Random Number Generation Based on Cryptocurrency Prices and Linear Feedback Shift Register

This paper presents the design and evaluation of an unpredictable random number generator (URNG) utilizing cryptocurrency prices. A URNG operates using a deterministic algorithm, while utilizing external entropy sources to generate random numbers that are practically unpredictable. The proposed URNG employs a linear feedback shift register (LFSR), whose sampling period are fluctuated by the cryptocurrency price Pt or its logarithmic return Rt. Using Bitcoin (BTC) price data, we simulated the proposed URNG and evaluated its randomness with the Diehard test. Our findings suggest that a 40-bit or longer LFSR, with sampling period fluctuations determined by either the least significant bit of Pt or the comparison of Rt to its average value, achieves satisfactory randomness quality. Future work includes evaluating the URNG with other cryptocurrencies and exploring new methods for entropy extraction.