Introduction: The Quest for Perfect Secrecy
In the landscape of future wireless networks (6G and beyond), security remains a paramount concern. Traditionally, wireless security has relied on upper-layer cryptographic protocols like RSA or AES. These methods provide "computational security," meaning they are safe only as long as the computational resources required to break them remain out of reach. However, the advent of quantum computing and the rapid cracking of post-quantum algorithms suggest that this foundation is increasingly fragile.
This research proposes a shift toward the gold standard of security: Shannon’s One-Time Pad (OTP). OTP offers "perfect secrecy," where an attacker cannot gain any information about the message regardless of their computing power. The primary challenge has always been the requirement for a secret key as long as the message itself. This paper introduces a framework to bridge the gap between theoretical perfect secrecy and practical wireless implementation by leveraging physical layer noise.
Proposed Architectures: Explicit and Implicit Encryption
The authors propose two distinct paths to approach OTP security within the physical layer (PHY):
1. Explicit Encryption: Noise-Based Key Generation (NBKG)
Traditional physical layer key generation relies on channel reciprocity—Alice and Bob measuring the same fading channel. However, fading is often too slow for high-speed data needs. The proposed NBKG instead utilizes local hardware and thermal noise as a source of randomness. To share these keys securely, the system employs constellation rotation and artificial noise injection. By using full-duplex capabilities, Alice can inject noise that she can cancel out, while an eavesdropper (Eve) is left with a high error floor, enabling high-rate key generation.
2. Implicit Encryption: Randomness Shaping (RS)
This "keyless" approach uses the inherent noise of the physical channel to prevent unauthorized decoding. A Randomness Shaper uses cryptographic primitives—such as One-way, Bidirectional, and Compressive Randomness Extractors—to diffuse and accumulate errors across a transmission block. This ensures that even if an eavesdropper has a momentarily clear channel, the accumulated randomness prevents them from decoding the message.
Quantifying Security: The DoA and DoSA Metrics
To measure how closely a system approaches OTP, the authors introduce two new metrics:
- Degree-of-Approaching (DoA): Quantifies the ratio of key entropy to message entropy, indicating how close the system is to the "perfect secrecy" condition.
- Degree-of-Synchronous-Approaching (DoSA): Measures efficiency by comparing the key generation rate to the channel's data capacity.
Experimental Validation and Results
The framework was evaluated through numerical simulations and a hardware testbed using Software Defined Radios (SDRs). In high Signal-to-Noise Ratio (SNR) regimes, the NBKG framework achieved a DoSA approaching 1, meaning secret keys were generated almost as fast as data was transmitted.
As shown in the constellation diagrams above, the proposed methods ensure that while the legitimate receiver (Bob) can recover the signal, the eavesdropper (Eve) sees only a uniform distribution of noise. In simulations, the Randomness Shaper successfully drove Eve’s Bit Error Rate (BER) to 0.5, the equivalent of random guessing.
Limitations and Future Directions
While promising, the approach faces practical hurdles. NBKG requires sophisticated full-duplex hardware and self-interference cancellation. Additionally, Randomness Shaping requires an estimate of the eavesdropper's error floor, which is difficult if the attacker is passive. Future work will focus on optimizing the trade-off between security overhead and total system throughput to minimize the "rate penalty" incurred by these advanced security measures.
Summary updated