Graph Layer Security: Encrypting Information via Common Networked Physics

Abstract

The proliferation of low cost Internet of Things (IoT) devices demands new encryption mechanisms over their wireless communication channel. Traditional public key cryptography (PKC) demands high computational power and is not suitable for low power IoT devices, making them vulnerable to more powerful eavesdroppers. Recent advances in physical layer security (PLS) exploits common wireless channel statistics to generate symmetrical keys, but require accurate channel estimation and a high communication signal-to-noise ratio (SNR). As most embedded and underground IoT devices operate in low communication SNR regimes, they cannot reliably use either PKC nor PLS encryption. Many IoT devices monitor underground networked assets such as water, oil, gas, and electrical networks. Here, we propose to exploit the monitored physical dynamics data to act as a basis for encrypting the digital information. Graph Layer Security (GLS) is proposed for the first time here, as a way to encode networked physical assets' information via their graph signal processing properties. Our approach is premised on the exploitation of networked correlation in nonlinear physical dynamics for encryption and decryption. We achieve this using Koopman operator linearisation and Graph Fourier Transform (GFT) sparsification. The resulting GLS encryption scheme, like PLS, do not require the exchange of keys or a public key, and is not reliant on wireless channel properties. Using real world examples, we demonstrate remarkably secure wireless communication encryption. We believe the technology has widespread applicability in secure health monitoring for Digital Twins in challenging radio environments and conclude our seminal paper with a discussion on future development challenges.