Neural networks

In this paper, systematic neural network modeling techniques are presented for microwave modeling and design using the concept of inverse modeling where the inputs to the inverse model are electrical parameters and outputs are geometrical parameters. Training the neural network inverse model directly may become difficult due to the nonuniqueness of the input-output relationship in the inverse model. We propose a new method to solve such a problem by detecting multivalued solutions in training data. The data containing multivalued solutions are divided into groups according to derivative information using a neural network forward model such that individual groups do not have the problem of multivalued solutions. Multiple inverse models are built based on divided data groups, and are then combined to form a complete model. A comprehensive modeling methodology is proposed, which includes direct inverse modeling, segmentation, derivative division, and model combining techniques. The methodology is applied to waveguide filter modeling and more accurate results are achieved compared to the direct neural network inverse modeling method. Full electromagnetic simulation and measurement results of Ku-band circular waveguide dual-mode pseudoelliptic bandpass filters are presented to demonstrate the efficiency of the proposed neural network inverse modeling methodology.

/pdf/neural-network-inverse-modeling-and-applications-to-59svabkg4a.pdf

Neural Network Inverse Modeling and Applications to Microwave Filter Design

An electronic device includes an amplifier circuit coupled to a linearizer. The amplifier circuit may receive a first input signal including first and second frequencies and generate a first output signal including a delta frequency signal at a delta frequency, which is the difference between the first frequency and the second frequency. The linearizer includes a signal detector circuit, a current-mirror circuit, a low pass filter, a phase shifter, and a bias circuit. The signal detector circuit may generate a second output signal. The current-mirror circuit may adjust an amplitude of a signal. The low pass filter may eliminate a portion of the second output signal having frequencies greater than the delta frequency. The phase shifter may generate a feedback signal corresponding to the delta frequency signal. An amplitude and/or a phase of the feedback signal is different from an amplitude and/or a phase of the delta frequency signal.

Method and apparatus for reducing intermodulation distortion in an electronic device having an amplifier circuit

One divided signal divided into two by a dividing circuit is inputted to a gate of a source grounded FET through a first matching circuit. In a drain of the FET, a second harmonic having a phase and an amplitude in accordance with an impedance of the first matching circuit is generated and extracted in a band pass filter and then the amplitude is adjusted in an attenuation circuit to input to an addition circuit through a second matching circuit. In the addition circuit, the output of the second matching circuit is added to another divided signal of the dividing circuit and inputted to a power amplifier. The impedance in the first matching circuit affecting the phase of the second harmonic generated from the FET is set so that a distortion component generated in the power amplifier is compensated for by the second harmonic inputted in the addition circuit.

Distortion compensating device and power amplifying device with distortion compensating function

This paper presents a hybrid (passive & active) power amplifier concept for a wideband high drain efficiency power amplifier design. The proposed design integrates for the first time a dual-band PA with an active second harmonic injection to achieve high efficiency across a continuous wideband frequency range of two octaves. The design utilizes a resistively loaded class B at the lower frequencies and a class J mode of operation at the upper frequency band. To maintain high efficiency during the transition between the two PA modes an active second harmonic injection at the output of the main transistor is employed through an addition of an auxiliary low power amplifier. To demonstrate the validity of the novel concept a demonstrator is realized around a 10 W GaN transistor with an average efficiency of 63% across 0.6–2.4 GHz at only modest gain compression of 1dB.

Wide band high-efficiency power amplifier design

A linearization technique that uses the injection of the fundamental signal second harmonics together with the fundamental signals at the amplifier input has been improved in this paper by introducing phase correction of the fundamental signals according to the differences in phases of the second harmonics. As the result, the improvement in third-order intermodulation power levels is approximately 40 dB for analog and 22 dB for digitally modulated signals for multichannel applications.

Improvement in second harmonics linearization technique for multichannel amplifiers

The wave approach to noise modeling of microwave transistors by including the correlation effect

A simple procedure for the extraction of intrinsic noise wave temperatures in the wave representations of microwave transistors is presented in this paper. A set of equations describing the noise parameters as a function of three equivalent noise temperatures is implemented within the circuit simulator Libra. After that, the wave noise model is defined as a new user-defined element of the Libra program library. Good agreement between modeled and measured noise parameters is observed.

MESFET noise modeling based on noise wave temperatures

In this article, a single-stage power amplifier and two-way Doherty amplifier are linearized by the technique that uses second harmonics and fourth-order nonlinear signals. Measurements of the linearization effects on the third- and fifth-order intermodulation products have been carried out. © 2012 Wiley Periodicals, Inc. Microwave Opt Technol Lett 55:425–430, 2012; View this article online at wileyonlinelibrary.com. DOI 10.1002/mop.27294

Linearization of power amplifiers by second harmonics and fourth-order nonlinear signals

In this paper two-way Doherty amplifier with the additional circuit for linearization has been realized and measurements of the linearization influence to the third-and fifth-order intermodulation products have been carried out. The linearization approach uses the fundamental signals' second harmonics and fourth-order nonlinear signals that are extracted at the output of peaking cell, adjusted in amplitude and phase and injected at the output of the carrier cell in Doherty amplifier.

https://ieeexplore.ieee.org/iel5/6089237/6102757/06102893.pdf

Natasa Males-Ilic

Papers

Improvement in second harmonics linearization technique for multichannel amplifiers

The wave approach to noise modeling of microwave transistors by including the correlation effect

MESFET noise modeling based on noise wave temperatures

Linearization of power amplifiers by second harmonics and fourth-order nonlinear signals

Linearization of two-way Doherty amplifier