A Privacy-Aware Blockchain Architecture for Secure Digital Voting

Abstract

Cryptographically motivated decentralized voting has already been on the rise around the world. However, over the last decade or so there are a wide range of concerns, from centralized controls and opaque aggregation methods to the delicate balance between vote anonymity and validated outcomes. Therefore, this paper proposes a fully decentralized e-voting architecture that leverages blockchain consensus for trustworthy, privacy-preserving voting, as well as automated verification, as a potential solution to these challenges. At a high level, the proposed architecture utilizes homomorphic encryption for security-critical aggregation, smart contract and Merkle-tree audit architecture. The validation mechanism also extends to multi-factor biometric authentication (facial recognition to one-time password) using both hardware and software. The architecture is designed to be modular to allow easy deployment across multiple blockchain implementations such as Ethereum, Hyperledger, and Polygon. Some preliminary comparisons show slight performance differences between these frameworks. To test the feasibility of the new system in a real-world scenario, a proof-of-concept candidate was developed that employs a web-based voting interface, an LBPH-based facial verification pipeline, a lightweight relational store containing all needed metadata, and a custom proof-of-work blockchain for enforced one-person-one-vote constraints. The results demonstrate that basic benchmarking and advanced mechanisms such as full homomorphic voting and Multichain Performance Testing demonstrate that a decentralized and cryptographically based approach supports trustworthy e-voting.