The process of validating transactions, adding new a block, solving a crypto puzzle, and minting (i.e., creating) new coins in Proof‐of‐Work (PoW)‐based blockchains (BC) is called mining.
For example, in the Bitcoin BC3, a new block is added to the BC every 10 minutes by the fastest miner that provided a solution that satisfies the PoW puzzle. In order to find this solution, a miner must spend significant computational power to calculate several possibilities (hash values) based on the information (e.g., transactions, headers, and timestamp) from the current block and the proposed one. Moreover, it competes with other miners to find a solution; thus, the more computational power it has, the more chances it has of finding a solution. As soon as the miner finds a solution, it broadcast the block with the solution to the BC network so that other miners can verify the correctness of the solution (this verification is straightforward as it requires just the calculation of one hash value). If the solution is correct, the BC protocol creates a new transaction containing the reward to miner who found the solution. This is an interactive process and starts again once the block is included.
Even though that this mining process, based on PoW, secures the BC network against doublespending, it generates a sustainability concern because of the amount of energy wasted to find the correct solution to the puzzle. As of September 2019, the Bitcoin network consumes 73.12 TWh of energy (depicted in Figure 1), which is compared to the consumption of the country of Austria.
In contrast to PoW, where the probability of the miner to find a solution to the puzzle is related to the amount of computational power that it possesses, in PoS‐based mining, this probability depends solely on the number of coins a node is willing to put at stake to secure the BC. In this sense, nodes with more stake are more likely to be chosen to validate transactions and include a block than nodes with fewer coins. This concept is based on the fact that nodes holding more coins in the BC would contribute to its security, as they have more to lose; in turn, wealthier nodes tend to become wealthier as they verify more blocks and collect transaction fees.
As the choice of the next miner is not based on computational power, but rather on wealth, the energy consumption of such a mechanism is negligible compared to PoW.
Delegated PoS was proposed to solve the issue of “nothing‐at‐stake” by introducing a democratic process in the selection of nodes that will add blocks in the BC. In dPoS, the probability of the node to be selected as the node, which will include the block depends on the wealth staked and on votes provided by other nodes in the BC. In contrast to PoS, it is not enough for a node to have staked wealth, it must also behave in a manner that ensures a correct state of the BC (e.g., only verifying transactions in a single fork). If a node starts to behave maliciously, other nodes in the BC can remove their voter and place in a node that is behaving correctly.
One problem that arises with the introduction of voting is that nodes might offer financial rewards for other nodes to place their votes in a particular node6, i.e., buying votes. Moreover, the set of verifiers nodes in dPoS‐based BCs is often restricted to a pre‐defined number (e.g., 21 verifiers in the case of EOS). Thus, leading a centralization of power in the BC.
PoA is a straightforward miming scheme in which only a set of “pre‐authorized” nodes have the right to write in a new block in the BC. This type of scheme is often found in private BCs, where a selected group of individuals are allowed to interact with the BC. The problem of such a scheme is that it is highly centralized, as the number of block verifiers (i.e., miners) is limited and known. Thus, they can perform malicious actions, such as censuring transactions.
Table 1 presents a comparison between the described alternatives to PoW‐based mining. It can be seen that as the energy consumption lowers, the level of centralization increases. This is due to the fact that whenever there is a set of trusted participants securing the BC, the consensus mechanism can be simplified (e.g., PoA). In contrast, when the participants are not known and do not trust each other, the consensus mechanism must ensure that it is computationally (e.g., PoW) or financially expensive (e.g., PoS) to propose a new block to be added in the BC.
Nevertheless, the academia and industry are engaged in providing consensus mechanisms6 that are both secure and energy efficient to be used in novel alternatives to PoW‐based mining.