Self-Balancing Semi-Hierarchical PCNs for CBDCs

We introduce a family of PCNs (Payment Channel Networks) characterized by a semi-hierarchical topology and a custom set of channel rebalancing strategies. This family exhibits two interesting benefits, if used as a platform for large-scale, instant, retail payment systems, such as CBDCs: Technically, the solution offers state-of-the-art guarantees of fault-tolerance and integrity, while providing a latency and throughput comparable to centralized systems; from a business perspective, the solution perfectly suits the 3-tier architecture of the current banking ecosystem (central banks / commercial banks / retail users), assigning a pivotal and peculiar role to the members of each tier. Furthermore, the cryptographic privacy of payments for retail users -- typical of PCNs such as the public Lightning Network -- is largely (possibly fully) retained. We study the system by simulating a scaled-down version of a hypothetical European CBDC, exploring the trade-offs among liquidity locked by market operators, payment success rate, throughput, latency, and load on the underpinning blockchain.