fool the network, the system can always find a consensus on the network status and transaction history. This is considered to be the highest degree of security that an algorithm of consensus can provide.
The future of Hedera Hashgraph
Safety. As a form of DTL, Hashgraph can change the structure of the modern Internet. The fact is that the Internet in its current form has a number of birth defects, the main of which is centralized servers and centralized data storage. It is because of this defect that hackers, BotNet, hacks, spam and DDoS attacks became part of the daily life of any Internet user.
Hashgraph can “cure” this defect by creating ecosystems where information is stored in the clear and is protected from hackers, falsification and DDoS attacks. Such systems will be able to serve anything - from the school network to the Pentagon’s security system. And almost everyone will be able to run them.