Adaptive Variation-Resilient Random Number Generator for Embedded Encryption

summary

Video file (mp4)

The gist

The gist The adaptive RNG extracts unbiased encryption-grade random number streams from physically driven entropy sources, mitigating variations in process and supply voltage through an adaptive

In short

The adaptive RNG extracts unbiased encryption-grade random numbers from physical sources like magnetic tunnel junctions (sMTJs). It uses an adaptive digitizer with a moving reference voltage to track and correct variations caused by process or supply voltage changes. This method ensures high-quality randomness suitable for embedded cryptography, achieving 100% pass rate on NIST tests across various throughputs.

Key concepts

Entropy Sources
These are the physical devices that generate true randomness, like thermal noise in electronics or stochastic switching in magnetic tunnel junctions (sMTJs). The challenge is that these sources are often biased by manufacturing variations and supply voltage fluctuations, which can skew the random numbers they produce.
Adaptive Digitizer
This is the core innovation. Instead of using a fixed reference voltage, this circuit uses a low-pass filter to create a moving reference voltage that tracks short-term average changes in the input signal. This allows the system to dynamically correct for drift or voltage mismatches in real-time.
Post-Processing (Mini Trivium)
After digitization, the resulting bit stream needs final refinement to meet strict cryptographic standards. The paper uses a simplified version of the Trivium cipher called Mini Trivium. This reduces hardware complexity while ensuring the final output has high statistical randomness suitable for encryption.
Variation Resilience
This refers to how well the RNG maintains its quality despite changes in operating conditions, such as fluctuations in supply voltage (VDD) or temperature (TMR). The adaptive digitizer actively compensates for these variations, significantly improving the system's robustness compared to standard TRNGs.

Terminology used across episodes

This episode discusses

The paper

Adaptive Variation-Resilient Random Number Generator for Embedded Encryption · Read on arXiv

Department of Computer Engineering, King Faisal University, Al-Ahsa, Saudi Arabia · Electrical Engineering Department, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia · University of California Santa Barbara · College of Computing & Data Science at Nanyang Technological University Singapore · School of Computing at Queen's University Kingston Ontario Canada · Massachusetts Institute of Technology

With a growing interest in securing user data within the internet-of-things (IoT), embedded encryption has become of paramount importance, requiring light-weight high-quality Random Number Generators (RNGs). Emerging stochastic device technologies produce random numbers from stochastic physical processes at high quality, however, their generated random number streams are adversely affected by process and supply voltage variations, which can lead to bias in the generated streams. In this work, we present an adaptive variation-resilient RNG capable of extracting unbiased encryption-grade random number streams from physically driven entropy sources, for embedded cryptography applications. The system's key feature is its adaptive digitizer with an adaptive reference voltage. As a proof of concept, we employ a stochastic magnetic tunnel junction (sMTJ) device as an entropy source. The impact of variations in the sMTJ is mitigated by the adaptive digitizer, which generates an adaptive short-term average reference voltage that dynamically tracks any stochastic signal drift or deviation, leading to unbiased random bit stream generation. The generated bit streams, due to their higher entropy, then only need to undergo simplified post-processing. A prototype of the adaptive RNG system was experimentally implemented using discrete electronic components and an FPGA for entropy source emulation. Statistical randomness tests based on the National Institute of Standards and Technology (NIST) test suite are conducted on bit streams obtained using the simulations as well as the discrete electronic component implementation, demonstrating that the bit streams consistently pass all 16 tests of the NIST SP 800-22 test suite with a 100% pass rate. Leveraging its simplified post-processing, the adaptive RNG shows consistent operation across a wide range of throughputs from 5 to 182 Mbps.

DOI: 10.1103/6yyk-321x

Transcript

Introduction to the show: ident: Security Radio. Generated commentary on the latest security and cryptography papers.

Nadia: I'm Nadia.

Elias: And I'm Elias, here with Priya.

Nadia: We're looking at the paper: "Adaptive Variation-Resilient Random Number Generator for Embedded Encryption".

Elias: The gist The adaptive RNG extracts unbiased encryption-grade random number streams from physically driven entropy sources, mitigating variations in process and supply voltage through an adaptive digitizer.

Nadia: First, who's behind it and why it matters.

Paper summary: Nadia: To recap this paper, the core thesis is tackling the issue where hardware TRNGs get biased due to process variations and supply voltage fluctuations in IoT devices.

Elias: They propose an adaptive RNG that uses a stochastic magnetic tunnel junction, or sMTJ, as an entropy source but mitigates its noise by having an adaptive digitizer with a reference voltage that tracks the signal's short-term average.

Priya: So what’s the main motivation driving this whole thing? Why do we need this specific kind of correction?

Nadia: Because conventional hardware TRNGs, based on classical electronics or emerging devices, are adversely affected by variations in fabrication process technology, which leads to bias in the generated bit streams.

Elias: Plus, for battery-powered IoT devices that fluctuate wildly with supply voltage changes, those fluctuations can also negatively affect the TRNG output.

Priya: So it’s a double whammy: manufacturing inconsistencies and power instability both mess up the randomness you need for security.

Nadia: Exactly. They show how this adaptive digitizer generates an adaptive moving reference voltage from the stochastic signal corresponding to its short-term average, which corrects for that drift or deviation.

Elias: This adaptation happens before the comparator, without needing any information about what the comparator ultimately outputs, using a low-pass filter circuit where the time constant tauLPF of that RC circuit dictates how fast it adapts to changes in Vstochastic.

Priya: So if I’m driving and my car's voltage dips slightly while I’m on the highway, this system is trying to keep the reference stable using what it sees right then?

Nadia: Precisely. It tries to maintain a short-term average of the noise itself, correcting for that drift or mismatch caused by those real-world conditions.

Elias: And they show that because the generated bit streams have higher entropy, they only need light-weight post-processing to ensure encryption-grade randomness is achieved.

Priya: That's good for hardware implementation because it means less complex circuitry is needed downstream to get the final security quality up where it counts.

Nadia: It simplifies the overall system architecture by reducing the complexity of what comes after the adaptive digitizer, which saves on hardware resources and power consumption in embedded systems.

Elias: So, in short, they’re presenting this adaptive RNG to extract high-quality unbiased random bit streams from physically driven entropy sources for embedded cryptography applications.

Priya: That feels like a good summary of the paper’s central claim, focusing on the resilience against those physical variations.

Conclusion: Nadia: So looking at this whole paper, "Adaptive Variation-Resilient Random Number Generator for Embedded Encryption," it really shows how crucial robust randomness is for IoT security.

Elias: It’s about taking a technology that has inherent noise—like sMTJs—and making it usable in real-world applications by adding intelligent adaptation to handle the physical inconsistencies we expect.

Priya: The authors are showing that you can build something that isn't just good under perfect conditions, but one that performs reliably even when the manufacturing process or power supply is shaky.

Nadia: That’s what they prove: this adaptive RNG doesn't need voltage pulsing to work, which is a big practical win for deployment in many embedded systems.

Elias: It really emphasizes the generality of the adaptive RNG concept because it can be implemented using various entropy sources, not just MTJs, as shown by their FPGA experiments.

Priya: So what’s the final word on this work? What does it mean for security researchers and designers in practice?

Nadia: It means that for embedded encryption in IoT devices, we have a method to get high-quality random numbers reliably without having to rely on fragile external stabilization methods.

Elias: They show that you can achieve consistent, high-quality randomness across a wide throughput range from five to one hundred eighty-two megabits per second.

Priya: It’s solid evidence that this design meets the statistical requirements for encryption, which is what we need before we trust it in a real security context.

Nadia: So, the main thing is that this approach provides a robust way to get unbiased random bits from physical sources for embedded encryption applications.

More episodes

← Home