Abstract High-bandwidth applications, from multi-gigabit communication and high-performance computing to radar signal processing, demand ever-increasing processing speeds. However, they face limitations in signal sampling and computation due to hardware and power constraints. In the microwave regime, where operating frequencies exceed the fastest clock rates, direct sampling becomes difficult, prompting interest in neuromorphic analog computing systems. We present the first demonstration of direct broadband frequency domain computing using an integrated circuit that replaces traditional analog and digital interfaces. This features a Microwave Neural Network (MNN) that operates on signals spanning tens of gigahertz, yet reprogrammed with slow, 150 MBit/sec control bitstreams. By leveraging significant nonlinearity in coupled microwave oscillators, features learned from a wide bandwidth are encoded in a comb-like spectrum spanning only a few gigahertz, enabling easy inference. We find that the MNN can search for bit sequences in arbitrary, ultra-broadband10 GBit/sec digital data, demonstrating suitability for high-speed wireline communication.Notably, it can emulate high-level digital functions without custom on-chip circuits, potentially replacing power-hungry sequential logic architectures. Its ability to track frequency changes over long capture times also allows for determining flight trajectories from radar returns. Furthermore, it serves as an accelerator for radio-frequency machine learning, capable of accurately classifying various encoding schemes used in wireless communication. The MNN achieves true, reconfigurable broadband computation, which has not yet been demonstrated by classical analog modalities, quantum reservoir computers using superconducting circuits, or photonic tensor cores, and avoidsthe inefficiencies of electro-optic transduction. Its sub-wavelength footprint in a Complementary Metal-Oxide-Semiconductor process and sub-200 milliwatt power consumption enable seamless integration as a general-purpose analog neural processor in microwave and digital signal processing chips.
more »
« less
Controlling Refraction Using Sub-Wavelength Resonators
We construct metamaterials from sub-wavelength nonmagnetic resonators and consider the refraction of incoming signals traveling from free space into the metamaterial. We show that the direction of the transmitted signal is a function of its center frequency and bandwidth. The directionality of the transmitted signal and its frequency dependence is shown to be explicitly controlled by sub-wavelength resonances that can be calculated from the geometry of the sub-wavelength scatters. We outline how to construct a medium with both positive and negative index properties across different frequency bands in the near infrared and optical regime.
more »
« less
- PAR ID:
- 10092999
- Date Published:
- Journal Name:
- Applied Sciences
- Volume:
- 8
- Issue:
- 10
- ISSN:
- 2076-3417
- Page Range / eLocation ID:
- 1942
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract The study of the impedance mismatch between the device and its surroundings is crucial when building an acoustic device to obtain optimal performance. In reality, a high impedance mismatch would prohibit energy from being transmitted over the interface, limiting the amount of energy that the device could treat. In general, this is solved by using acoustic impedance matching layers, such as gradients, similar to what is done in optical coatings. The simplest form of such a gradient can be considered as an intermediate layer with certain qualities resting between the two media to impedance match, and requiring a minimum thickness of at least one quarter wavelength of the lowest frequency under consideration. The desired combination(s) of the (limited) available elastic characteristics and densities has traditionally determined material selection. Nature, which is likewise limited by the use of a limited number of materials in the construction of biological structures, demonstrates a distinct approach in which the design space is swept by modifying certain geometrical and/or material parameters. The middle ear of mammals and the lateral line of fishes are both instances of this method, with the latter already incorporating an architecture of distributed impedance matched underwater layers. In this paper, we develop a resonant mechanism whose properties can be modified to give impedance matching at different frequencies by adjusting a small set of geometrical parameters. The mechanism in question, like the lateral line organ, is intended to serve as the foundation for the creation of an impedance matching meta-surface. A computational study and parameter optimization show that it can match the impedance of water and air in a deeply sub-wavelength zone.more » « less
-
Orthogonal frequency division multiplexing (OFDM) is a candidate technique to provide high-speed data transmissions for optical communication systems. For intensity modulation and direct detection (IM/DD) optical communication systems, the peak transmitted power limitation of light sources and nonnegative transmitted signal constraints can result in nonlinear distortions from clipping. In this paper, we propose a clipping enhanced optical OFDM (CEO-OFDM) for IM/DD communication systems to reduce the clipping effects. CEO-OFDM transmits the information that results from clipping the peak power, which allows the use of a higher modulation index to improve the signal to noise ratio in exchange for a larger bandwidth. For the same transmitted data rate, CEO-OFDM can achieve a lower bit error rate than DC-biased optical OFDM (DCO-OFDM), asymmetrically clipped optical OFDM (ACO-OFDM) and unipolar OFDM (U-OFDM). By using a larger modulation constellation size, the proposed CEO-OFDM can support a higher throughput than other techniques when the same bit error rate is achieved.more » « less
-
Linear ultrasonic (LU) techniques used by majority of the researchers working in material damage monitoring, are reliable for detecting relatively large defects. If the defect dimensions are in the range of the ultrasonic wavelength or larger, then those defects can be detected by analyzing the scattered ultrasonic fields. Material damage affects LU parameters such as ultrasonic wave speed and attenuation and their changes are detectable for relatively large defects. However, for small defects (when defect dimensions are significantly smaller than the wavelength of the propagating signal) the changes in the LU parameters are too small to detect or measure reliably. For detecting small defects engineers often use high frequency ultrasonic signals to make the defect dimensions greater than the wavelength. However, high frequency signals attenuate quickly and therefore, only very small regions near the ultrasonic probe can be inspected in this manner. Inspecting large structures by high frequency ultrasonic signals requires moving the probe mechanically from one point to the next and therefore, can be very time consuming. Nonlinear ultrasonic (NLU) technique on the other hand does not need to satisfy this restricting condition that the wavelength of the signal must be smaller than the defect size. NLU works well when the signal wavelength is much larger than the defect size. Therefore, relatively low frequency signals that can propagate a long distance and monitor a large area of a structure can be used for NLU measurements. Different NLU techniques can be used for detection and monitoring of small damages in a specimen. A relatively new NLU technique called the Sideband Peak Count-Index or SPC-I technique has been developed by the author and his collaborators. SPC-I technique for monitoring different types of materials – composites, metals, concrete, and other cement-based materials has been found to be effective. Along with those success stories of SPC-I the effect of topography and the advantage of topological sensing is also discussed in this presentation.more » « less
-
Abstract The powerful high‐frequency/very high frequency radio emissions that occur during lightning flashes can be used as a signal of opportunity to study the bottom side ionosphere. The lightning emission is bright, broad spectrum, and short in duration, providing an ideal signal of opportunity for making ionograms. This study continues previous work in Obenberger et al. (2018), where the direct line of sight signal from lightning can be cross correlated with megahertz frequency radio telescope observations to reveal ionogram traces created from the reflected lightning signals. This process was further developed to automate production of ionograms made from individual lightning flashes over the course of several hours, as well as create new techniques to detect the lightning signal using the all‐sky‐imaging mode. By using the Long Wavelength Array Sevilleta radio telescope as an interferometer, the point of reflection of the lightning signal for each frequency of the ionogram can be located in the ionosphere, instantaneously revealing density gradients within the ionosphere on minute time scales. We also explore the minimum size stations required for the application of this technique, which we found to be at least 32 antennas.more » « less
An official website of the United States government

