Quantifying the Sensitivity and Unclonability of Optical Physical Unclonable Functions

Abstract

According to recent estimates, [1] the looming internet of things and worldwide information exchange by the year 2018-2023 will produce a global data stream of around tens zettabytes per annum. This requires secure and reliable authentication methods in order to protect private information and to safeguard access to personal devices and services. The currently widespread techniques to this end rely on the permanent storage of digital secret keys in electronic devices, for example, in smartphones, car keys, bank cards, passports, or computers. Unfortunately, the last decades have seen an explosion of attacks that can extract such keys unnoticedly, including sophisticated malware and physical methods. [2][3][4] This obviously calls for new authentication approaches with improved security features.The use of nondigital primitives such as physical unclonable functions (PUFs) constitutes a promising new avenue in this context. [5][6][7][8] PUFs are randomly structured physical systems which exhibit a complex input-output or, in PUF parlance, "challenge-response" behavior that is unique to each PUF. Their uncontrollable individual disorder on small length scales makes them practically unclonable, even for their original manufacturer. Due to their physical nature, randomness, and unclonability, PUFs can disable various popular attack vectors compared to classical, permanently stored keys: For example, their physical nature obviously prevents that PUFs are stolen remotely over a purely digital data connection by attackers. [2,9] As another example, PUFs allow the short-term derivation of individual secret key material in devices, avoiding the long-term and attack-prone presence of secrets in digital memory. This usually complicates key extraction [2] and side channel attacks. [9] Finally, some special,

Flexible Physical Unclonable Functions Based on Non-deterministically Distributed Dye-Doped Fibers and Droplets

“…Herein, it is important to highlight that to generate different challenges a Mersenne “twister” algorithm has been used. For each challenge response pair, an authority database is generated by enrolling the collected CRPs that will be used in the future for speckle comparison during the authentication process ,, (more details in Experimental Section). Once the CRPs are collected and stored, the authentication process that is used is based on the following steps.…”

“…Here, we used a 270 pixel × 360 pixel camera with 40 FPS for this task. All of the details of the experimental setup are reported in refs and . The readout requires <5 min.…”

Designing a photonic physically unclonable function having resilience to machine learning attacks

“…Because PUFs that use light-encoded data as challenge and response are less susceptible to side-channel threats that collect CRPs, previous work has studied several optical/photonic PUF architectures. 37,39,[44][45][46][47][48][49][50][51][52][53] Designs include optical scattering PUFs, which use the movement of scattered light around nano-sized obstacles to obtain responses, [44][45][46][47] and "silicon" photonic PUFs that use the interaction of light in a "chaotic silicon microcavity" to do the same. [49][50][51] Other designs control light's traversal through PIC components, 37,44 and Ref.…”

Hybrid Camouflaged Anticounterfeiting Token in a Paper Substrate

“…[ 2,19–22 ] In general, physical unclonable functions are defined as individual physical signatures whose intrinsic unpredictability produces a unique and “unclonable” response when interrogated by a specific challenge, following the so‐called challenge‐response pair (CRP) scheme. [ 23–26 ] Depending on the number of responses associated to a physical unclonable function, its strength can be classified as “weak” or “strong.” [ 27 ] Once interrogated, weak PUFs produce at least one response as in case of anti‐counterfeiting tags. [ 5,6 ] Strong PUFs instead are more complex systems that under interrogation produce a large number of independent responses from the same device.…”