In this letter, a GaAs-based wideband spoof surface plasmon polariton (SSPP) waveguide with super compact size is presented at millimeter-wave regime. The periodic SSPP unit cell consists of a longitudinal slot and a meander slot in transversal direction, and its corresponding waveguide has good low-pass filtering feature. The momentum of microstrip line for measurement feeding to the SSPPs can be well matched without any gradient transition structures. To apply this SSPP waveguide into the design of bandpass filter (BPF), a row of vias is added with a certain distance from the SSPP unit cells. For demonstration, two on-chip examples, i.e., SSPP waveguide and SSPP-based BPF, are fabricated operating at different frequencies. Good agreement between simulations and measurements validates the propagation characteristics and implementation of the proposed SSPPs.
13 July 2020 in IEEE Transactions on Electromagnetic Compatibility
The field prediction of an unknown electromagnetic interference (EMI) source within a complex electromagnetic environment requires complex radiation field formulas. In this article, we propose an artificial neural network (ANN) method to predict the field by using the scanned phaseless near-field from the EMI source. The near-field magnitude is scanned using a near-field probe first. After that, an ANN is trained to present the mapping between the observation point and the radiation field. A set of free-space Green's functions are used as the input of the ANN, and the magnitude of the radiation field is taken as the output of the ANN. With the help of Green's function, the trained ANN can accurately predict the radiation field outside the scanning regions. The feasibility of this method is verified using numerical and measurement experiments. The proposed method can realize source reconstruction in a complex electromagnetic environment.
As a crucial frequency selection device in modern communication systems, the microwave filter plays an increasingly prominent role. There is a great demand for the multi-objective design of microwave filters. The filter’s performance affects the quality of the whole communication system directly. However, traditional multi-objective electromagnetic (EM) optimization design demands repetitive EM simulations to adjust the physical parameters of the microwave filters. Accordingly, using electromagnetic simulation directly to design optimization is quite expensive. Given this situation, this paper applies a novel surrogate model based on one-dimensional convolutional autoencoders (1D-CAE) into the multi-objective algorithm evolutionarily based on decomposition (MOEA/D) for the first time. This approach uses MOEA/D as the multi-objective optimizer, and a novel low-complexity surrogate model based on one-dimensional convolutional autoencoders (1D-CAE) is constructed to predict the expensive EM simulation results. The surrogate model based on 1D-CAE is used to generate the results of scalar subproblems of MOEA/D, which greatly improves the design efficiency. Compared with the traditional design methods based on an EM solver, this method not only effectively optimizes multiple design objectives but also completes the design of microwave filters in a shorter time. The proposed method is verified using the design of a sixth-order ceramic filter and a seventh-order metal cavity filter.
13 March 2023 in IEEE Antennas and Wireless Propagation Letters
A new approach by using an N +3 coupling matrix for synthesizing the filtering antenna is proposed. Compared with the traditional one, the coupling matrix synthesis method mainly is applied to design filters. Due to the introduction of transistor parameters, the N +3 coupling matrix achieves the prediction of filtering performance and gain characteristics for the filtering antenna. This filtering antenna, as a multifunctional device with filtering and radiation functions, can therefore be well combined with the proposed matrix to achieve the integrated synthesis of the filtering antenna. Secondly, a novel multimode broadband filtering patch antenna is designed as an example to verify the effectiveness of this theory. Three modes and a patch mode are excited within the passband, resulting in a wide operating bandwidth. Multiple paths between the input/output stub lines are formed by three resonant modes and direct source-antenna coupling to produce four radiation nulls outside the desired band, achieving good out-of-band suppression. Finally, the measured results of the fabricated antenna prototype have demonstrated a good filtering response in the 3.44-3.72 GHz band. Specifically, this proposed antenna has achieved 6.9 dBi gain and out-of-band suppression above 30 dB, which is suitable for 5G communication.
30 September 2020 in IEEE Journal on Multiscale and Multiphysics Computational Techniques
This paper proposes a new method to predict the radiation field in a complex environment. The magnetic field probe is used to scan the magnitude of the radiation field, and a cascade-forward neural network (CFNN) is introduced to establish the non-linear relationship between the Green's function and the magnitude of the radiation field. The trained CFNN can replace the real radiation source in a complex environment and produce the same radiation field, which is helpful to analyze electromagnetic interference (EMI) problems. The innovation of this paper is to establish an equivalent source based on only scanning the magnitude of the radiation field. At the same time, the electric current source and the magnetic current source of the radiator are equivalent to electric dipoles and magnetic dipoles, respectively. After CFNN training, we can predict the radiation field beyond the scanning plane. Simulation and measurement examples are used to verify the effectiveness of the proposed method.
20 December 2021 in International Journal of RF and Microwave Computer-Aided Engineering