RESUMO
At present, most millimeter wave 5G systems operate at frequencies ranging from 24 GHz to 39 GHz. Nevertheless, the new 5G release is going to increase the supported 5G spectrum into the 60 GHz band. In this communication, we discuss a methodology of converting a modular antenna array, which was originally designed for the 17 GHz unlicensed band, to the emerging 60 GHz 5G band. The antenna array is of a modular architecture, allowing a simple extension to a higher number of elements. The feeding structure is composed of substrate-integrated waveguides (SIW), allowing relatively simple and cheap manufacturing. As revealed by a sensitivity analysis, the frequency up-conversion significantly increases the need for precision of the used manufacturing technology. Subsequently, the structure is optimized to minimize the repercussions of manufacturing variations. The properties of the converted array are studied when equipped with parasitic patches to increase the axial ratio (AR) bandwidth.
RESUMO
In the paper, a 3D knitted fabric is used for the design of a circularly polarized textile-integrated antenna. The role of the radiating element is played by a circular slot etched into the conductive top wall of a textile-integrated waveguide. Inside the circular slot, a cross slot rotated for about 45° is etched to excite the circular polarization. The polarization of the antenna can be changed by the rotation of the cross slot. The antenna has a patch-like radiation pattern, and the gain is about 5.3 dBi. The textile-integrated feeder of the antenna is manufactured by screen printing conductive surfaces and sewing side walls with conductive threads. The antenna was developed for ISM bands 5.8 GHz and 24 GHz. The operation frequency 24 GHz is the highest frequency of operation for which the textile-integrated waveguide antenna has been manufactured.
RESUMO
This work presents and evaluates a 12-electrode intracranial electroencephalography system developed at the National Institute of Mental Health (Klecany, Czech Republic) in terms of an electrical source imaging (ESI) technique in rats. The electrode system was originally designed for translational research purposes. This study demonstrates that it is also possible to use this well-established system for ESI, and estimates its precision, accuracy, and limitations. Furthermore, this paper sets a methodological basis for future implants. Source localization quality is evaluated using three approaches based on surrogate data, physical phantom measurements, and in vivo experiments. The forward model for source localization is obtained from the FieldTrip-SimBio pipeline using the finite-element method. Rat brain tissue extracted from a magnetic resonance imaging template is approximated by a single-compartment homogeneous tetrahedral head model. Four inverse solvers were tested: standardized low-resolution brain electromagnetic tomography, exact low-resolution brain electromagnetic tomography (eLORETA), linear constrained minimum variance (LCMV), and dynamic imaging of coherent sources. Based on surrogate data, this paper evaluates the accuracy and precision of all solvers within the brain volume using error distance and reliability maps. The mean error distance over the whole brain was found to be the lowest in the eLORETA solution through signal to noise ratios (SNRs) (0.2 mm for 25 dB SNR). The LCMV outperformed eLORETA under higher SNR conditions, and exhibiting higher spatial precision. Both of these inverse solvers provided accurate results in a phantom experiment (1.6 mm mean error distance across shallow and 2.6 mm across subcortical testing dipoles). Utilizing the developed technique in freely moving rats, an auditory steady-state response experiment provided results in line with previously reported findings. The obtained results support the idea of utilizing a 12-electrode system for ESI and using it as a solid basis for the development of future ESI dedicated implants.
