Last month we were pleased to announce the , an EVM-compatible ZKVM solution, released the 25th of July, 2022. ZKEVM has been a hot topic itself over the past couple of weeks, and upon the release of OlaVM, the paper managed to receive some honorable attention from prominent people in the industry, amongst them, one being (who also is the main author of the ), we would like to thank Daira for her feedback. Daira brought up a few important questions in regards to the design decisions, one of them relating to the choice of hash function in and signature algorithms. The exact comment can be seen in the tweet attached below. OlaVM Whitepaper Daira Hopwood Zcash protocol ECDSA Schnorr What Daira is referencing can in simple terms be described with the following: The security level of is collision resistant, so it cannot be regarded as a Random Oracle. For the ECDSA and Schnorr signature algorithms, in order to satisfy sufficient security, the choice of hash function needs to be regarded as Random Oracle. In order to understand this better, we need to understand some cryptographic concepts first. Sinsemilla hash 1. Security Properties of a Cryptographic Hash Function (CHF) According to the definition in the paper , the security properties corresponding to a CHF are divided into the following three categories: Cryptographic Hash-Function Basics - Given H(x), it is computationally difficult to determine x, i.e., it is computationally infeasible to discover any input which created a specific output. Preimage Resistance - Given x, it is computationally difficult to find some different value x’ such that H(x) == H(x’), i.e., it is computationally infeasible to find any second input which has the same output as any specified input. Second Preimage Resistance - Expanding the concept of Second Preimage Resistance, it is computationally infeasible to find any two arbitrary inputs, x and y such that H(x) == H(y), i.e., different inputs which hash to the same output. Collision Resistance An important aspect to be noted is that, if you satisfy the conditions for Collision Resistance, you also satisfy the conditions for Second Preimage Resistance, however this does not guarantee that you are satisfying the conditions for Preimage Resistance. 2. What is a Random Oracle? A is described in the following model: Random Oracle Imagine you have a black box. In the box lives a gnome, with a big book and some dice. We can add some input data into the box (an arbitrary sequence of bits). Given some input data that the gnome did not see beforehand, he uses his dice to generate some output, uniformly and randomly, in some conventional space (the space of oracle outputs). The gnome also writes down the input and the newly generated output in his book. If he is given some formerly known input data, the gnome turns to his book, in order to recover the output he generated and returned the last time, and returns it again. To briefly summarize the behavior of Random Oracles, assuming the input is x: If x has been used as input data before, it will directly return the corresponding H(x). If x has never been used as input data, the Random Oracle will generate a string consisting of 0.1 in the range of values . completely randomly It might be noted that: here means that even the Random Oracle itself does not know what the final value will be, it has no rules to follow. This is the main difference from hash functions. Any hash function has its own calculation rules. Complete randomness But in the real world, it is quite difficult to achieve an actual Random Oracle, therefore, we need to find a potential candidate for Random Oracles, and we need to make the output as random as possible. Hash functions are a good choice. A secure hash function needs to satisfy the following conditions: Preimage Resistance, Second Preimage Resistance, and Collision Resistance. A hash function that can be used as Random Oracle must satisfy these three properties, but a hash function that satisfies these three properties may not necessarily be used as Random Oracle, there is a necessary and insufficient relationship between them. More details can be found in What is the “Random Oracle Model” and why is it controversial? 3. Requirements for Using Hash Functions in ECDSA and Schnorr Signature Algorithm As mentioned in the papers and , the hash functions in both ECDSA and Schnorr signature algorithms need to be considered Random Oracles before they are safe. According to the previous description, this hash needs to satisfy all the security properties of CHF: preimage-resistance, 2nd-preimage resistance, and collision resistance. On the security of ECDSA with additive key derivation and presignatures On the Exact Security of Schnorr-Type Signatures in the Random Oracle Model 4. About the Sinsemilla Hash Function The Sinsemilla hash function was designed by , and the bottom layer relies on . Under a fixed-length input, the Sinsemilla hash function satisfies the Collision Resistance property, but not the Preimage Resistance property. Please refer to Daira’s for further clarification. Daira Hopwood and Sean Bowe ECDLP(Elliptic Curve Discrete Logarithm Problem) answer According to the , the original intention of designing the Sinsemilla hash function is to make full use of the advantages of being Lookup-friendly during the execution of the zero-knowledge proof algorithm Halo2, in order to improve the execution efficiency of Halo2. Therefore, the Sinsemilla hash function is a Lookup-friendly hash function, which is more suitable for computation of promises and computation of Merkle Tree roots. Zcash protocol specification 5. Conclusion We would like to thank for pinpointing this and her guidance, we will continue gathering feedback to continuously optimize and improve the OlaVM design, making it more efficient and secure. The Sinsemilla hash function will still be used in suitable modules in the OlaVM design. Regarding the usage of hash functions for signatures, we will opt for an appropriate and secure hash function, such as e.g the - or hash functions. Daira Hopwood Poseidon Reinforced Concrete

Zcash

Oracle

Twitter

Going Deep into the Specification for Marlin

How to Design the ZKVM Circuit

Nominated for 2022 - HackerNoon Contributor of the Year - Optimization

Nominated for 2022 - HackerNoon Contributor of the Year - Algorithms

Nominated for 2022 - HackerNoon Contributor of the Year - Privacy

Nominated for 2022 - HackerNoon Contributor of the Year - Architecture

Nominated for 2022 - HackerNoon Contributor of the Year - Blockchain

Too Long; Didn't Read

Understanding the Sinsemilla Hash Function in OlaVM

Understanding the Sinsemilla Hash Function in OlaVM

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

A Study on Parallel Execution: Everything You Need to Know

3 Features I Would Pay For in a Decentralized Social Network

A Necessary Evolution of Privacy and Data Protection on Blockchain Networks

Aergo — The 4th Generation Enterprise Blockchain Protocol.

An Intro to Solace De-Fi Insurance Protocol

Blockchain Primitives and their Advantages for Web 3.0 Development

A Study on Parallel Execution: Everything You Need to Know

3 Features I Would Pay For in a Decentralized Social Network

A Necessary Evolution of Privacy and Data Protection on Blockchain Networks

Aergo — The 4th Generation Enterprise Blockchain Protocol.

An Intro to Solace De-Fi Insurance Protocol

Blockchain Primitives and their Advantages for Web 3.0 Development

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps