zkStream: a Framework for Trustworthy Stream Processing

Authors:
Open Access
Publisher website
Google Scholar
Abstract
In stream processing, managing sensitive information in a timely manner while ensuring trust remains a significant challenge. When parties without a priori trust cooperate to execute a streaming application, it is difficult to ensure that sensitive data is kept confidential while guaranteeing that every party executes their code honestly. This paper presents zkStream: a framework that leverages signatures and zero-knowledge proofs (ZKP) to add trust to streaming applications that run in the edge cloud, guaranteeing data confidentiality, provenance, and computational integrity. We introduce two optimizations to minimize the computational overhead associated with ZKPs, making our framework suitable for real-world applications. We validated our solution with existing benchmarks for streaming applications. Our method achieves an end-to-end latency that is between 6.5 and 15× faster than a naive implementation, demonstrating its potential for industrial adoption where trust is critical.

This publication has 52 references indexed in Scilit: