Public Randomness Protocol Using VDF Staging for Low Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating randomness in public participation events lack security, scalability, and usability, as well as suffer from high result-publication latency, making them unsuitable for large-scale applications.
Innovation Solution
A novel protocol called HeadStart, which divides the contribution phase into stages and uses verifiable delay functions (VDFs) to reduce result-publication latency while ensuring unpredictability and bias resistance, allowing participants to contribute and verify results efficiently on commodity devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If protocols based on commitments, PVSS, and threshold signatures are used to ensure security, then unpredictability is improved, but result-publication latency increases due to multiple phases with high communication overhead
Solution Approach 1:
The protocol segments the contribution phase into multiple stages, where each stage collects contributions and publishes Merkle tree roots independently. This segmentation allows the result-generation phase to start earlier by processing completed stages in parallel, reducing the overall result-publication latency while maintaining security through staged verification.
Solution Approach 2:
The protocol performs preliminary computation of Merkle tree roots and audit paths for each contribution stage during the contribution phase itself. By preparing these verification components in advance, the system eliminates the need for lengthy post-contribution verification phases, thereby reducing result-publication latency without compromising unpredictability.
2Reliability
If protocols based on delay functions and VDFs are used to prevent precomputation, then security is improved, but result-publication latency increases due to long result-generation time
Solution Approach 1:
The protocol divides the contribution collection into multiple parallel stages, each with its own Merkle tree construction. By segmenting the work this way, the system can process and verify contributions in smaller batches rather than waiting for all contributions to accumulate, significantly reducing the effective result-publication latency while maintaining the security guarantees of VDFs.
Solution Approach 2:
Each contribution stage independently generates its own Merkle tree root and audit path, making the verification process self-contained for each stage. This self-service approach allows multiple stages to be processed in parallel without interfering with each other, reducing the total time to publish results while maintaining security.
3Loss of time
If randomness beacon protocols decrease beacon interval to reduce latency, then result-publication latency is reduced, but contribution phase duration is shortened which affects fairness
Solution Approach 1:
The protocol segments the contribution phase into multiple stages that can run in parallel or sequentially, allowing the total contribution collection period to be extended across multiple beacon intervals. Each stage maintains its own Merkle tree and verification path, so the system can accommodate longer contribution periods without increasing the latency of individual result publications, thereby maintaining both fairness and low latency.
Data Source
AI summary
A method of generating randomness by public participation may comprise: communicating with the commodity devices to execute a protocol comprising a setup phase, a contribution phase and a result-generation phase, wherein: in the setup phase, parameters are initialized, a verifiable delay function is setup, and the parameters are published; the contribution phase is divided into at least one first stage, published parameters are provided, random values are received, and a Merkle tree root and Merkle tree audit paths are published in each of the first stage; and the result-generation phase is divided into at least one second stage of the same number as that of the first stage, each second stage is dedicated to one of the first stage ahead of the second stage for a period, and in each second stage, computation is performed to generate a result of randomness which is published.