ZkregPlus: Privacy-Preserving Content Compliance for Filecoin Storage

# Open Grant Proposal: `ZkregPlus: Privacy-Preserving Content Compliance for Filecoin Storage`

**Project Name:** `ZkregPlus: Privacy-Preserving Content Compliance for Filecoin Storage`

**Proposal Category:** `Research & protocols`

**Individual or Entity Name:** Hofstra University (TBA - If Hofstra does not
approval the funds payment structure, I will apply as an individual).

**Proposer:** `xfu2006`

**Project Repo(s):** `https://github.com/xfu2006/ZkRegPlus.git`

The primary repository is currently private but will be open-sourced. Related prior work and foundational libraries by the PI include:
- https://github.com/xfu2006/IZPR
- https://github.com/xfu2006/zkreg
- https://github.com/xfu2006/halo_pse (Enhanced with CQ lookup arguments)

**(Optional) Filecoin ecosystem affiliations:** `None`

**(Optional) Technical Sponsor:** `None`

**Do you agree to open source all work you do on behalf of this RFP under the MIT/Apache-2 dual-license?:** `Yes`

# Project Summary

ZkregPlus addresses the critical challenge of privacy-preserving content
compliance within the Filecoin ecosystem. Proving non-membership of
a committed string against large regular expression (regex)
collections, such as the 38k industry-standard malware signatures in ClamAV,
is a notoriously difficult problem due to exponential state-space explosion.
Existing solutions like Zombie and Reef struggle to scale to these massive,
complex datasets, often becoming computationally infeasible for
provers or failing to support full PCRE regex features.

Our project introduces a novel tri-value logic based abstraction framework
that provides comprehensive support for the full ClamAV regex rule-set.
By incorporating various recent zk-Snark techniques (such as
CQ/Logup lookup, SuperNova, and Mangrove), we further improve
the prover performance. ZkregPlus achieves zero false positives and zero false negatives in empirical evaluations of over 700MB of Linux binary executables
against the full ClamAV signature set. Furthermore, the system is at least
two orders of magnitude faster than current state-of-the-art techniques.
Preliminary data demonstrates that even for "hard-to-prove" 80MB executables
in Linux, proofs are completed in 12 hours on standard GCP C2D instances
(448GB RAM), with significant room for further performance gains
through enhanced parallelism.

The project is currently 70% complete, with the core folding/zkSnark prover
architecture and regex discharging algorithms already implemented in about
40k LOC of Rust. This proposal seeks to bridge the remaining 30% of
development, focusing on performance optimizations and integration-ready
tooling for Filecoin Storage Providers. By enabling verifiable,
privacy-respecting data auditing, ZkregPlus provides a scalable
path for Storage Providers to ensure content compliance without
compromising the confidentiality of user data.

## Impact

The primary pain point this project addresses is the fundamental tension
between data privacy and regulatory compliance on decentralized storage
networks. Filecoin Storage Providers (SPs) face increasing pressure to
ensure they do not host illegal, malicious, or prohibited content.
However, traditional auditing methods require plaintext access,
which completely undermines the privacy guarantees that
attract users to the Filecoin network. Without a scalable way
to prove content compliance in zero-knowledge, SPs are
forced to choose between significant legal/reputational risks
and compromising user confidentiality.

By getting this right, ZkregPlus enables a "Compliance-as-a-Service"
layer for the Filecoin ecosystem. This allows SPs to mandate
verifiable proofs of non-membership against specific malware or
prohibited content databases (like ClamAV) before accepting a storage deal.
The benefit is a cleaner, more secure network that remains trustless
and private. The risk of failing to solve this is
that SPs might refuse encrypted data due to fear of hidden liabilities,
thereby limiting Filecoin's utility for enterprise-grade,
sensitive data storage.

Success of this project results in a standardized, high-performance
auditing protocol that SPs can integrate into their existing workflow.
This will significantly impact the enterprise storage market by
providing a technically sound mechanism for compliance that satisfies
both legal departments and privacy-conscious clients.
In the long term, this infrastructure can scale beyond
malware detection to support custom regex-based policy enforcement
for private organizations and government entities using
the Filecoin network.

## Outcomes

The final outcomes and deliverables for this project will focus on
transforming our research prototype into a production-ready framework for
the Filecoin ecosystem. The primary software deliverables include:

1. **Open-Sourced ZkregPlus Framework:** A high-performance Rust
implementation of the abstraction based regex discharging pipeline and
the FoldPot folding Folding/SNARK architecture. This core library will
be fully documented and licensed under MIT/Apache-2.
2. **ClamAV/SNORT IDS Integration:** Empirical validation and pre-processing
logic for the ClamAV and SNORT Intrusion Detection System ruleset. This
demonstrates the framework's versatility by extending its capabilities from
malware scanning (ClamAV) to network-level security and protocol compliance.
3. **Filecoin Audit Tooling (CLI & API):** A production-ready command-line
interface and API designed specifically for Storage Providers (SPs) and users.
This tool will allow for the seamless generation and verification of
non-membership proofs during the storage deal lifecycle.
4. **Hardware-Specific Performance Benchmarks:** A comprehensive research paper or technical report detailing the design of the system, and performance
profiles across various GCP and local hardware configurations, providing SPs
with clear guidance on resource requirements for large-scale data auditing.

We will measure the success of the project using the following metrics:
- **Generality:** Achieving a >99% discharge rate with zero false positives/negatives for the full ClamAV signature sets. (>75% SNORT signature sets)
as our technique currently do not support look-aheads.
- **Latency:** Optimizing the system to handle standard 1GB datasets
within commercially viable timeframes (targeting a significant reduction
from current benchmarks via parallelism).
- **Resource Efficiency:** Maintaining RAM usage within the limits of
standard high-performance cloud instances even for complex
"hard-to-prove" binaries.
- **Integration Readiness:** Successful verification of the CLI/API by
at least one independent developer or Storage Provider peer
within the Protocol Labs network.

## Data Onboarding

N/A. This is primarily a research and protocol development project.
However, we would like to inquire if the Filecoin Foundation recommends
that we upload our primary test datasets. They all can be obtained
from public open-source repos.

## Adoption, Reach, and Growth Strategies

**Target Audience:** Our primary audience consists of Filecoin Storage Providers (SPs) who require automated content compliance tools, and enterprise/government clients who need to store sensitive, encrypted data on decentralized networks while remaining regulatory-compliant.

**Current Engagement:** We are currently engaging with the academic and ZK-research communities. We plan to present our results at major Web3 and ZK-focused venues (such as ZK Summit, or academic conferences such as USENIX Sec, NDSS, or CCS) to demonstrate the technical viability of ZkregPlus.

## Development Roadmap

The project is currently 70% complete. The following roadmap outlines the remaining 30% of development, focused on optimization, expanding to SNORT signatures, and building Filecoin-specific integration tools.

### Milestone 1: Core Optimization and SNORT IDS Integration
* **Duration:** 6 Months (May 2026 – October 2026)
* **Personnel:** Dr. Xiang Fu (PI), Student Assistant (to be hired)
* **Functionality:**
- Implement advanced parallelism within the FoldPot architecture to reduce proving time for "hard" high-memory binaries.
- Develop pre-processing logic for the full SNORT IDS ruleset. Further
improve or optimize the ClamAV signature sets.
- Conduct empirical validation on both ClamAV and SNORT sets.
* **Funding:** $30,000 ($20,000 Staffing, $10,000 Computing)

### Milestone 2: Filecoin Tooling and API Development
* **Duration:** 3 Months (November 2026 – January 2027)
* **Personnel:** Dr. Xiang Fu (PI), Student Assistant (to be hired)
* **Functionality:**
- Develop a production-ready command-line interface (CLI) for Filecoin Storage Providers to automate non-membership proof verification.
- Build a REST API for proof generation, allowing integration into various storage deal-flow systems.
- Conduct initial integration testing with established Filecoin SP nodes.
* **Funding:** $13,000 ($10,000 Staffing, $3,000 Computing)

### Milestone 3: Validation, Benchmarking, and Open Source Release
* **Duration:** 3 Months (February 2027 – April 2027)
* **Personnel:** Dr. Xiang Fu (PI)
* **Functionality:**
- Perform exhaustive end-to-end stress testing and finalize the performance benchmarking report across multiple hardware configurations.
- Officially open-source the full ZkregPlus codebase under MIT/Apache-2 license.
- Submit a research paper about the system (or at least make it
available at crypto e-print).
- Present project results at major Web3/ZK conferences and Filecoin community working groups.
* **Funding:** $7,000 ($2,000 Staffing, $5,000 Travel Support)

## Total Budget Requested

| Milestone # | Description | Deliverables | Completion Date | Funding |
|===|===|===|===|===|
| 1 | Optimization & SNORT Integration | Optimized Core + SNORT Pre-processor | Oct 31, 2026 | $30,000 |
| 2 | Filecoin Tooling & API | Auditing CLI + REST API | Jan 31, 2027 | $13,000 |
| 3 | Validation & Open Source | Benchmark Report + Public Repo + Presentations | Apr 30, 2027 | $7,000 |
| **Total** | | | | **$50,000** |

## Maintenance and Upgrade Plans

The ZkregPlus framework will be maintained as a core project within Dr. Xiang Fu’s research lab at Hofstra University. Post-grant, we plan to:
- Monitor and integrate advancements in folding schemes (e.g., further optimizations in Sonobe or Cyclefold).
- Engage with the Filecoin community to refine the auditing CLI based on real-world Storage Provider feedback.
- Seek long-term sustainability through potential integration with Filecoin's official auditing standards or further research grants.

# Team

## Team Members

- **Dr. Xiang Fu:** Principal Investigator. Professor of Computer Science at Hofstra University. Lead architect of ZkregPlus and FoldPot.
- **Student Assistant (To be hired):** Implementation support, testing, and documentation.

## Team Member LinkedIn Profiles

- [Dr. Xiang Fu](https://www.linkedin.com/in/xiang-fu-2b724b/)

## Team Website

- Faculty Profile: [https://sites.hofstra.edu/xfu/](https://sites.hofstra.edu/xfu/)

## Relevant Experience

Dr. Xiang Fu is an expert in zero-knowledge proofs and formal methods, with a focus on practical, high-performance ZK auditing systems. His recent research (2020–2025) specifically targets scalable ZK solutions for auditing and malware detection:

1. Aman Luthra, James Cavanaugh, Hugo Renzzo Olcese, Michael Raymond, Reina
Eustache, Rina M. Hirsch, and Xiang Fu. 2025. Mitigating Ponzi schemes by
zero-knowledge auditing. *Information Security Journal: A Global
Perspective* 34, 2 (2025), 167–190.
https://doi.org/10.1080/19393555.2024.2404216
2. Diya Krishnan and Xiang Fu. 2024. Towards zero knowledge argument for
double discrete logarithm with constant cost. *Theoretical Computer
Science* 1018 (2024), 114799. https://doi.org/10.1016/j.tcs.2024.114799
3. Trevor Conley, Nilsso Diaz, Diego Espada, Alvin Kuruvilla, Stenton Mayne,
and Xiang Fu. 2024. IZPR: Instant Zero Knowledge Proof of Reserve. In
*Financial Cryptography and Data Security. FC 2024 International
Workshops: CoDecFin, WTSC, and Voting* (LNCS 14746). Springer, 225–239.
4. Michael Raymond, Gillian Evers, Jan Ponti, Diya Krishnan, and Xiang Fu.
2023. Efficient Zero Knowledge for Regular Language. In *Security and
Privacy in Communication Networks: 19th EAI International Conference,
SecureComm 2023, Proceedings, Part I* (LNICST 567). Springer, 369–394.
5. Aman Luthra, James Cavanaugh, Hugo Renzzo Olcese, Rina M. Hirsch, and
Xiang Fu. 2020. ZeroAUDIT. In *Proceedings of the 36th Annual Computer
Security Applications Conference (ACSAC '20)*. ACM, 798–812.
https://doi.org/10.1145/3427228.3427254

The ZkregPlus project is a direct evolution of this work, combining the regular language ZK research from *SecureComm 2023* with the advanced folding techniques required for the Filecoin network's scale.

## Team code repositories

- https://github.com/xfu2006/IZPR
- https://github.com/xfu2006/zkreg
- https://github.com/xfu2006/halo_pse (halo_pse extension with CQ lookup)

# Additional Information
`

**How did you learn about the Open Grants Program?**
`
Through independent research via Google Search regarding funding opportunities for decentralized storage and cryptography.
`

**Contact Email:**
`Xiang.Fu@hofstra.edu`

**Additional Evaluation Information:**
We would like to emphasize to the review committee that the foundational research and core implementation of ZkregPlus (including the FoldPot folding/snark architecture and the 5-stage discharging algorithms) are already **70% complete**.
- A full technical paper draft detailing the theoretical proofs,
and performance evaluations is available for review upon request.
- The primary Rust codebase currently resides in a private GitHub repository. We are more than happy to grant access to the Filecoin grants team for technical evaluation prior to the official open-source release.