Research On Secure And Efficient Blockchain Consensus Mechanism

As a distributed database,blockchain has the technical advantages of openness,transparency,immutability,anonymity,etc.Based on these characteristics,blockchain technology has created a reliable "cooperation" mechanism,and has been widely used in ehealth,internet of things,social management,financial services,sharing economy and other fields in recent years.As one of the key technologies underlying blockchain,consensus algorithm is responsible for solving the consistency problem of computational data in distributed networks,i.e.,how transaction data can be consistently agreed among blockchain network nodes.The operational efficiency of the consensus algorithm directly affects the stability of the blockchain system,but there are problems in the existing blockchain consensus scheme such as low security,serious offline canvassing and negative participation of nodes,which hinder the promotion and development of blockchain technology.Therefore,from the perspective of improving the blockchain consensus algorithm,this thesis studies the classical distributed consensus algorithm and its improved algorithm,mainly including the following:(1)An efficient blockchain consensus scheme based on blinded voting is proposed to address the problems of offline bribery,node identity leakage,and vulnerability to attacks of the DPOS algorithm.Firstly,anonymous ballots are generated by using associative ring signature and double-blinding mechanism to hide candidate node order information and effectively resist denial-of-service attacks against master nodes.Second,according to the differentiated role history behavior credit evaluation model,the historical behavior of nodes under different roles is comprehensively evaluated and their current credit values are comprehensively calculated to ensure that candidate nodes participate in the consensus process honestly and efficiently;finally,the proposal block verification message is constructed through threshold signatures to achieve a Byzantine fault-tolerant consensus algorithm with high trustworthiness and low latency with reduced communication overhead.The experiments show that the blockchain system can maintain low latency while improving voting fairness and reducing the probability of collusion and denial-of-service attacks,providing algorithmic guarantees for efficient operation of the blockchain system.(2)Aiming at the problems of uneven distribution of benefits and negative participation in the existing Delegated Proof of Stake(DPOS)algorithm,combined with the differentiated role history behavior credit evaluation model,this scheme designs an improved DPOS algorithm model based on trust grouping.The algorithm model divides nodes into different groups based on credit evaluation value,and conducts voting election by groups,and sets up voting instant reward mechanism to solve the problems such as low motivation of nodes to vote in the existing scheme while improving election efficiency;introduces abnormal behavior prosecution mechanism and bookkeeping reward dividend mechanism to promote nodes’ income balance,strengthen mutual supervision among nodes,and ensure fair and reliable election process.Finally,the evolutionary game idea is applied to model the native DPOS and the improved scheme,respectively.Experimental analysis shows that the improved algorithm proposed in this thesis substantially improves user participation,effectively guarantees node income balance,and promotes the long-term benign development of the blockchain system.