Decentralized E-Voting Anonymity and Scalability via Parallel Differential Privacy on Hyperledger Besu

Abstract

Blockchain - based e-voting systems use decentralized ledgers to securely record encrypted votes, ensuring immutability, transparency, and tamper-proofverification without central authorities. These systems such as the existing BP-Vot framework integrate smart contracts, k,ε differential privacy, and self-sovereign identities (SSI) to balance transparency, security, and voter anonymity in remote elections. BP-Vot deploys on Hyperledger Besu, using a single pivot candidate to redistribute votes probabilistically (θ=1/d), achieving 98%+ vote approximation accuracy via Min-Max regression and 1s/TX latency—24% better than prior art—while proving robustness against reconstruction attacks. However, limitations include single-pivot privacy fragility in low- volume elections, partial reliance on election authority for key registration, and untested scalability beyond 50k votes or additional nodes.This work proposes a multi pivot parallel differential privacy extension, dynamically selecting multiple pivots for distributed noise injection, fully integrated with SSI Web3 wallets and immutable contracts. Benefits encompass superior anonymity for millions of votes, eliminated centralization risks, and optimized performance for national-scale deployments. Evaluation confirm linearprivacy/accuracy gains with vote volume, independent of candidate count, enabling GDPR-compliant, trustless e-voting superior to state-of-the-art. Keywords: Blockchain, e-voting systems, decentralized ledgers, BP-Vot framework, Hyperledger Besu, Min-Maxregression, SSI Web3 wallets