State of the Art of Consensus Mechanism
Most of the consensus mechanisms currently used are based on the Proof of Work mechanism and the Proof of Stake mechanism. Proof of work can achieve block consistency (Han & Liu, 2017), but its large amount of computational power is almost all used to do useless work, which leads to the waste of computing power. Jake Billings (2017) proposed a proof of work mechanism based on the image of interest (Billings, 2017), which detects whether the image is meaningless while generating the block. The mechanism replaces the nonce generation part. In this process, nodes read the images and calculate their second-degree entropies. For the images whose second-degree entropies satisfy the specific requirements, the mechanism judges that it as an interesting image and read the image content for encoding to generate a nonce. The workload proof mechanism utilizes computational resources to some extent, but its application scenarios are rare and simple.
In addition to the method of proof of work, there is a consensus mechanism based on proof of stake. Different from the proof of work, it allows nodes to participate in elections by investing in virtual resources, thus obtaining consensus from distributed networks (Garcia-Alfaro et al., 2017). It requires the node to invest a certain amount of digital currency on the new block as a voucher to participate in the generated block. The consensus mechanism randomly assigns the qualifications for verifying transactions and generating blocks based on the proportion of funds invested by each node. This approach, to some extent, stops the node from verifying fraudulent transactions, which results in the loss of funds it invests. However, since the precondition of this method is that the node already has a certain wealth when it participates in the generation of the block, this method cannot be used at the beginning of the blockchain generation. Most consensus mechanisms that use the proof of stake approach also use the method of proof of work.