Novel Postprocessing and Sampling Point Optimization Techniques for Enhancing Quality of Randomness in MTJ-Based TRNGs

True random number generators (TRNGs) play a crucial role in cryptography, hardware security, statistical analysis, simulations, and device modeling. However, ensuring sufficient randomness when extracting binary bits from hardware devices is of utmost importance. The state-of-the-art techniques in the field lack comprehensive exploration for enhancing the quality of randomness in hardware-based TRNGs. This article proposes two techniques to enhance the quality of randomness in TRNGs by leveraging the advantages of voltage-controlled magnetic anisotropy (VCMA) in magnetic tunnel junctions (MTJs). The first technique involves postprocessing using von Neumann extraction (VNE) followed by rotation symmetric Boolean functions (RSBFs). The VCMA-MTJ serves as a source of randomness, offering benefits like low energy consumption, high throughput, and a compact design. A 5-variable RSBF is proposed that is invariant under circular translation with very good cryptographic properties. The combination of the proposed RSBF and VNE applied to raw data extracted from the VCMA-MTJ-based TRNG significantly enhances the quality of randomness in the binary stream. Additionally, the sampling point optimization technique effectively showcases true randomness by harnessing device stochasticity and process variations without requiring additional circuitry. The results from the National Institute of Standards and Technology (NIST) test suite demonstrate that the bit-streams generated using these techniques exhibit true randomness. Moreover, the hardware performance results show that the sampling point optimization technique is more efficient in terms of throughput and energy consumption as compared to the postprocessing scheme, primarily due to the absence of additional circuitry.