Deception with Side Information in Biometric Authentication Systems

TL;DR

This paper analyzes the likelihood of successful deception in biometric authentication systems when adversaries have correlated side information, establishing the optimal exponent of deception probability through information-theoretic methods.

Contribution

It introduces a novel analysis linking biometric deception with rate distortion theory, providing the first characterization of deception probability exponents with side information.

Findings

Derived the optimal deception probability exponent.

Established the connection between biometric deception and rate distortion with side information.

Provided both achievability and converse proofs.

Abstract

In this paper, we study the probability of successful deception of an uncompressed biometric authentication system with side information at the adversary. It represents the scenario where the adversary may have correlated side information, e.g.,~a partial finger print or a DNA sequence of a relative of the legitimate user. We find the optimal exponent of the deception probability by proving both the achievability and the converse. Our proofs are based on the connection between the problem of deception with side information and the rate distortion problem with side information at both the encoder and decoder.