No abstract

Abstract

GMSK-like signal, with 2 MHz of bandwidth in both cases to strengthen the possible appearance of memory effects. The objective is to calibrate the invariance of the model to a different input signal type, using in this case a modulation format similar to the 2G mobile system, GSM. The NMSE values of the modeled predictions were evaluated Ϫ14 and Ϫ10 dBm as a function of the unit sampling time, providing values of Ϫ32.5 and Ϫ22.5 dB, respectively, with no correction applied to calculate the error. Similar results were obtained for the MP model, allowing to conclude that the VBW model has a performance as good as the MP when validated with a different signal of this type. An improvement of about 0.8 -1 dB was obtained for VBW against MP, when the error was calculated with respect to an optimum amplitude-and time-adjusted version of the original output signal [8]. In this case, the NMSE values for the VBW model were in the range of Ϫ38 and Ϫ31 dB for the input levels of Ϫ14 and Ϫ10 dBm, respectively. The normalized measured and modeled spectra for the higher output level are depicted in Figure 4 for the case Q ϭ 2 showing an excellent agreement with experimental results. The real and imaginary parts of the PA output complex envelope are represented in Figure 5 for the same input level. CONCLUSIONPerformance of the Volterra behavioral model for wideband amplifiers has been extensively analyzed in this work. The VBW model had been theoretically demonstrated in a previous work by the authors exploiting the knowledge of the amplifier behavior in the frequency domain. A 2-MHz-wide WCDMA-like waveform was used as probing signal to model a 915 MHz, 34 dB-gain amplifier. The experimental envelope output samples of the amplifier driven near the 1-dB compression point were used to extract the Volterra kernels and comparison of the calculated and measured signals has shown a high correspondence with normalized error of Ϫ41 dB. Even in a moderately nonlinear operation at Ϫ10 dBm, the error remains small, in the order of Ϫ33.5 dB. For the two input power levels, the VBW model certainly outperforms the results of the MP model. ACKNOWLEDGMENTSThis work was supported by the Spanish National Board of Scientific and Technological Research (CICYT) measurements of scattering matrix parameters over a large dynamic range (90 dB or more), it employs a heterodyning receiver technique involving a frequency synthesized source and a double frequency converter. Because of a large size and a high price tag, its use is restricted to laboratory environments. As many microwave applications require measurements in confined spaces or harsh environments, there is a need to have its portable and low-cost alternative. The low-cost six-port reflectometer introduced in [1-3] can serve such a purpose. This device has one port reserved for a microwave source, the second port for a device under test (DUT), and the remaining four ports terminated with (scalar) power detectors. By performing mathematical transformations on power values measured at four port...

This work provides design and analysis of a beam steering reflectarray antenna, designed at 26 GHz, based on the mechanical rotation of the array. Unit cells based on circular ring elements are designed, and scattering parameter measurements have been carried out for obtaining progressive phase distribution. The unit cell measurements demonstrated a maximum reflection loss of 4 dB with a total phase range of almost 360 • . Symmetricity of the array and the resonant elements has been exposed to steer the main beam by tilting the reflectarray at different angles. A beam steering range of more than ±60 • has been demonstrated by varying the tilt angle from +30 • to −30 • . The designed 20×20 element array provided a maximum gain of 26.47 dB at 0 • which reduced to 19.8 dB at 61.9 • in the elevation plane. On the other hand, the reflectarray antenna demonstrated a maximum bandwidth of 13.1% with a minimum side lobe level of −25.9 dB.

This work provides design and characterization of millimeter wave reflectarray antenna based on unit cells as well as periodic reflectarray design at 26 GHz. Diverse unit cell design configurations have been investigated based on simulations and scattering parameter measurements for feasibility of designing an optimum performance 5G reflectarray antenna. It has been demonstrated that the rectangular patch element provided minimum reflection loss of 1.02 dB and maximum bandwidth of 560 MHz. However, the phase error for rectangular patch element was observed to be 80°, which is much higher as compared to 10° and 13° in the case of rectangular ring and circular ring elements respectively. Periodic reflectarray antennas were also designed with main beam focused at 40° in the azimuth plane achieved with tilted array instead of using an offset feed. Radiation pattern measurements were carried out for further characterization where a maximum gain of 26.7 dB was provided by reflectarray designed with circular ring elements with variable radius. On the other hand, rectangular patch element array provided higher 1 dB gain drop bandwidth of 13.6% as compared to circular ring element reflectarray, which demonstrated a bandwidth of 13.1%. However, side lobe levels were observed to be higher at −18.4° for rectangular patch element reflectarray as compared to −19.4° in the case of circular ring elements based reflectarray antenna.