Programa del congreso

This contribution presents the nonlinear characterization up to 50 GHz of a PIN diode commonly used as switch for reconfigurable devices at microwave applications. Nonlinear models are extracted by means of X-parameter measurements supported on accurate calibration and de-embedding procedures.

Results are validated by S-parameter measurements in the low power signal regime and by harmonic measurements in the large signal regime. The use of these models for reconfigurable antennas assessment is discussed.

The implementation of a non-quasi-static model for Schottky diodes into a commercial Computer-Aided Design tool is described. This formulation relies on a non-quasi-static charge definition that depends on two quasi-static functions: charge and delay. The model considers the use of the Nonlinear Function Sampling operator in order to extract each of this quasi-static components. This model implementation has been tested under Harmonic Balance analysis with some numerical examples. The performance of the proposed element observed in these preliminary tests allows being optimistic in the upcoming validation under realistic experimental conditions.

In this paper, a direct learning architecture is employed for the digital predistortion of a power amplifier with a concurrent dual-band input signal. Conventional dual-band behavioral models need to cope with a large number of regressors, being prone to algebraic issues that arise in the computation of the inverse of the covariance matrix. Ridge regularization is applied to the modelling equation in order to improve the stability of the learning process and the precision of the estimated coefficients in validation. Experimental results for the linearization performance achieved with the proposed approach are provided for a class AB power amplifier driven by two concurrent 10-MHz 5G New Radio signals.

The implementation of Wireless Communication Systems through Software Defined Radio (SDR) platforms are haven special attention from the researcher community. Our research team has developed a Co-operative System SDR platform to demonstrate the viability and advantages to use A\&F and D\&F protocols in LTE network. However, currently, the 5G Mobile Communications are receiving enormous attention and our platform should be upgraded to increase the required capacities to implement New Radio (NR) relaying technologies. In this paper, the nonlinearity study of Power Amplifier (PA) to increasing the benefits of our Relay Node Co-operative System SDR platform has been performed. The study consists in several experiments in a realistic laboratory environment. Besides, other tests conducted, Non-Line-of-Sight (NLOS) scenario, 64-QAM modulation scheme, as well as LTE signal are considered. From the results, the Error Vector Magnitude (EVM) of the enhanced SDR platform shows better performance and allows to increase the reach of the Backhaul Link (Base Station-to-Relay Node Link). Furthermore, this work constitutes the next stage for the implementation of a 5G NR Co-operative System SDR platform.