Investigating Security Properties of Polkadot Using Graph Analysis

Abstract

Interest in cryptocurrencies has been steadily rising ever since Bitcoin and blockchain were introduced in 2009. Consequently, considerable efforts have been directed toward investigating the large amount of transactions data in prominent cryptocurrencies. Using network science and graph theory, researchers were able to characterize major activities on the cryptocurrency networks, including discovering security issues in the form of unusual transaction patterns. In this thesis, we conduct the first systematic study on the Polkadot cryptocurrency, which adopts a unique design based on the principles of sharding, that addresses scalability and interoperability shortcomings in existing blockchain systems. By leveraging graph analysis, we characterize the transactions and users network of Polkadot, including finding abnormal patterns and outliers. We also investigate the temporal properties and evolution of the network since its launch in May 2020. The results of our analysis provide us with novel insights into the dynamics of Polkadot as a developing multi-chain system. We also review Polkadot's architecture in detail and discuss its limitations. We hope that our work would inspire researchers to further study Polkadot, addressing some of its limitations in terms of its performance, security, and network aspects.

Date of Award	2022
Original language	American English
Awarding Institution	HBKU College of Science and Engineering