A Micro-python library which provides an interface to generate a random number using the ESP32-C3's hardware RNG.
A Micro-python library which provides an interface to generate a random number using the ESP32-C3's hardware RNG.
The code uses several hardware devices of the MCU to generate random noise values using additional entropy sources. The SAR ADC and Wi-Fi are exploited to provide quality inputs to the hardware random number generator integrated into the ESP32-C3. It reads a random value from the 'RNG_DATA_REG' register.
This uses the same method to generate random numbers as the ESP32. Only the registry address changes (0x600260B0)
- Library : esp32_c6_rng.py
- Library test: test_esp32_c6_rng.py
Source: ESP32-C3 Technical Reference Manual V.1.1 - Chapter 25 Random Number Generator (RNG) - Pages 528/529 - link
The ESP32-C6 RNG generates random integer numbers as 32-bit values.
Random value #1: 146820682
Random value #2: -196285670
Random value #3: -866286180
Random value #4: -1000580641
Random value #5: 1642066338
Random value #6: -1467530287
Random value #7: -697217448
Random value #8: -1944603219
Random value #9: 1952318979
Random value #10: 1012424606
The code was implemented specifically for an Espressif ESP32-C3 microcontroller on a Seeed Studio XIAO ESP32C3 Module Link
Esptool indicates ESP32-C3 (QFN32) (revision v0.4)
Quad Flat No-lead 32 pins (QFN32).
MicroPython v1.22.2 on 2024-02-22; ESP32C3 module with ESP32C3
Time taken to generate xxxxxx values: xx seconds (avg)
Throughput: xxxxxx bytes/sec
xxxxx random values / sec.
Time taken to generate 8000 values: 837 seconds
Throughput: 38.23 bytes/sec
Number of random values per second: 9.56
Needless to say, it is rather slow !
Clearly, the adoption of HP SRAM on the C6 has greatly improved things compared to the SRAM on the C3 in terme of latency.
(www.fourmilab.ch) John Walker
- Sample size: xx MB
- Total generated: xxxxxx values
My first attempt to generate random values gives on 44446 numbers, an entropy of 3.555333 bits per byte. although the MCU is slow, it fulfills its task perfectly.
(https://webhome.phy.duke.edu/~rgb/General/dieharder.php) Robert G. Brown
Sample size: xx MB
Total generated: xxxxxxx values