Attendees

1. @Sara Faour

2. @David Burnett

3. @Thomas Watteyne

4. @Titan Yuan

5. @Fil Maksimovic

6. @Kristofer PISTER

Minutes

• @Sara Faour

  • Discussed with @Fil Maksimovic about options for PUFs. Maybe oscillators are a good option.

  • Published article using SRAM to generate PUF data. SRAM likely better because it is less effected by aging, temperature, etc. SRAM is considered a weak PUF because it is only available when powered.

  • @Thomas Watteyne question: anything stable within the oscillator, etc., to repeat the same conditions? Also possible to use the differences between oscillators. 128 bits is the goal here.

  • Radio and ADC are OK sources of randomness

• @Titan Yuan

  • ADC plotting and more working! Working on improving reliability. SCuM seems to crash occasionally, maybe because of printf overload described last week. Goal is to continuously tune chip as it drifts over time so it can run for many hours without recalibration. Will base work on @Fil Maksimovic's SCuM star code to continuously listen. Clock sampling the IF waveform is the first to drift out of spec.

  • SCuM star includes monitoring IF to continuously tune LO. Openmote sends constant length packets and retunes IF sampling clock based on time between SFD and CRC. 2 MHz clock

  • Ordered 10 UART transceivers (product with clear plastic box around blue PCB)