A Bitcoin payment network with reduced transaction fees and confirmation times

The high transaction fees and confirmation times made Bitcoin unfeasible for many applications when the payments are in small amounts and require instant approval. As a result, many other cryptocurrencies were introduced for addressing these issues, but the Bitcoin network is still the most widely used payment system. Without doubt, to benefit from its network of users, there is a need for novel solutions that can successfully address the problems about high transaction fees and transaction verification times. Recently, payment network ideas have been introduced including the Lightning Network (LN) which exploits off-chain bidirectional payment channels between parties. As off-chain links can be configured to perform aggregated transactions at certain intervals without writing to blockchain, this would not only reduce the transaction fees but also decrease the verification times significantly. Nevertheless, LN still deploys relay nodes which charge fees and its growth leads certain nodes to become monopolies which defeat the very purpose of decentralization. Despite the thought that LN will provide a "scale-free network mechanism" along with high decentralization, how to form such a network among multiple parties is never studied before. Therefore, in this paper, by exploiting the LN, we offer to form a purely decentralized payment network which will allow Bitcoin to handle a lot of transactions with enough capacity. The idea is to form a network of retailers (i.e., nodes) which do business with Bitcoin by connecting them through off-chain links (i.e., payment channels) based on the business needs. The problem is initially modeled as a network optimization but since scalability would be a concern the next steps follow a heuristic approach where certain links are pruned to force evenly distributed payment flows while minimizing the total investments made to create initial off-chain links. The evaluations demonstrate the extent of scalability of the network along with the trade-offs between the distribution of flows and the initial flow capacities of the nodes. (C) 2020 Published by Elsevier B.V.