Bourns College of Engineering

A scheme for secure communications, called “Secret-message Transmission by Echoing Encrypted Probes (STEEP)”, is revisited. STEEP is a round-trip scheme with a probing phase from one user to another and an echoing phase in the reverse direction. STEEP is shown to be broadly applicable to yield a positive secrecy rate in bits per channel use even if the receive channels at eavesdropper (Eve) are stronger than those between legitimate users in both forward and reverse directions. This paper focuses on STEEP in the following settings: using Gaussian probing signal and Gaussian linear encryption over MIMO Gaussian channel (G-STEEP); using phase-shift-keying probing signal and a nonlinear encryption over SISO channel (P-STEEP); and a variation of G-STEEP for multiple access communication (M-STEEP). In each of the settings, Eve is assumed to have any given number of antennas, and STEEP is shown to yield a positive secrecy rate subject to a sufficiently large power in the echoing phase, as long as Eve’s receive channel in the probing phase is not noiseless. It is also shown that G-STEEP, subject to asymmetric large powers in forward and reverse directions, has its secrecy rate approaching the secret-key capacity based on Gaussian probing signal over MIMO Gaussian channel. STEEP does not require secure feedback channel, collaborative third party, in-band full-duplex or reciprocal channels between users, but only needs a design for echoing encrypted probes, asymmetric power allocation and/or collaborative round-trip coding.

Virus-based nanocarriers have shown great potential for noninvasive delivery of drugs, diagnostics, and imaging agents to hard-to-reach anatomical locations. Yet, they largely depend on diffusion for transport, often lacking the force to actively penetrate biological barriers, and navigation to guide therapeutic agents. In these studies, the M13 bacteriophage, a linearly shaped virus, was converted from passive nanocarrier to actively propelled, fuel-driven nanomotor. Using the distinctive low symmetry of its capsid, a single Pt nanoparticle was added to one end of the M13 virus to form a tadpole-like structure. The Pt/M13 head/tail nanomotors exhibited notably enhanced diffusion in the presence of hydrogen peroxide fuel, and significantly improved uptake by SVOK3 ovarian cancer cells in vitro. Given the successes of the M13 bacteriophage as a nanocarrier, the demonstration of this simple, but comparatively mobile M13-based nanomotor platform represents an important step in advancing the potential therapeutic efficacy of viral nanocarriers.

The rapid growth of generative AI has placed significant strain on traditional data center infrastructures and existing power grids, leading to soaring energy demands and environmental burdens that may disproportionately affect the local communities. Shifting AI inference from the cloud to edge devices could potentially reduce the reliance on network connections, enhance user privacy, and alleviate the escalating pressure data centers impose on the local electricity grid. In this work, we present a case study examining the environmental footprint and energy consumption when deploying a generative AI model on cloud and edge platforms. To this end, we model and evaluate the water consumption and carbon emissions associated with AI inference across these deployment scenarios. Our empirical results demonstrate that, for several state-of-the-art generative AI models deployable on both cloud and edge devices, a reduced environmental footprint is observed for edge platform deployments. More specifically, edge platforms can achieve over 90% energy savings while reducing carbon emissions and water consumption by more than 80%. Putting the accuracy and latency performance aside, these findings highlight the potential of edge inference to lower the energy demands and environmental footprint of generative AI compared to cloud-based inference.

Microorganisms (e.g., bacteria, fungi, and viruses) add indispensable functionality to a range of electrospun polymer materials and devices. The optimal distribution of bioactive agents on either the interior or exterior of the fiber is application-specific. Current microbe surface immobilization strategies and core-confinement techniques continue to pose a number of challenges. Here, we explore a simple strategy, utilizing electrostatic forces, to control the migration and surface concentration of the M13 bacteriophage within near-field electrospun poly(vinyl alcohol) (PVA) microfibers. Both the surface charge of the electrospun virus and the applied electric field polarity altered microbe placement. When doped with Rhodamine 6G (R6G), the circular microfiber cross sections formed active whispering gallery mode (WGM) resonators. These relatively high-quality (Q) optical cavities enabled us to sensitively probe the virus content of their outer layer, while functioning as label-free optical biosensors with phage-based streptavidin biorecognition elements. Coulomb forces displayed significant control over M13 surface coverage during near-field electrospinning, increasing biosensor response by nearly a factor of 4 to 1310 nM streptavidin. These findings are an important demonstration of electrostatic forces as a simple, yet adaptable method to enhance biohybrid fiber functionality and performance by tailoring microbe distribution.

The ending of Moore's Law calls for innovations in integrated circuit (IC) technologies and chip designs. Heterogeneous integration (HI) emerges as a pathway towards smart future chips for more Moore time and for beyond-Moore time, featuring systems-on-integrated-chiplets (SoICs) and advanced micro-packaging (μ-packaging). Reliability, particularly with regard to electrostatic charge (ESD) failure, is a major challenge for 3D SoIC chips in μ-packaging, which is an emerging design-for-reliability challenge for future chips. This perspective article articulates that interposer-based ESD protection will be an important potential solution for 3D SoIC chips in μ-packaging against the devastating ESD failure problem.

Interplay between topological electrons and magnetic ordering enables efficient electrical control of magnetism. We extend the Kane-Mele model to include the exchange coupling to a collinear antiferromagnetic (AFM) order, which allows the system to exhibit the quantum anomalous Hall and quantum spin Hall effects in the absence of a net magnetization. These topological phases support a staggered Edelstein effect through which an applied electric field can generate opposite non-equilibrium spins on the two AFM sublattices, realizing the Néel-type spin-orbit torque (NSOT). Contrary to known NSOTs in AFM metals driven by conduction currents, our NSOT arises from pure adiabatic currents devoid of Joule heating, while being a bulk effect not carried by the edge currents. By virtue of the NSOT, the electric field of a microwave can drive the AFM resonance with a remarkably high efficiency, outpacing the magnetic field-induced AFM resonance by orders of magnitude in terms of power absorption.

A magnon is a collective excitation of the spin structure in a magnetic insulator and can transmit spin angular momentum with negligible dissipation. This quantum of a spin wave has always been manipulated through magnetic dipoles (that is, by breaking time-reversal symmetry). Here we report the experimental observation of chiral spin transport in multiferroic BiFeO3 and its control by reversing the ferroelectric polarization (that is, by breaking spatial inversion symmetry). The ferroelectrically controlled magnons show up to 18% modulation at room temperature. The spin torque that the magnons in BiFeO3 carry can be used to efficiently switch the magnetization of adjacent magnets, with a spin–torque efficiency comparable to the spin Hall effect in heavy metals. Utilizing such controllable magnon generation and transmission in BiFeO3, an all-oxide, energy-scalable logic is demonstrated composed of spin–orbit injection, detection and magnetoelectric control. Our observations open a new chapter of multiferroic magnons and pave another path towards low-dissipation nanoelectronics.

The recent proliferation of large language models (LLMs) has led to divergent narratives about their environmental impacts. Some studies highlight the substantial carbon footprint of training and using LLMs, while others argue that LLMs can lead to more sustainable alternatives to current practices. We reconcile these narratives by presenting a comparative assessment of the environmental impact of LLMs vs. human labor, examining their relative efficiency across energy consumption, carbon emissions, water usage, and cost. Our findings reveal that, while LLMs have substantial environmental impacts, their relative impacts can be dramatically lower than human labor in the U.S. for the same output, with human-to-LLM ratios ranging from 40 to 150 for a typical LLM (Llama-3-70B) and from 1200 to 4400 for a lightweight LLM (Gemma-2B-it). While the human-to-LLM ratios are smaller with regard to human labor in India, these ratios are still between 3.4 and 16 for a typical LLM and between 130 and 1100 for a lightweight LLM. Despite the potential benefit of switching from humans to LLMs, economic factors may cause widespread adoption to lead to a new combination of human and LLM-driven work, rather than a simple substitution. Moreover, the growing size of LLMs may substantially increase their energy consumption and lower the human-to-LLM ratios, highlighting the need for further research to ensure the sustainability and efficiency of LLMs.

Traditional systems for indoor pressure sensing and human activity recognition (HAR) rely on costly, high-resolution mats and computationally intensive neural network-based (NN-based) models that are prone to noise. In contrast, we design a cost-effective and noise-resilient pressure mat system for HAR, leveraging Velostat for intelligent pressure sensing and a novel hyperdimensional computing (HDC) classifier that is lightweight and highly noise resilient. To measure the performance of our system, we collected two datasets, capturing the static and continuous nature of human movements. Our HDC-based classification algorithm shows an accuracy of 93.19%, improving the accuracy by 9.47% over state-of-the-art CNNs, along with an 85% reduction in energy consumption. We propose a new HDC noise-resilient algorithm and analyze the performance of our proposed method in the presence of three different kinds of noise, including memory and communication, input, and sensor noise. Our system is more resilient across all three noise types. Specifically, in the presence of Gaussian noise, we achieve an accuracy of 92.15% (97.51% for static data), representing a 13.19% (8.77%) improvement compared to state-of-the-art CNNs.

Among the fascinating phenomena observed in two-dimensional (2D) magnets, the magneto-exciton effect stands out as a pivotal link between optics and magnetism. Although the excitonic effect has been revealed and exhibits a considerable correlation with the spin structures in certain 2D magnets, the underlying mechanism of the magneto-exciton effect remains underexplored, especially under high magnetic fields. Here we perform a systematic investigation of the spin-exciton coupling in 2D antiferromagnetic NiPS3 under high magnetic fields. When an in-plane magnetic field is applied, the exceptional sharp excitonic emission at ~1.4756 eV exhibits a Zeeman-like splitting with g ≈ 2.0, experimentally identifying the exciton as an excitation of dominant triplet-singlet character. By examining the polarization of excitonic emission and simulating the spin evolution, we further verify the correlation between excitonic emission and Néel vector in NiPS3. Our work elucidates the mechanism behind the spin-exciton coupling in NiPS3 and establishes a strategy for optically probing the spin evolutions in 2D magnets.