In this paper, a balanced-to-balanced power divider is proposed, for the first time, with arbitrary power division, which can be regarded as the balanced form of a Gysel power divider. The constraint rules are provided for its mixed-mode and single-ended S-parameters. The six-port network is analyzed by simplifying it to two-port networks with other ports matched at the central frequency. Its critical characteristic impedances are then calculated analytically by our derived equations according to the desired differential-mode power division ratio. The maximum achievable power division ratio is 1:4.69² when the characteristic impedances are limited within the realizable range of 20-120 Ω. The impacts of several freely selected design parameters on the operating bandwidth are explored numerically. A prototype is realized by microstrip lines and lumped resistors with the power division ratio of 1:3². The balanced-to-balanced performances of unequal power division, low differential-mode insertion loss, good suppression of common-mode noises, and mode conventions have been demonstrated by the simulated and measured results of the balanced-to-balanced power divider prototype.

Frequency selective surface (FSS) is critical for electromagnetic (EM) radiation protection due to its high spatial filtering performance, especially for active FSS. Recently, the artificial neural network (ANN) has shown great potential in solving EM inverse problems and rapid industrial design. In such an inverse model with ANN, it establishes the relationship between the given inputs of S-parameters and the desired structure parameters or material parameters. However, faced with applications where S-parameters vary in a large frequency range with different curve shapes, such as multiband microwave devices, equal interval sampling may result in high-dimensional inputs and will require a more complicated neural network. In this work, we present a Fourier subspace-based deep learning method (FS-BDLM) for FSS inverse design, where the dimension of the input is largely reduced by using Fourier subspace to represent the most salient features of the desired S-parameter performance. Compared with existing deep learning methods, the proposed technique makes inverse neural models more compact and more stable to noise contaminations. The validation of the proposed FS-BDLM is conducted both numerically and experimentally through two dual-passband FSS design examples, where the well-designed FSS is fabricated to validate the technique.

A dual-band differential shifted-feed microstrip grid array antenna (DSF-MGAA) with two parasitic patches is presented in this communication. By moving the feed location, a conventional MGAA (C-MGAA) can be turned into a SF-MGAA for dual-band operation. The high-frequency band of SF-MGAA corresponds to the operation mode of the C-MGAA, and its low-frequency band is similar to that of the patch or meshed patch antenna. A new differential feeding scheme and two parasitic patches are adopted to improve the impedance bandwidth of SF-MGAA. The proposed DSF-MGAA can cover a dual-band of 3.7 and 28.5 GHz. The measured -10 dB impedance bandwidths of 3.4% and 4.5%, which correspond to the peak gains of 8.98 and 19.2 dBi, are obtained in the low-and high-frequency bands, respectively. The DSF-MGAA can be a potential candidate for the future dual-band wireless communication.

Substrate integrated coaxial line (SICL) is a kind of planar transmission line, which can guarantee the propagation of transverse electromagnetic mode wave within a wide band. In this paper, a multichannel SICL array with coaxial lines placed in both vertical and horizontal directions is proposed for high-speed parallel data transmission. The adjacent channels of the SICL array share outer conductors for high-density integration. Moreover, the outer conductor is enhanced by metal fence to suppress the crosstalk between adjacent channels of the SICL array. In addition, a quasi-coaxial transition is designed for the proposed SICL array. A systematical design procedure is given to determine the physical dimensions of SICL array with transitions. The SICL array has flat group delay so that data can be transmitted without significant phase distortion. It is suitable for high-density integration and high-speed data transmission due to the low crosstalk and high bandwidth. To verify the advantage of the proposed SICL array, a 15 x 3 channel prototype is designed and fabricated in low temperature co-fired ceramic. The measured S-parameters agree well with the simulated results. The maximum bandwidth of a single channel of the SICL array is up to 56 GHz and the coupling of adjacent channels in the SICL array is less than -30 dB. Data rate of 30 Gb/s is measured for each channel with good eye diagram and low bit error rate, and hence, a total 1.35-Tb/s data rate can be achieved with the proposed 15 x 3 channel SICL array.

Frequency transfer is a key challenge in machine learning as it allows researchers to go beyond in-range analyses of spectrum properties towards out-of-the-range predictions. Traditionally, to predict properties at a specific frequency, targeted spectrum is included in training data for a deep neural network (DNN). However, due to limitations of measurement or computation source, training data at some frequencies are hardly accessible, especially for multi-physics problems. In this work, we propose a multi-physics deep learning framework (MDLF) consisting of a multi-fidelity DeepONet, a Euler latent dynamic network, and a data-analytical inversion network. Without the knowledge about multi-physics response, MDLF is successfully generalized to unseen frequency bands for both parametric and free-form metasurface by dynamically utilizing a Euler latent space and single-physics information. Moreover, an inversion method is introduced to incorporate hybrid a priori in inverse design of metasurface. Under EM-thermal coupling, we verify the proposed MDLF numerically and experimentally.

A new type of half-mode substrate integrated waveguide (SIW) periodically loaded with different lumped elements and structures is proposed in this paper. The propagation constants, Bloch impedances, and voltage distributions of the Floquet TE-modes are all characterized by solving the eigenvalues of the generalized transmission matrix of a single period. Several typical capacitor-loaded, inductor-loaded, corrugated, and electromagnetic-bandgap (EBG)-loaded half-mode SIWs were fabricated to demonstrate their modal dispersion behavior. The measured and simulated S -parameters, group delays, and simulated electric field distributions in these novel structures are presented, with good agreements obtained for all samples. In particular, their unique controllable cutoff frequencies, multiple bandgaps, and separated monomode regions are also illuminated numerically and experimentally. To further show their applications, a tunable phase shifter is developed using a varactor-loaded structure, with its 3-dB bandwidth of 50% and a maximal tunable range of 88°. In addition, a bandpass filter based on EBG-loaded half-mode SIW is also realized and presented with its special characteristics. It demonstrates that the filter has a sharp upper slope side and a good performance of spurious suppression.