Hoi-Shun Lui

The non-ionizing nature of microwave radiation and the low cost of microwave electronics offer innovative solutions for medical diagnosis, treatment, and health monitoring

Mutual coupling between antennas and their surrounding environment has been a recognized phenomenon since the inception of antennas [1]. When an antenna element within an array operates as a transmitter, its radiated field induces electromagnetic currents in nearby elements, leading to reradiation. This coupled field can distort radiation patterns and change the terminating impedance. Both effects are typically considered as undesirable, although they may also be exploited to achieve higher performance in some cases. During receiving operations, the received signals comprise both the desired incoming signals and unwanted coupled fields from adjacent antennas and surrounding objects [2].

At microwave frequencies, radar cross-section (RCS) measurements are usually performed by placing the target in the far-field region of the antenna. The wavefront of the radiating field from the antenna can be approximated as planar, ensuring that the incident field and the power interact with the target independently of the antenna. However, for electrically large targets, the required distance becomes significant, posing challenges for implementation. Scaled-model RCS measurements offer an alternative solution. RCS measurements at terahertz and optical frequencies typically require a collimated beam as the source, where the intercepted power and RCS become dependent on the excitation. To address this dependency, researchers have proposed modifying the RCS definition to account for the intercepted power and to analytically formulate the scattering problem. However, such modifications require prior knowledge of the target’s geometry and material properties, which are often not readily available in practice. This also limits the study to only canonical targets. In this paper, we propose an alternative approach for modelling the intercepted power. The Gaussian beam is decomposed into a number of plane waves travelling to different directions using the theory of plane wave spectrum. The scattering problem is solved using the full-wave method of moment. Through theoretical proofs and numerical examples involving spheres and a non-canonical target, with a scaled-model aircraft, we demonstrate that the original RCS definition can serve as a good approximation for scaled measurements, provided that the beam waist is approximately four times the target’s dimensions. These findings provide valuable guidelines for radar engineers when performing scaled measurements using collimated beams. The results, which match those obtained from full-model measurements, enable us to predict the RCS of full-scale targets. This capability facilitates various target-related applications, such as target characterization, classification, detection, and even recognition.

A comparison of the Matrix Pencil Method and the Cauchy Method is studied. The two methods both serve as natural resonant frequency extraction algorithm but is based on information in different domains. Their performances are analyzed using synthesized data and full-wave simulated data. Both methods present the capability to extract NRFs from targets. Numerical results are presented to show the similarity and differences between the two methods.

The choice of matching medium can directly affect the amount of power transmitted into and through human bodies, which impacts on the quality of the reconstructed images in microwave-based medical imaging, and thus the diagnosis accuracies. In this paper, the amount of the transmitted and reflected power of multi-layer planar healthy and hemorrhagic stroke brain models under different choices of matching medium are determined. These multi-layer planar brain tissue models are made up of at least 17 tissue layers with realistic dielectric properties. A layer of blood with different thickness is introduced to model the case of hemorrhagic stroke. Our results show that matching medium with low relative permittivity provides better intrinsic contrast between healthy brain and hemorrhagic brain. Analysis on how the reflection coefficient and transmittance vary under different matching medium are included to provide better insights to the findings.

Reliable radar target recognition has long been the holy grail of electromagnetic sensors. Target recognition based on the singularity expansion method (SEM) uses a time-domain electromagnetic signature and has been well studied over the last few decades. The SEM describes the late time period of the transient target signature as a sum of damped exponentials with natural resonant frequencies (NRFs). The aspect-independent and purely target geometry and material-dependent nature of the NRF set make it an excellent feature set for target characterization. In this chapter, we aim to review the background and the state of the art of resonance-based target recognition. The theoretical framework of SEM is introduced, followed by signal processing techniques that retrieve the target-dependent NRFs embedded in the transient electromagnetic target signatures. The extinction pulse, a well-known target recognition technique, is discussed. This chapter covers recent developments in using a polarimetric signature for target recognition, as well as using NRFs for subsurface sensing applications. The chapter concludes with some highlights of the ongoing challenges in the field.

Algebraic Reconstruction Algorithms (ARAs) is a family of algorithms that reconstructs computed tomography (CT) images in an iterative manner. Due to the ill-posedness nature of the reconstruction problem, any small noise in the projection data can cause large error in the reconstruction result. Therefore, it is important to study how the noise in the projection data could affect the outcome of the reconstructions. In this paper, projection data sets with different noise levels will be used to illustrate how noise would impact the image reconstruction process. Image reconstruction performance of two of the ARAs, namely the Algebraic Reconstruction Technique (ART) and Simultaneous Iterative Reconstructive Technique (SIRT), under different settings will be studied using Monte-Carlo simulation. Our findings show that the existence of noise affects the convergence behavior of the algorithms. With a proper choice of relaxation factor, both algorithms are able to reach to a similar level of accuracy. Comparing between the two algorithms, SIRT shows a superior stability over ART which is preferred. The findings provide us with further insights on how noise will impact the convergence behavior of the algorithms, as well as practical guidelines for optimizing the image reconstruction performance. Such findings are useful when we apply the algorithms to projection data sets which are highly contaminated by noise.

In microwave-based medical diagnosis, reflections of electromagnetic waves from the air–skin interface can be reduced with a proper choice of matching medium such that electromagnetic radiation can be effectively penetrated into and through human bodies. Due to the heterogeneous and lossy nature of tissue and the fact that the electromagnetic problem varies as a function of matching medium, frequency, incident angles and polarization, it is not trivial to analyze the entire propagation phenomena. In this study, transmission of electromagnetic wave through a multi-layer planar human tissue model with different matching medium is considered. The reflected, absorbed and transmitted power components are computed to quantify the transmission performance of the matching medium. Our results show that although reflection can be minimized through a proper choice of matching medium, it does not necessary result in an optimal transmission as the transmitted signal is severely attenuated inside the tissue. Our results also show that the oblique incidence plane wave components, which in general correspond to the non-zero spatial frequency components of a near field distribution, can better penetrate through the tissue when a low-permittivity matching medium is used. These findings suggest that a better signal integrity can be obtained when a low-permittivity matching medium ( ε r between 5 and 30) is used.

This article investigates the accuracy of direction-of-arrival (DOA) estimations using compact antenna arrays under the influence of mutual coupling. We model the entire electromagnetic problem using the moment method such that the undesirable mutual coupling effect is taken into account accurately. We decouple the received signals using the receiving-mutual-impedance method (RMIM) and perform the DOA estimation using the matrix pencil method (MPM) through a Monte Carlo simulation approach. Although the signals are strongly distorted by mutual coupling, we show that increasing the number of elements within a small array aperture can significantly improve the estimation accuracy. The findings show that the RMIM can effectively compensate for mutual coupling for arrays with an aperture of a factor of a wavelength. Such findings are crucial to the development of compact arrays in various applications.

Radar cross‐section (RCS) measurements of complex, large objects are usually performed on scale models so that the measurement is carried out in a well‐controlled environment. The feasibility of RCS measurement using a terahertz (THz) quantum cascade laser (QCL) via laser feedback interferometry is explored. Numerical simulations show that the RCS information embedded in the nonlinear interferometric signals obtained from simple targets can be retrieved through numerical fitting of the well‐known excess phase equation. The method is validated experimentally using a THz QCL and the results are well matched with those obtained from numerical simulations.