Harvesting RTN for True Random Number Generators and Physical Unclonable Functions

Random Telegraph Noise (RTN) is an intriguing entropy source that can be exploited to develop lightweight cryptographic primitives. Its utility in Physical Unclonable Functions (PUFs) has been established recently, while its potential for True Random Number Generators (TRNGs) applications has been timidly explored. However, the integration of RTN-based TRNGs while maintaining dual-use capability with PUFs remains a novel concept. This paper explores how with a suitable entropy harvesting function, previously shown to be effective for an RTN-based PUF architecture, it is also possible to extract the necessary entropy for building a resilient TRNG. The proposed RTN-based TRNG is tested using three basic randomness metrics, as well as through industry-grade NIST SP 800-22 standards tests.