Articles in 2023

The Editorial Board and Editorial Team are delighted to present a selection of short Research Highlights describing some of our favourite Communications Engineering publications of 2023.

Carver and colleagues build a neural network-based classifier, which can distinguish the subcategories of the non-rapid eye-movement sleep state in Parkinson’s disease patients, by analysing local field potential recordings from the subthalamic nucleus. By studying the impact of sampling rate and frequency bands, the researchers achieve optimal complexity/accuracy trade-off of their methodology.

Signal processing is key to communications and video image processing for astronomy, medical diagnosis, autonomous driving, big data and AI. Menxi Tan and colleagues report a photonic processor operating at 17Tb/s for ultrafast robotic vision and machine learning.

Fufei An and colleagues report a solution-based strategy to prepare large-area, freestanding amorphous quasi-2D carbon nanomembranes from coal-derived carbon dots. Their mechanical and dielectric properties make these attractive materials for applications in nanoelectronic devices such as memristors and 2D-based transistors.

Accurate control of excitation and inhibition balance change is important for brain perceptual and motor learning. Yokoyama and Kitajo report a data assimilation-based method to track the excitation and inhibition balance change via analysing electroencephalography signals. The method was validated using data for human beings in sleep.

Sun and colleagues report real-time hippocampal calcium imaging for jointly capturing spatial, visual, and auditory information. Using machine learning techniques, they demonstrate a rapid and accurate all-optical brain-computer interface for decoding mice’s hippocampus activity.

Lu and coauthors use two orthogonal polarisations to separate downstream and upstream data flows in connected fibre-free-space optics-5G wireless communication. They achieve an undisruptive signal transmission and enable a high-throughput interface.

Liu, Zhang and Fang propose and experimentally demonstrate an optimization method for shape morphing of origami spring structures to approximate complex 3D curves by introducing virtual creases.

Mueller matrix microscopy is capable of mapping tissue architecture at the subcellular level. Wan, Dong and colleagues report an unsupervised learning approach to identify pathological structures by clustering polarization features in Muller matrix images. This approach enables the identification of microstructures subtypes invisible in nonpolarized optical images.

Yanqing Su, Wanmin Lin colleagues report the construction of an instance-based learning classifier by non-specific hybridization of DNA sequences both in simulation and in a small-scale experiment. The approach is simple, robust and scalable to larger classification applications.

Costa and co-authors detected an earthquake in Mexico using conventional polarisation optics within a trans-oceanic fibre-optic cable connecting Los Angeles, USA with Valparaiso, Chile. Their approach enables non-invasive monitoring and localization of seismic waves on the seabed.

Bablich and colleagues report an Intrinsic Photomixing Detector (IPD), a cost-effective amorphous silicon device, enabling high-bandwidth, long-distance, and high accuracy Time-of-Flight optical ranging even in low light conditions. The device allows for large-scale integration on silicon or flexible platforms.

Anton Molina and colleagues demonstrate the fabrication of absorbent materials from the drought tolerant plant agave sisalana (sometimes known as sisal). The findings suggest a route toward local production of the absorbent component of menstrual pads in low- and middle-income countries, where there is more limited access to sanitary products.

Yao Xiao and colleagues report a graph-based machine learning algorithm for categorising and then more efficiently distributing code segments across different processing units in heterogeneous hardware platforms. The approach leads to more efficient processing, vital for future developments harnessing AI, for example in fields such as autonomous vehicles and machine vision.

Oscar Recalde-Benitez and colleagues report a FIB-based sample preparation process to limit current leakage during operando TEM experiments, thus improving the accuracy of device nanocharacterization under operating conditions. The methodology results in leakage currents that are small compared to device currents, which enables the analysis of operating stack devices inside the microscope.

Inspired by the construction of insect eyes, Gkanias and colleagues built a navigation sensor which finds the North using skylight polarisation. An array of polarization analysers detects the Sun’s location even in cloudy skies, providing a low-complexity compass for autonomous vehicles and robots.

Yuyang Wei and colleagues use 3D printing to fabricate tactile sensors with an auxetic structure design capable of detecting both forces and temperature for use in prosthetics and other robotic applications.

Nanomaterial solutions to sustainable development goals are hindered in their path to commercialisation by an early-stage reliance on single metric optimisation. Here we formulate the PSEC challenge (Performance, Scalability, Environment and Cost) to integrate broader sustainability thinking with precise technical solutions and so enable successful commercialisation of these advanced materials.

Elastocaloric cooling devices utilize the latent heat that shape memory alloys generate when subjected to stress to generate large temperature variations. Wang and colleagues discuss key challenges of the practical application of elastocaloric cooling systems including materials fabrication, force actuation and heat transfer enhancement.

Gate-all-round field effect transistors (FETs) with channels fabricated from highly stacked nanowires can enhance the drive current of such devices for a fixed footprint. Chen, Liu, and colleagues fabricated FETs with as many as 16 highly stacked Ge_0.95Si_0.05 nanowires and 12 nanowires without parasitic channels using wet etching.