perVRF

perVRF stands for per-entity verifiable random function, a cryptographic primitive designed to produce a verifiable random value that is bound to a specific entity or context. Like a standard verifiable random function (VRF), a perVRF yields a pseudo-random output and a proof that the output was computed correctly from a secret key and a given input. The distinguishing feature is the explicit binding of the randomness to a particular entity, identifier, or contextual parameter, enabling per-participant randomness without central coordination.

How it works in practice involves a VRF construction that takes an input, such as a protocol round, a node identifier, or other contextual data, and uses a per-entity or per-context key or input binding to generate a pair consisting of a random value and a cryptographic proof. Anyone with the corresponding public key can verify the proof and confirm that the output is the correct VRF value for that specific entity and input. Binding can be achieved by including the entity’s identifier in the input or by using per-entity keys.

Applications for perVRF include leader election or slot assignment in distributed and blockchain systems, per-node randomness for fair lottery or resource scheduling, and per-participant randomness in governance or consensus protocols. This binding reduces cross-entity leakage of randomness and supports per-entity fairness without relying on a global randomness beacon.

Security considerations include the standard VRF properties of unpredictability and verifiability, along with the per-entity binding guarantees. Risks involve key compromise, misbinding of identifiers, or biased inputs. Implementations typically rely on established VRF schemes (for example, elliptic-curve VRFs) with an explicit per-entity binding either through input construction or per-entity keys.