Ticket-based multi-strand method for increased efficiency in proof-of-work based blockchains

TL;DR

This paper tackles the limited transaction throughput of proof-of-work blockchains by introducing a ticket-based multi-strand architecture. It replaces a single linear chain with $n=2^p$ semi-independent chains, where each new block requires a ticket that encodes which chain it may belong to, thus enabling parallel block production. A block is tied to its ticket through a signature from the ticket owner, linking the PoW effort to the resulting block without a second PoW phase. The approach aims to preserve security under decentralization, reduce the impact of targeted attacks, and improve scalability, with practical considerations for implementation and user wallets across multiple chains.

Abstract

This paper outlines a method aiming to increase the efficiency of proof-of-work based blockchains using a ticket-based approach. To avoid the limitation of serially adding one block at a time to a blockchain, multiple semi-independent chains are used such that several valid blocks can be added in parallel, when they are added to separate chains. Blocks are added to different chains, the chain index being determined by a ``ticket'' that the miner must produce before creating a new block. This allows increasing the transaction rate by several orders of magnitude while the system is still fully decentralized and permissionless, and maintaining security in the sense that a successful attack would require the attacker to control a significant portion of the whole network.