An on-chain governance model based on particle swarm optimization for reducing blockchain forks

Abstract

Blockchain technology has emerged as one of the most promising technologies of the past few years. However, it has encountered a number of challenges, including forks and low processing rates when compared to other payment systems. It is a natural process for most public blockchain systems to fork over time since forks are a natural part of the process. In order to resolve these issues, the side fork transactions will be discarded during chain reorganization (reorg) after these issues have been resolved. As a result of the requirement for clients to wait for more confirmations before being able to complete transactions, there is a slight delay in the process. The probability of a blockchain network forking to the greatest extent possible should be reduced as much as possible. This will improve the security, speed, and efficiency of the blockchain network. To the greatest extent possible. Understanding the behavior of blockchain systems is crucial to optimizing parameters. In this reserach, we propose a novel fork model that incorporates previously unconsidered parameters, such as the network delay and degree of validation. We have conducted a series of experiments using a blockchain simulator on the Ethereum network and the EIP-1559 specification to verify the validity of our proposed method. It is less likely that a fork will occur if the validation degree is decreased and the marginal cost of the miners is increased (as in EIP-1559). This part indicates that the probability of a fork occurring is reduced by approximately 10%. Further, the experimental results demonstrate that our method is highly accurate in predicting forking probabilities. By developing sharding, we were able to address the problem of low processing rates, but it also enhances the scalability of the network by making it more efficient. There are still some questions that need to be answered, however, regarding the impact of this on the likelihood of a fork. The second objective of this thesis is to determine whether adding new shards to a blockchain impacts the likelihood of forks in that blockchain as a whole. As a first step towards achieving this goal, a new simulator that facilitates simulation of sharded networks has been developed as a first step towards achieving it. We then examined the impact of sharding on the occurrence of forks as a result of sharding. There have been a number of experiments conducted on two EIP-1559 enabled networks, each with 60 and 120 nodes, in which a number of experiments were carried out. As a result of adding one shard, it has been found that 60% of the orphan blocks are reduced on average when one shard is added. Furthermore, we present a model that reduces forks by 23% and 15%, respectively, when there are 60 or 120 nodes in the network. The processing speed of blockchain networks is one of the most significant disadvantages associated with these networks. In order to achieve this, it may be necessary to implement sharding, which is capable of solving this problem. This will lead to an improvement in the scalability of the network as a result. It was difficult to determine how the sharding process might impact the likelihood of forks arising as a consequence of the sharding process in the current study. In the third part of this thesis, we performed a number of experiments using 120 nodes on the network EIP-1559 to achieve this objective. Our analysis indicates that adding a shard to the system results in a 60% reduction in the number of orphan blocks. There has also been the development of a novel on-chain governance model that uses particle swarm optimization in order to reduce the likelihood of forks between different shards. Based on the results of our study, we are confident that the proposed on-chain governance model reduces the risks associated with forking and maintains a positive user experience. As a contribution to public blockchains, the thesis provides an adaptive blockchain with learning capabilities. Consequently, there will be a reduction in the likelihood of forks occurring, leading to a more scalable blockchain architecture. Based on the local data inputs that networks receive, this solution enables networks to learn the optimal configuration. Based on the experimental results, the proposed solution has demonstrated better performance and scalability than the current state of the art.

Date	7 Dec 2022
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Kaiwen Zhang (Supervisor)