Definition of Blockchain

There’s no unique definition of the term blockchain; there’s also a different usage of the term depending on its meaning. Bashir (2017)Imran Bashir. Mastering Blockchain. Packt Publishing, Limited, 2017. and Drescher (2017)Daniel Drescher. Blockchain Basics: A Non-Technical Introduction in 25 Steps. Berkely, CA, 2017. Apress. give various technical definitions, listed below.

Blockchain

A blockchain is:

• A distributed shared ledger of transactions. The transactions are ordered and grouped into blocks.

• A data structure. It’s basically a linked list that uses hash pointers in order to achieve a chronology of transactions.

• A decentralized (or distributed) consensus mechanism whereby all peers eventually come to an agreement regarding the state of transactions.

Any real-life application of a blockchain running system is a combination of all these points, namely a combination of a data structure and a decentralized consensus mechanism, which is applied in order to maintain the accuracy and consistency of data that is stored in a ledger, where the ledger is distributed via a purely peer-to-peer system. Drescher (2017) calls this “a suite of technologies," which is used to achieve the data integrity of a public shared ledger. We’ll introduce this approach in the following sections.

PoW vs. BFT blockchains

There are two different types of blockchains:

• Permissioned—private—or BFT blockchains: The central property of this type is that admission to the blockchain is restricted. Thus, there is a closed set of known nodes that participate in consensus. Since the nodes are whitelisted, there’s no risk of a Sybil attack, and therefore costly consensus protocols such as Proof-of-Work are not needed. Vukolic (2015)Marko Vukolic. The quest for scalable blockchain fabric: Proof-of-work vs. bft replication. In Jan Camenisch and Dogan Kesdogan, editors, iNetSeC, volume 9591 of Lecture Notes in Computer Science, pages 112-25. Cham, Switzerland, 2015. Springer. defines this type as Byzantine Fault Tolerant blockchains (PFBT).

• Permissionless—public—or PoW blockchains: The central property of this type is that the admission to the blockchain is unrestricted, thus everyone is able to join the network anonymously and thus is allowed to participate in consensus. Vukolic (2015)Marko Vukolic. The quest for scalable blockchain fabric: Proof-of-work vs. bft replication. In Jan Camenisch and Dogan Kesdogan, editors, iNetSeC, volume 9591 of Lecture Notes in Computer Science, pages 112-25. Cham, Switzerland, 2015. Springer. defines this type as Proof-of-work blockchains, since “permissionless participation is made possible by PoW, as PoW inherently addresses the Sybil attack”.

An important consequence is that the consensus mechanism has to be chosen according to the admission to the blockchain as they consider different adversaries. Thus, private networks, where the participants are whitelisted, offer a wider range of consensus protocols to pick from, because there’s no risk of a Sybil attack. One possible solution is the PBFTPractical Byzantine Fault Tolerance protocol.

However, in a fully open network, where everyone can participate anonymously, more costly consensus mechanisms like PoW are indispensable. BFT protocols only consider the number of corrupted nodes that may grab any number of votes. Thus, the BFT protocol comes up with its well-known assumption that an adversary holds less than $33 \%$ of the voting power (see this lesson). Christidis (2016)Konstantinos Christidis and Michael Devetsikiotis. Blockchains and smart contracts for the internet of things. IEEE Access, 4:2292-303, 2016. outlines that in a PBFT blockchain, the protocol determines “a ‘primary’ (leader) node that acts as block miner; the leader can be changed by the rest of the network via a so-called ‘view-change’ voting mechanism, if it crashes or if it exhibits arbitrary behavior (Byzantine faults)” (Konstantinos Christidis et al. (2016)Konstantinos Christidis and Michael Devetsikiotis. Blockchains and smart contracts for the internet of things. IEEE Access, 4:2292-303, 2016.)

On the opposite, PoW blockchains consider the total computational (hash) power that an adversary controls. A PoW-based blockchain tolerates any adversary which has less than $50 \%$ of computing power in the network at least theoretically. We will see in section that a PoW blockchain is vulnerable to any attacker hashrate.

A complete high-level comparison between PoW and BFT blockchains is given by Vukolic (2015)Marko Vukolic. The quest for scalable blockchain fabric: Proof-of-work vs. bft replication. In Jan Camenisch and Dogan Kesdogan, editors, iNetSeC, volume 9591 of Lecture Notes in Computer Science, pages 112-25. Cham, Switzerland, 2015. Springer., who concludes that PoW and BFT blockchains are placed at the two opposite ends of the spectrum of node scalability: whereas PoW-based blockchains offer good node scalability with poor performance (only a few transactions per second), BFT blockchain offers a good performance $\left(>10000\right.$ transactions per second$)$ for only a very small number of replicas $($ $n \leq 20$ nodes$)$.

Furthermore, Vukolic (2015) outlines that the main advantage of BFT blockchains over PoW is that they all satisfy consensus finality, meaning that a valid block that has been appended successfully to the blockchain will never be removed, which allows an “immediate confirmation of the final inclusion of a transaction into the blockchain." In contrast, consensus finality is not satisfied by PoW-based blockchains since they are prone to temporary forks that occur when different blocks have been discovered simultaneously by different miners and now compete for a place in the blockchain, even if all nodes are honest, as we’ll show in this lesson. Thus, a transaction needs to be followed by several blocks in order to increase the probability that it will not end in an orphan block and therefore will be removed from the blockchain.