blockchains need geospatial decentralization validators should be globally distributed to ensure regulatory resilience, robustness, and fairness our research dives in we do empirical analysis, propose GPoS mechanism and run emulations using GeoDec

empirical analysis 1. data collection: gather validator data for location and stake. Solana, Aptos and Avalanche data is pulled through their APIs. Sui Mysten team helped us gather data. For Ethereum and other Cosmos SDK chains, the challenge is to differentiate between a full node and a validator. We tracked beacon nodes in Ethereum to learn about their short lived subscription to attestors

2. Quantifying decentralization proving that existing metrics such as Gini, Nakamoto does not capture geospatial nature, we design metric to capture geospatial decentralization. We design a graph: validators are nodes, stake is node weight and edge is inverse of distance between nodes Find the influence of validators using Eigenvector centrality. We then take Gini of Eigenvector centrality to quantify the inequality in influence.

Seeing most blockchains are not geospatially decentralized, we questioned how to improve it? assuming proof of location, what if we consider not just stake in voting power, but also the geospatial diversity that validator contributes we call it GPoS: Geospatially-aware Proof of Stake

finally, we run emulations on GeoDec emulator to study the impact of shift in voting power and impact of GPoS. GeoDec runs emulations on HotStuff, CometBFT with validators at custom locations, artificial latencies using netem the emulations show minimal overhead with GPoS