arXiv subjects

Ehsan Afshari

Publications and source records attributed to Ehsan Afshari.

6 recordsLinked to original sources

arXiv2026

A Study on THz Plasmonics in a CMOS Continuum Transistor Array

This work addresses the limitations of CMOS at terahertz (THz) frequencies, where charge transit time and parasitic capacitances restrict the maximum operating frequency, fmax. As transistor dimensions shrink, reduced current handling capabilities further challenge CMOS, necessitating novel circuit design approaches for the THz domain. By leveraging the plasma characteristics of electron channels in CMOS transistors, this study explores a potential solution for THz signal amplification. Key mechanisms in plasma wave amplification within a continuum transistor array (CTA) formed by 28 nm fully depleted silicon-on-insulator (FD-SOI) CMOS transistors are investigated. A hydrodynamic transport model combined with Pierce's theory is presented to describe plasma wave propagation along the CTA. Simulations of gated amplifiers demonstrate the potential for THz signal amplification in advanced fabrication nodes. Finally, a proof-of-concept plasma wave amplifier operating at 700 GHz has been designed and fabricated, exhibiting amplification along the plasma wave propagation path.

eess.SY↗

arXiv2026

Super-resolution ranging using a sub-terahertz self-injection-locked frequency-modulated radar

Sub-terahertz (sub-THz) and terahertz (THz) frequency-modulated continuous-wave (FMCW) radars have opened a plethora of scientific and industrial applications, especially in the imaging field. While strong candidates for sub-THz/THz FMCW radar imagers are implemented using photonic methods, there is a desire to achieve the full integration and portability that only electronics can offer. However, integrated electronic sub-THz/THz FMCW radars have significantly lower bandwidth (< 100 GHz) than photonic-based radars, restricting the radar range resolution to the millimeter scale (> 1.5 mm). In addition, the electronic FMCW radar's broad bandwidth comes with increased transmitter phase noise, consequently degrading the radar range accuracy. Here, we present a sub-THz fully-integrated autodyne frequency-modulated (AFM) radar utilizing a self-injection locking (SIL) mechanism that fundamentally overcomes the aforementioned challenges of FMCW radars. The AFM radar supports an exceptionally wide effective bandwidth extending into the terahertz sweep range by forming an intermediate frequency comb spectrum in a quadratic receiver, unlocking the path for super-resolution ranging. Furthermore, SIL significantly reduces the transmitter's phase noise, allowing high-accuracy range measurements. We theoretically describe and experimentally demonstrate the SIL operation of the AFM radar. The proposed radar experimentally achieves sub-millimeter range resolution and a range accuracy of < 0.002%, enabling the imaging of covered printed letters with micrometer features.

eess.SP↗

arXiv2023

FuNToM: Functional Modeling of RF Circuits Using a Neural Network Assisted Two-Port Analysis Method

Automatic synthesis of analog and Radio Frequency (RF) circuits is a trending approach that requires an efficient circuit modeling method. This is due to the expensive cost of running a large number of simulations at each synthesis cycle. Artificial intelligence methods are promising approaches for circuit modeling due to their speed and relative accuracy. However, existing approaches require a large amount of training data, which is still collected using simulation runs. In addition, such approaches collect a whole separate dataset for each circuit topology even if a single element is added or removed. These matters are only exacerbated by the need for post-layout modeling simulations, which take even longer. To alleviate these drawbacks, in this paper, we present FuNToM, a functional modeling method for RF circuits. FuNToM leverages the two-port analysis method for modeling multiple topologies using a single main dataset and multiple small datasets. It also leverages neural networks which have shown promising results in predicting the behavior of circuits. Our results show that for multiple RF circuits, in comparison to the state-of-the-art works, while maintaining the same accuracy, the required training data is reduced by 2.8x - 10.9x. In addition, FuNToM needs 176.8x - 188.6x less time for collecting the training set in post-layout modeling.

cs.LG↗

arXiv2017

On Probability of Support Recovery for Orthogonal Matching Pursuit Using Mutual Coherence

In this paper we present a new coherence-based performance guarantee for the Orthogonal Matching Pursuit (OMP) algorithm. A lower bound for the probability of correctly identifying the support of a sparse signal with additive white Gaussian noise is derived. Compared to previous work, the new bound takes into account the signal parameters such as dynamic range, noise variance, and sparsity. Numerical simulations show significant improvements over previous work and a closer match to empirically obtained results of the OMP algorithm.

cs.IT↗

arXiv2016

A New Performance Guarantee for Orthogonal Matching Pursuit Using Mutual Coherence

In this paper we present a new coherence-based performance guarantee for the Orthogonal Matching Pursuit (OMP) algorithm. An upper bound for the probability of correctly identifying the support of a sparse signal with additive white Gaussian noise is derived. Compared to previous work, the new bound takes into account the signal parameters such as dynamic range, noise variance, and sparsity. Numerical simulations show significant improvements over previous work.

cs.IT↗

arXiv2016

Smart Detector Cell: A Scalable All-Spin Circuit for Low Power Non-Boolean Pattern Recognition

We present a new circuit for non-Boolean recognition of binary images. Employing all-spin logic (ASL) devices, we design logic comparators and non-Boolean decision blocks for compact and efficient computation. By manipulation of fan-in number in different stages of the circuit, the structure can be extended for larger training sets or larger images. Operating based on the mainly similarity idea, the system is capable of constructing a mean image and compare it with a separate input image within a short decision time. Taking advantage of the non-volatility of ASL devices, the proposed circuit is capable of hybrid memory/logic operation. Compared with existing CMOS pattern recognition circuits, this work achieves a smaller footprint, lower power consumption, faster decision time and a lower operational voltage. To the best of our knowledge, this is the first fully spin-based complete pattern recognition circuit demonstrated using spintronic devices.

cs.ET↗