IMPROVING SCALABILITY AND OPTIMIZING MESSAGE DISSEMINATION ON THE XRP LEDGER

Blockchains have become crucial in numerous fields, notably within financial applications.
The advent of Bitcoin coincided with the emergence of a global economy highly dependent
on digital transactions, stimulating the development of numerous blockchains that aim to ad-
dress different challenges. The XRP Ledger (XRPL) is an example of a blockchain designed
to address cross-currency payments. Unlike its counterparts, the XRPL does not employ
resource-intensive consensus mechanisms such as Proof of Work (PoW), employing in-
stead a Federated Byzantine Agreement consensus protocol, in which subnets of validators
vote to determine the next version of the ledger.
The XRPL network allows any node to act as a validator, but its consensus mechanism
relies on a trusted group of validators called the Unique Nodes List(UNL). In its current im-
plementation, the XRPL relies on a single UNL, curated by the XRP Ledger Foundation
(XRPLF). The dependence on this centrally managed UNL raises issues regarding authority
distribution and network scalability. Although the XRPL is decentralized in terms of com-
munication, it does not feature dynamic trust overlay rearrangement, resulting in limited
authority dispersion and possible performance problems as the network grows.
This thesis investigates ways to enhance the dispersion of authority and node-wise scal-
ability of the XRPL. First, we explore a tool that has the potential of diversifying UNLs to
achieve higher trust autonomy, we then propose the use of pubsub dissemination to reduce
message overhead and improve scalability. The research also addresses key questions on
the impact of pubsub mechanisms on the performance of the XRPL, aiming to provide insight
into tuning parameters for optimal network efficiency. The ultimate goal is to determine how
pubsub dissemination can enhance the scalability and performance of blockchains using
FBA consensus, specifically within the context of the XRPL.