GAPG: a Heuristic Greedy Algorithm for Grouping Storage Scheme in Blockchain

The decentralization and immutability of blockchain make it have a wide application prospect, however, as the number of transactions increases, the storage cost of each node will also increase. So blockchain will be facing the bottleneck of insufficient scalability. Consensus Unit(CU) is one of grouping storage scheme for improving the scalability of blockchain, which organizes different nodes into one cluster and allocates all the blocks of the whole chain to the nodes in each cluster. In this paper, we propose an optimization problem that assigning blockchain data to nodes and minimized the total query cost, this is a variant of generalized basic probability assignment problem, so we propose a heuristic algorithm named GAPG to solve this problem. Numerical results demonstrate that the proposed algorithm is superior to random assignment and greedy algorithms for clusters with a large number of nodes.