13 January 2020

Together on the blockchain: Finding consensus in a decentralised network

An essential element of the blockchain and the technology behind it is the validation of transactions, i.e. confirmation that data in the blockchain ledger is entered rightfully1. But how, exactly, can data entered in a decentralised ledger be validated? Who does the validating? These questions arise once you look deeper into the workings and structure of transactions on the blockchain.

The idea behind distributed ledger technology (DLT) (e.g. a blockchain) is to have a decentralised ledger that is not controlled by a single person/authority but by the network itself. This results in a trust problem, since many network participants usually have to trust a central "trusted party". To avoid having a central authority controlling the decentralised network, consensus algorithms are used to determine if book entries were validly made.
This is not a new issue. It has existed since the dawn of distributed computing (e.g. for synchronising distributed clocks).
There are several ways to reach consensus on a distributed network. In the DLT/blockchain world, the most popular applied consensus mechanisms are:

  • The Proof of Work (POW) mechanism, which is the first blockchain consensus mechanism and was initially used by Bitcoin. The execution of a proof of work mechanism (which is essentially the attempt to reach a certain result of an arithmetic problem) in the context of blockchains is called "mining". The miners try to find a result with certain properties by performing billions of arithmetic operations. If a miner reaches the correct result, the miner will be remunerated. The process of recording a transaction on the blockchain is as follows: (i) transactions are grouped into one block; (ii) the miners check whether these transactions are legitimate by performing the proof of work calculations; (iii) the first miner who to find the solution receives the block payment; and (iv) the validated transactions are appended to the blockchain in the form of a new block.
  • The Proof of Stake (POS) mechanism, based on the idea that those members on the blockchain who own the most tokens have an interest in keeping the network maintained. The decisive factor is therefore the stake of a user, i.e. the proportion of the total amount of tokens they own. The larger the share, the more likely it is that this user will be selected to mine the next block. Broadly speaking, compared to proof of work, the proof of stake mechanism is more like a joint-stock corporation. Whoever owns a larger share in the company normally receives more voting rights entitling them to make decisions.

Proof of Work vs. Proof of Stake
These two mechanisms are quite different. POW calls for validation by the first person to solve the arithmetic problem, meaning that many people are competing at the same time and an enormous amount of computing effort is required to solve the same problem, a lengthy, resource intensive and expensive process. POS is more environmentally friendly, since the user to be validated is randomly selected depending on the token they hold. The higher the proportion of tokens, the higher the chance of being selected. In addition, the validators of a POS mechanism are incentivised to maintain the network, as they actually hold coins of the blockchain on which they are validating, whereas in a POW system, a miner may own none of the coins they are mining and simply wants to maximise their profits without improving the network.

In short, one of the key issues of a decentralised network is how to find consensus. Practical mechanisms such as POW and POS exist, but they are not perfect (high energy consumption, ability to influence transactions). We will see in the future which mechanism will prevail and allow people to work "together" on the blockchain.

1 What is Blockchain?



austria vienna