Leveraging the Goldfinger Attack in Blockchain Based on the Topological Properties

Abstract

In this paper, we provide a new approach to modeling and analyzing the Goldfinger attacks in blockchain networks, based on the topology of a peer-to-peer network; this paper studied the impact of the Goldfinger well known as 51% attacks in the case of the ring, mesh and fully connected topology. The outcome of simulation has proven that in the case of fully connected topology, the attacker node with 501 (hash/s) has found 155 blocks in 60 s and the rest of the nodes, on average, 58 per node; this gives the control of the balance to the blocks of node J with 131 balance and 43 balance on average for nine nodes, in practice, the node with more than 50% will monopolize the balance. The Goldfinger occurred very worst in the fully connected topology and monopolized the network because of topologically connections; the attacker in fully connected topology is a direct and duplicated connection with the nine nodes. In terms of connection duplicated bidirectional or directional, the fully connected topology presents a hard fork compared to ring and mesh topology after the attack.