Collusion Attack Analysis and Detection of DPoS Consensus Mechanism

Abstract

AbstractWith the development of blockchain technology, the increasing safety accidents result in huge economic losses in blockchain systems. Delegated Proof of Stake (DPoS) selects the witness nodes to produce blocks by voting, leading to the quick confirmation of transactions. As one of the widely used consensus mechanisms in public blockchain, DPoS is still threatened by attacks. In this paper, an analysis method for collusion attacks of DPoS consensus mechanism is proposed. Meanwhile, we analyze the behavioral motivations of malicious nodes and detect the attacks that exist in the voting process of DPoS. First, the coalitional game is the basic form of cooperative game, which can be used to analyze the structure, strategy and benefits of cooperative game. We build a coalitional game model to analyze motivations of DPoS nodes that launched collusion attacks. And then we use the Shapley-Shubik power index and Banzhaf power index in weighted voting games of DPoS, which calculated different values that DPoS suffered attacks during the voting phase. Experimental results show that collusion attacks in DPoS can be effectively detected by this method. In addition, the analysis results can further contribute to the security of the DPoS blockchain system.KeywordsBlockchainConsensus mechanismDPoSCollusion attackCoalitional gameWeighted voting game