Acousto-Optic System & Method for generating PUFs

Method to generate physical unclonable functions (PUFs) which uses acousto-optic systems.
System that allows to reduce complexity while maintaining high integrability and reliability. Allows miniaturization.

Market need: A PUF consists of a physical system interacting with a physical entity, generating for each input (challenge) a unique, stable in time and difficult to predict output (response), generating challenge-response pairs. This output can provide a digital identity to physical circuits and build secure authentication mechanisms. Commonly memory-based and delay-based PUFs are considered weak against quantum computing and modelling attacks.
Strong PUFs are systems in which multiple different inputs can be generated, each resulting in a unique output. This increases the complexity of the mechanism against modelling attacks. However, their high sensitivity to environmental effects and misalignments in the readout system makes them prone to error.

Proposed solution: The method comprises an acousto-optic system for generating strong PUFs. In this system the acousto-optic medium modulates an optical beam to generate multiple patterns used to probe a PUF based in optical scattering. The responses are unfeasible to predict or simulate, making this PUF safe against tampering attacks.
Multiple challenge-response pairs can be generated by varying the properties of the acoustic signal or by using multiple acousto-optic mediums could arranged sequentially.
In addition, this acousto-optic system can be implemented as a photonic integrated circuit (PIC), enabling its miniaturisation and integration in Internet-of-Things (IoT) devices.

Competitive advantages

High number of Challenge-Response pairs. An electrical signal controls the acoustic wave generator and can modify the modulation of the optical beam by the system. Moreover, different acousto-optic systems can be concatenated, further increasing the numbe
Increased Reliability. No precision alignment, mechanical stages or lenses are required, reducing complexity and probability of error.
Integrability and Miniaturization. The method can be implemented on a chip.