Search

Accurate and robust localization is a fundamental need for mobile agents. Visual–inertial odometry (VIO) algorithms exploit the information from the camera and inertial sensors to estimate position and translation. Recent deep-learning-based VIO models attract attention as they provide pose information in a data-driven way, without the need of designing hand-crafted algorithms. Existing learning-based VIO models rely on recurrent models to fuse multimodal data and process sensor signals, which are hard to train and not efficient enough. We propose a novel learning-based VIO framework with external memory attention that effectively and efficiently combines visual and inertial features for state estimation. Our proposed model is able to estimate pose accurately and robustly, even in challenging scenarios, for example, on overcast days and water-filled ground, which are difficult for traditional VIO algorithms to extract visual features. Experiments validate that it outperforms both traditional and learning-based VIO baselines in different scenes. [ABSTRACT FROM AUTHOR]

The design of a W-band BCB-air cavity-backed antennas is presented in this letter. The three-dimensional (3-D) capacitance model of the cavity-backed structure is analyzed and the design methodology is illustrated in detail. The proposed antenna is fabricated by using our in-house silicon-based MEMS photosensitive composite film fabrication process by combining localized backside etching technology and through silicon ring trench. The measured impedance bandwidth (−10 dB) of the antenna is from 88.9 to 93.7 GHz with a maximum gain 8.2 dBi and 95% radiation efficiency at 91 GHz, higher the gain and radiation efficiency by 2.2 dB and 35%, respectively, as compared with the case without TSRT. The measured results show close agreement with the simulated results. The proposed antenna can be further 3-D heterogeneously integrated with other components because of low interconnection loss to enhance the high-frequency performance for future millimeter wave radar and communication systems. [ABSTRACT FROM AUTHOR]

In this article, quarter-mode substrate-integrated waveguide (QMSIW) filters featuring ultrawide stopband and flexibly allocated transmission zeros (TZs) are proposed. Eight higher modes with resonant frequencies up to 4.7 times of the fundamental mode are highly suppressed. It is realized by the combination of three approaches without extra circuit areas. First, a slot is cut in the first and last QMSIW resonators to make their resonant frequencies of higher modes different from those of other resonators. Second, avoid exciting the specific higher modes by setting the feed point in the area where the electric field strength of the higher modes is weak. Third, mismatch the higher modes when matching the passband. It is realized by using a stripline impedance transformer with its main body not exceeding the area of QMSIW resonators. Furthermore, except for the TZs derived by the classical cross-coupling matrix, two more independently controlled TZs are generated. The packaged filter is fabricated by using a benzocyclobutene (BCB)-based microelectromechanical systems (MEMS) process, which is suitable for millimeter-wave 3-D system-on-package (SOP) applications. The fabricated thin-film filter is measured with an insertion loss of 2.96 dB at 29.4 GHz and a fractional bandwidth (BW) of 17%. Higher mode suppressions up to 138.3 GHz are achieved. [ABSTRACT FROM AUTHOR]

In this article, a discontinuous Galerkin time-domain (DGTD) method hybridized with Robin transmission condition (RTC-DGTD) is proposed for the transient thermal analysis of 3-D integrated circuits (ICs). As a kind of domain decomposition method (DDM), the DGTD method employs a term named numerical flux for the information exchange of solutions between neighboring subdomains. The numerical flux is a function of both temperature $T$ and heat flux q, and thus for the heat equation, apart from $T$ , the variable q has to be solved as well. To reach this, the traditional DGTD method decomposes the second-order partial-differential thermal equation into two first-order ones, which results that both $T$ and q need to be solved in the whole computational domain, thereby the computational cost is extremely expensive. In this work, to get this trouble-treated, the proposed DGTD method is directly applied to discretize the second-order heat equation. An RTC is established at the interfaces of subdomains, which serves as the second governing equation and meanwhile bridges another connection of solutions in adjacent subdomains. In this way, the new variable q is now only present at the interfaces of subdomains. Consequently, the total number of unknowns is significantly reduced compared with that in the traditional DGTD method. To further speed up the time-marching scheme, an adaptive time-stepping method in terms of the proportional–integral–derivative (PID) control algorithm is employed. Besides, by introducing a restriction operator, the globally coupled matrix equation is torn into a number of small matrix equations pertinent to each subdomain, which is later solved via a finite-element tearing and interconnecting (FETI)-like method. To validate and verify the accuracy and efficiency of the proposed algorithm, several representative examples are studied and compared with COMSOL simulations. [ABSTRACT FROM AUTHOR]

In this letter, a two-material-filled ridge half-mode substrate integrated waveguide (2M-RHMSIW) is proposed for monomode bandwidth (MMB) enhancement. The cutoff frequency of the dominant mode decreases with the permittivity of the material filling in the bottom layer increasing. The cutoff frequency of the second mode increases as the permittivity of the material filled in the top layer decreases. Therefore, the proposed structure uses low-permittivity material for the top substrate layer and high-permittivity material for the bottom substrate layer, to widen the MMB. In addition, a series of air holes are drilled in the top substrate layer to reduce its effective permittivity, which further widens the MMB. A prototype of the proposed structure is fabricated under PCB process, which MMB is from 2 to 14 GHz. [ABSTRACT FROM AUTHOR]

In this work, to characterize the radiated emission from electronic circuits in shielding enclosure, an improved electric dipole-based source reconstruction method (SRM) is developed. Moreover, by resorting to this reconstructed equivalent source, the estimation of electromagnetic interference (EMI) between different circuit modules in the enclosure can be conveniently and accurately evaluated. Different from the free-space SRM, the equivalent dipoles of the proposed SRM are directly placed over the shielding box enclosed circuit board, and the numerical Green’s function (NGF) is developed to bridge the connection between the equivalent dipoles and the planar scanned magnetic near-field. With the NGF strategy, the effects of surrounding environments (including the presence of substrate, ground plane, and shielding enclosure) are inclusively considered, which makes the proposed SRM valid for any complex EM environments in theory. Since the proposed SRM also has the capability of reproducing the radiated emission inside the shielding enclosure, it is critically helpful to evaluate the EMI between different circuit modules inside the shielding enclosure. To reach this aim, the Rayleigh-Carson Reciprocity theorem is referred, in which the original EM coupling problem is decomposed into the “forward problem” and “reverse problem.” A Huygens box enclosing the victim circuit module is built and the EM fields on the box are obtained using the reconstructed dipoles in both problems. Then the EMI can be calculated accordingly. Therefore, the novelty and merits of the proposed approach are threefold: 1) the physically-based equivalent dipole model considers the interactions between the radiation sources and the surrounding objects, leading to the accurate prediction of radiated emissions both outside and inside the enclosure; 2) due to the utilization of the NGF, there is no approximation involved in the solution of the dipole’s radiation, thus improving the calculation precision; and 3) the near-field coupling between multiple electronic circuit modules inside the shielding enclosure can be identified precisely, which is useful for diagnostics of radiation sources. The effectiveness of this algorithm is verified by representative numerical examples. [ABSTRACT FROM AUTHOR]