12 February 2015 in Optics Express
Graphene’s relatively poor absorption is an essential obstacle for designing graphene-based photonic devices with satisfying photo-responsivity. To enhance the tunable light absorption of graphene, appropriate excitation of localized surface plasmon resonance is considered as a promising approach. In this work, the strategy of incorporating periodic cuboid gold nanoparticle (NP) cluster arrays and cylindrical gold NP arrays with Bragg reflectors into graphene-based photodetectors are theoretically studied by the boundary-integral spectral element method (BI-SEM). With the BI-SEM, the models can be numerically analyzed with excellent accuracy and efficiency. Numerical simulation shows that the proposed structures can effectively engineer the light absorption in graphene by tuning plasmon resonance. In the spectra of 300 nm to 1000 nm, a maximum light absorption of 67.54% is observed for the graphene layer with optimal parameters of the photodetector model.
22 September 2017 in Optics Express
In this work, we propose a numerical solver combining the spectral element - boundary integral (SEBI) method with the periodic layered medium dyadic Green’s function. The periodic layered medium dyadic Green’s function is formulated under matrix representation. The surface integral equations (SIEs) are then implemented as the radiation boundary condition to truncate the top and bottom computation domain. After describing the interior computation domain with the vector wave equations, and treating the lateral boundaries with Bloch periodic boundary conditions, the whole computation domains are discretized with mixed-order Gauss- Lobatto-Legendre basis functions in the SEBI method. This method avoids the discretization of the top and bottom layered media, so it can be much more efficient than conventional methods. Numerical results validate the proposed solver with fast convergence throughout the whole computation domain and good performance for typical multiscale nano-optical applications.
4 March 2020 in IEEE Transactions on Antennas and Propagation
An efficient and accurate spectral-element spectral-integral method is developed for rigorously modeling arbitrary inhomogeneous objects with doubly periodicity embedded in a layered medium background. It is an improvement over the traditional finite-element boundary-integral method. In the proposed approach, the domain decomposition between the interior spectral-element domain and the outer surface boundary-integral domain enables the boundary-integral domain to be solved by a far more efficient spectral integral method using a uniform mesh; without any loss of generality, it leaves the interior spectral-element domain with flexibility to model any inhomogeneous objects. For the boundary-integral domain, with a uniform mesh and rooftop basis functions, we can exploit the Toeplitz structure of the boundary-integral matrices with the fast Fourier transform algorithm, and apply the Calderón preconditioner to achieve a high iteration convergence rate that will not slow down as the problem gets larger. With the periodic layered medium Green’s functions, the layers without objects can be treated as background and the discretization is avoided. We show that this method is very promising for large-scale problems, because the CPU time and memory required with respect to the number of unknowns grows slowly. We validate this method with a commercial finite element solver and show its ability by solving a large-scale optical lithography problem.
7 July 2016 in IEEE Transactions on Antennas and Propagation
In order to directly utilize the dyadic Green's functions in surface integral equations (SIEs), a novel double-arctan transformation for nearly hypersingular integrals is proposed in this communication. This new transformation is flexible and applicable to nearly hypersingular integrals in the forms of Ȓ Ȓ/R 3 , Ȓ/R 3 , and 1/R 3 over the curved surfaces by a fully numerical method. With the help of the sigmoidal transformation to improve the stability of this new singularly handling method, there results an efficient solution for the third-order near-singularity problems in SIEs. Moreover, the proposed method is also effective for the lower orders of the nearly singular integral kernels. With typical testing cases, the performance of this method is fairly evaluated, and its validity and stability is well demonstrated.
16 June 2022 in IEEE Transactions on Computational Imaging
Existing deep compressive sensing (CS) methods either ignore adaptive online optimization or depend on costly iterative optimizer during reconstruction. This work explores a novel image CS framework with recurrent-residual structural constraint, termed as $\mathrm{R}^{2}$CS-NET. The $\mathrm{R}^{2}$CS-NET first progressively optimizes the acquired samplings through a novel recurrent neural network. The cascaded residual convolutional network then fully reconstructs the image from optimized latent representation. As the first deep CS framework efficiently bridging adaptive online optimization, the $\mathrm{R}^{2}$CS-NET integrates the robustness of online optimization with the efficiency and nonlinear capacity of deep learning methods. Signal correlation has been addressed through the network architecture. The adaptive sensing nature further makes it an ideal candidate for color image CS via leveraging channel correlation. Numerical experiments verify the proposed recurrent latent optimization design not only fulfills the adaptation motivation, but also outperforms classic long short-term memory (LSTM) architecture in the same scenario. The overall framework demonstrates hardware implementation feasibility, with leading robustness and generalization capability among existing deep CS benchmarks.
8 March 2022 in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
Exploration wells are liquid-filled boreholes drilled into formations with different geophysical and petrophysical properties. These boreholes support axisymmetric, flexural, and quadrupole family of guided modes that can probe radially varying formation properties at different frequencies. Radially varying formation properties are caused by drilling induced fractures or near-wellbore stress concentrations. This work describes a novel workflow that inverts borehole flexural and Stoneley dispersions to obtain radially varying formation mass density, shear and bulk moduli away from the borehole surface. An integral equation relates fractional changes in guided mode velocities at different frequencies caused by fractional changes in radially varying mass density, shear and bulk moduli from a radially uniform reference state. A solution of this integral equation is based on extending the Backus-Gilbert (B-G) method for obtaining radial profile of a single to radial profiles of three formation properties away from the borehole surface. Inverted radial profiles from synthetic flexural and Stoneley dispersions have been validated against input formation parameters used to generate the synthetic (measured) dispersions.