DPoS区块链共识机制的改进研究

针对委托权益证明(delegated proof of stake,DPoS)共识机制存在恶意节点相互勾结以及权益分配不合理的两大问题,提出了一种改进方案.首先,引入RBF神经网络模型,根据节点在整个区块链上工作表现的历史数据和相关属性计算综合信任值,使得通过综合信誉值选举出的节点更加权威可信;同时,加入基于动态博弈的信誉激励机制,增加了节点恶意攻击的成本,使得共识节点的出块更加安全.其次,利用沙普利值对节点权益进行合理划分,使得节点的权益得到了分散,增强了"去中心化"程度,降低了"财阀统治"的可能性.实验结果表明,改进后的DPoS共识机制能够有效抵御"腐败攻击",增强了系统的稳定性和安全性,具有潜在的应用价值.

Research on improvement of DPoS blockchain consensus mechanism

Aiming at the two major problems of the DPoS consensus mechanism that malicious nodes collude with each other and the unreasonable distribution of stakes, this paper proposed an improvement scheme. Firstly, it introduced the RBF, neural network model to calculate the comprehensive trust value based on the historical data and related attributes of the node''s performance on the entire blockchain, making the nodes elected through the comprehensive reputation value more authoritative and credible. At the same time, it added a reputation incentive mechanism based on dynamic games to increase the cost of malicious attacks on nodes and make the block generation of consensus nodes more secure. Secondly, it used the Shapley value to reasonably divide the stake of nodes, which enhanced the degree of "decentralization" and reduce the possibility of "plutocratic rule". Experimental results show that the improved DPoS consensus mechanism can effectively resist "corruption attacks", enhance the stability and security of the system, and has potential application value.