With the development of cryptography and the popularization of message encryption, the need for truly random numbers has increased. Generating a large pool of numbers requires well-designed algorithms. They should be unpredictable and unbreakable by a potential attack. Raw mathematical solutions do not provide a perfectly random sequence of numbers. The result of the generators will depend on the input data, so random numbers must also be passed to the generator for the output to be random. Entropy helps determine the predictability of the message entering the random generator. The purpose of this study was to investigate a rectifying antenna as a new source of entropy for a random number generator. The work included a brief overview of the field. Popular and non-obvious solutions to the problem were mentioned. The voltage appearing at the rectifying antenna was investigated. A measurement system, based on an STM32 microcontroller was designed to address the problem. The software was written to accurately measure successive voltage values. A script written in Python was used to receive data. In addition, the program allowed visualization and comparison of the collected data on graphs. Tests of received radio noise in different environments were realized. The values of the pulse coming from an electromagnetic wave with a frequency of 2.4 GHz, generated by a cell phone were examined.
The portable measurement system was based on an STM32f303re microcontroller. Data was acquired using the build in ADC and with the use of a DMA timer to match the form of a signal. Harvested data were the fuel of the SHA-256 Digest function.
In addition, the random number generator was implemented based on the work of the Pierre L'Ecuyer. Whole random number generator process using the collected values was examined. Developed bits mixer to add the extra entropy ✨
Python script was written to collect data from the serial port & visualize the frequency of number values occurrence. Different environments and radio noise sources were examined.
