Method for modeling synchronous double-frequency power amplifier based on real number time delay neural network

Abstract

The invention provides a method for modeling a synchronous double-frequency power amplifier based on a real number time delay neural network, which comprises the following steps of: establishing a real number time delay neural network model of the synchronous double-frequency power amplifier; taking two paths of broadband multi-carrier signals as baseband input signals of the synchronous double-frequency power amplifier; acquiring input and output data of the synchronous double-frequency power amplifier; and determining the parameter of the real number time delay neural network model and training the model. The nonlinear characteristic and memory effect of the synchronous double-frequency power amplifier can be well described, and the method for modeling the synchronous double-frequency power amplifier based on the real number time delay neural network is convenient to implement, and has low complexity and high convergence rate and accuracy.

Description

technical field [0001] The invention relates to a method for establishing a behavior model of a dual-frequency power amplifier, in particular to a modeling method for a synchronous dual-frequency power amplifier based on a real number delay neural network. Background technique [0002] With the continuous growth of wireless communication system services, limited spectrum resources become increasingly crowded. In order to improve spectrum utilization efficiency, various modulation techniques have emerged one after another. Although the new modulation method improves the utilization rate of spectrum, it also puts forward higher requirements on the RF front end. Most of these modulation methods are non-constant envelope, and the higher the efficiency of the modulation method, the larger the peak-to-average ratio of the signal. In order to make the power amplifier work with high efficiency, the power amplifier mostly works in the saturation region or even the cut-off region. Fo...

AI Technical Summary

Problems solved by technology

[0008] However, most of the current power amplifier modeling is for single-frequency power amplifiers, that is, for single-input and single-output power amplifier modeling, and less for dual-frequency power amplifiers, that is, for multiple-input and multiple-output power amplifier modeling

Because the signals of the two frequency bands in the dual-frequency power amplifier will have mutual influence (such as inter-band cross modulation) that the single-frequency power amplifier has never encountered, which increases the nonlinear effect of the component. The output model description is more complex than single-input and single-output, which is inconvenient to implement and has low precision

Images