Inspiration In modern cybersecurity, security is only as strong as its randomness. Standard software-based Pseudo-Random Number Generators rely on deterministic algorithms—meaning if an attacker discovers the initial seed, the entire output sequence becomes predictable. Real cryptographic randomness cannot come from code alone; it requires physical, non-deterministic entropy harvested directly from natural physical phenomena.
What It Does Our Zener TRNG is a hardware-based True Random Number Generator. It harvests micro-level quantum noise from a Zener diode, isolates and amplifies the signal through active operational filtering, and processes the raw entropy via an Arduino Uno to output unbiased, high-entropy binary streams for cryptographic applications.
How We Built It Quantum Noise Harvesting We reverse-bias a Zener diode into its reverse-breakdown regime using a battery DC source.
Band Bending and Quantum Tunneling: At high reverse bias, the silicon energy bands bend sharply. Electrons probabilistically tunnel through the narrowed depletion barrier from the P-type to the N-type semiconductor region—a purely quantum mechanical process that is fundamentally unpredictable.
Avalanche Ionization: High-velocity electrons collide with silicon atoms, producing secondary charge cascades influenced by dynamic lattice vibrations.
Thermal Stability: Tunneling exhibits a negative temperature coefficient, while avalanche breakdown exhibits a positive one. The chosen voltage threshold creates an ideal balance that naturally stabilizes the noise output against thermal drift.
Analog Signal Conditioning and Filtering DC Blocking Capacitor: Isolates the high DC bias voltage from the Zener diode while passing pure AC microvolt noise, protecting the amplifier from saturation.
Gain Staging Op-Amp: Amplifies weak millivolt noise fluctuations using feedback resistors into dynamic volt-scale waveforms centered at the midpoint of the Arduino ADC input range.
Digital Processing and Debiasing Pipeline High-Speed ADC Sampling: The Arduino Uno samples the conditioned analog noise waveform.
Von Neumann Debiasing: Removes physical circuit biases by evaluating consecutive bit pairs, keeping mismatched pairs and discarding identical pairs to guarantee equal probability.
Interface and Display: Outputs real-time dynamic bitstream diagnostics onto an I2C LCD Display, toggled via a pushbutton.
Challenges We Faced DC Isolation: Blocking the Zener diode DC voltage without dampening high-frequency AC shot noise required fine-tuning the coupling stage.
Op-Amp Centering: Biasing the amplifier output precisely to the midpoint was critical to prevent waveform clipping and avoid digital bias during ADC conversion.
Power Supply Interference: Isolating quantum breakdown noise from external electromagnetic noise required switching from noisy USB power to pure battery DC lines.
What We Learned The physical interplay between quantum tunneling and avalanche breakdown in Zener junctions.
Practical active signal processing, DC decoupling, and gain staging for microvolt entropy sources.
Applying algorithmic Von Neumann debiasing to transform continuous physical noise into balanced digital bits.
Log in or sign up for Devpost to join the conversation.