Abstract
Modern incentive programs, such as airline miles and digital rewards, often compromise user privacy through centralized tracking. While privacy-preserving alternatives exist, they typically suffer from high computational overhead and limited verifiability. This paper introduces Boomerang, a novel decentralized protocol that leverages 2-cycle elliptic curves, blind signatures, and zero-knowledge proofs to provide an efficient, unlinkable, and publicly verifiable incentive system. Our implementation demonstrates significant performance gains, with issuance times of 7.25ms and the capacity to handle over 15 million requests per day on the Solana blockchain.
Introduction
Incentive systems are ubiquitous in the digital economy, yet most current implementations are centralized and privacy-invasive. Providers typically track every transaction, creating detailed behavioral maps of their users. While researchers have proposed privacy-preserving alternatives like Black-Box Accumulators (BBA+), these systems often rely on computationally expensive pairing-based cryptography, lack robust public verifiability, and possess security gaps regarding user "pooling" of points.
Boomerang addresses these challenges by creating a system that is fast enough for millions of users, requires no trusted setup, and allows for public verification on a blockchain. By moving away from pairing-based methods toward discrete-logarithm (DLOG) based tools, Boomerang achieves state-of-the-art performance while maintaining strong security guarantees.
System Architecture
Boomerang is built upon several modern cryptographic primitives designed for efficiency and scalability:
- Accumulation Tokens System (ATS): The protocol utilizes multi-use tokens bound to a user's secret key. Points are managed using the homomorphic properties of Pedersen Commitments, allowing for mathematical operations on encrypted values.
- Curve Trees: To manage tokens at scale without a trusted setup, Boomerang employs "Curve Trees" built on 2-cycle elliptic curves (specifically the secp256k1/secq256k1 cycle). This structure enables extremely fast zero-knowledge membership proofs.
- Blind Signatures (ACL Scheme): To ensure unlinkability, the issuer signs a user's token "blindly." This prevents the issuer from linking a signature to a specific transaction when the user later spends their points.
- Zero-Knowledge Proofs (ZKP): The protocol uses Sigma-protocols and Bulletproofs for range proofs, allowing users to prove they possess sufficient points without revealing their total balance.
The Issuance Protocol
The core of the Boomerang system lies in its interactive protocols between the Client and the Issuer. During the issuance phase, the client generates a commitment to their token and a zero-knowledge proof of its validity. The issuer then verifies this proof and provides a blind signature, ensuring the user can later spend the points without being tracked.
As shown in the issuance workflow, the process involves a series of cryptographic exchanges that establish a new state for the user's token ($tk_0$) while maintaining the privacy of the underlying identifiers ($ID_0$). The use of blind signature schemes (BSA) is critical here, as it allows the issuer to authorize the token without learning its specific value or identity.
Performance Evaluation
Boomerang was evaluated using a Rust implementation (~25,000 lines of code) tested on commodity hardware and the Solana blockchain. The results demonstrate a significant leap in efficiency over previous state-of-the-art protocols:
- Latency: The Issuance procedure takes only 7.25ms, compared to 72ms for BBW and 172ms for BBA+. Spending points takes roughly 98ms, more than 3x faster than BBW (304ms).
- Throughput: A single Solana validator can verify over 15.5 million reward requests per day.
- Cost-Effectiveness: The estimated cost per transaction on AWS is approximately $0.000051 for issuance, making it viable for large-scale commercial use.
Security and Fraud Prevention
Despite its focus on privacy, Boomerang includes robust anti-fraud mechanisms. The system can detect double-spending by revealing a mathematical "tag" during transactions. If a user attempts to spend the same token twice, the system can derive their long-term public key to identify the fraudster, protecting the integrity of the incentive pool without compromising the privacy of honest users.
Conclusion and Future Work
Boomerang represents a significant advancement in decentralized incentive systems, offering a high-performance alternative to privacy-invasive centralized models. While the current protocol reveals the specific value being spent to the verifier, future work aims to harden privacy further by hiding the spent amount and making the smart contract layer elliptic-curve agnostic to support multiple blockchain ecosystems.
Summary updated