Physical unclonable functions (PUFs) serve as a hardware source of private information that cannot be duplicated and have applications in hardware integrity and information security. Here we demonstrate a photonic PUF based on ultrafast nonlinear optical interactions in a chaotic silicon micro-cavity. The device is probed with a spectrally-encoded ultrashort optical pulse, which nonlinearly interacts with the micro-cavity. This interaction produces a highly complex and unpredictable, yet deterministic, ultrafast response that can serve as a unique "fingerprint" of the cavity and as a source of private information for the device's holder. Experimentally, we extract 17.1-kbit binary keys from six different photonic PUF designs and demonstrate the uniqueness and reproducibility of these keys. Furthermore, we experimentally test exact copies of the six photonic PUFs and demonstrate their unclonability due to unavoidable fabrication variations.

We demonstrate a one-dimensional optical phased array on an integrated silicon platform for operation at 1.55 µm. Light is emitted end-fire from the chip edge where the waveguides are terminated. The innovative design and high confinement afforded by the silicon waveguides enables λ/2 spacing (775-nm pitch) at the output. Steering is achieved by inducing a phase shift between the waveguides via integrated thermo-optic heaters. The device forms a beam with a FWHM angular width of 17°, and we demonstrate beam steering over a 64° range.A multitude of applications depend on the ability to quickly and accurately steer a laser beam, ranging from fiber endoscopes in biological imaging to LiDAR scanning in large-scale landscape surveying and autonomous automobile navigation. In addition, high-speed beam steering has the potential to impact a host of new technologies from free space optical communications to next-generation projectors. The conventional approaches to realizing optical beam steering are mechanical in nature and require bulky motors and gimbals to physically move the entire laser system, or alternatively tilt-mirrors through galvanometer-based or micromechanical devices. Mechanical systems are fundamentally limited in speed by their inertia. Commercial galvanometer tilt-mirror devices have maximum steering speeds of approximately 10 kHz, whereas the fastest MEMS devices can reach up to 420 kHz [1]. Both types of steering systems can be limited in range as well, with most systems subtending less than 40° total. These limitations have recently prompted the development of nonmechanical beam steering techniques based on integrated optical phased arrays (OPAs) in silicon photonic platforms, which steer by forming a beam from the interference of multiple phase-controlled optical emitters.

The hallmark of the information age is the ease with which information is stored, accessed, and shared throughout the globe. This is enabled, in large part, by the simplicity of duplicating digital information without error. Unfortunately, an ever-growing consequence is the global threat to security and privacy enabled by our digital reliance. Specifically, modern secure communications and authentication suffer from formidable threats arising from the potential for copying of secret keys stored in digital media. With relatively little transfer of information, an attacker can impersonate a legitimate user, publish malicious software that is automatically accepted as safe by millions of computers, or eavesdrop on countless digital exchanges. To address this vulnerability, a new class of cryptographic devices known as physical unclonable functions (PUFs) are being developed. PUFs are modern realizations of an ancient concept, the physical key, and offer an attractive alternative for digital key storage. A user derives a digital key from the PUF’s physical behavior, which is sensitive to physical idiosyncrasies that are beyond fabrication tolerances. Thus, unlike conventional physical keys, a PUF cannot be duplicated and only the holder can extract the digital key. However, emerging machine learning (ML) methods are remarkably adept at learning behavior via training, and if such algorithms can learn to emulate a PUF, then the security is compromised. Unfortunately, such attacks are highly successful against conventional electronic PUFs. Here, we investigate ML attacks against a nonlinear silicon photonic PUF, a novel design that leverages nonlinear optical interactions in chaotic silicon microcavities. First, we investigate these devices’ resistance to cloning during fabrication and demonstrate their use as a source of large volumes of cryptographic key material. Next, we demonstrate that silicon photonic PUFs exhibit resistance to state-of-the-art ML attacks due to their nonlinearity and finally validate this resistance in an encryption scenario.

We present a secure communication system constructed using pairs of nonlinear photonic physical unclonable functions (PUFs) that harness physical chaos in integrated silicon micro-cavities. Compared to a large, electronically stored one-time pad, our method provisions large amounts of information within the intrinsically complex nanostructure of the micro-cavities. By probing a micro-cavity with a rapid sequence of spectrally-encoded ultrafast optical pulses and measuring the lightwave responses, we experimentally demonstrate the ability to extract 2.4 Gb of key material from a single micro-cavity device. Subsequently, in a secure communication experiment with pairs of devices, we achieve bit error rates below 10 at code rates of up to 0.1. The PUFs' responses are never transmitted over the channel or stored in digital memory, thus enhancing the security of the system. Additionally, the micro-cavity PUFs are extremely small, inexpensive, robust, and fully compatible with telecommunications infrastructure, components, and electronic fabrication. This approach can serve one-time pad or public key exchange applications where high security is required.

Demonstrated is a record serial transmission rate of 42.8 Gbit/s by sending four-level pulse amplitude modulated (PAM-4) data over a low-loss Megtron 6 electrical backplane. The data is processed and equalised using early -late gate timing recovery and a finite-impulseresponse-based least-mean-squares equaliser. Bit error rate performance is reported over several different links.