Behavior Modeling Method of Power Amplifier Based on Clock-Recurrent Neural Network

Abstract

The invention discloses a behavioral modeling method of a power amplifier based on a clock cycle neural network, which overcomes the problems of a large number of iterations of a common neural network model, poor performance on long-term memory effects, and the like. This method uses the characteristic that the output of the recurrent neural network is not only related to the immediate input, but also related to the historical input, and is used to describe the memory effect of the power amplifier. On this basis, the weight of the hidden layer of the ordinary cyclic neural network is divided into several modules, each module has its own cycle, and the weight of this module only updates the weight in its own cycle, thereby reducing the number of weight updates to speed up Training of neural network models. This method can well describe the nonlinear characteristics and memory effect of the power amplifier, and has high precision.

Description

technical field [0001] The invention relates to a power amplifier behavior modeling method based on a clock cycle neural network, belonging to the technical field of nonlinear system modeling and analysis applications. Background technique [0002] The power amplifier is an important module of the transmitter and is a complex nonlinear system. In order to make the power amplifier work with high efficiency, the transistors in the power amplifier mostly work in the near-saturation region or even the cut-off region, so the power amplifier often produces serious nonlinear distortion, and because of the influence of the equivalent reactance of the device, the power amplifier will produce memory effect. [0003] The modeling methods of power amplifiers can be divided into physical modeling and behavioral modeling. Physical modeling needs to know the specific structure inside the circuit and a proficient grasp of circuit knowledge to establish; while behavioral modeling only need...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the deficiencies in the prior art, the object of the present invention is to provide a power amplifier behavior modeling method based on a clock cycle neural network, which solves the problem that the traditional neural network model only performs well in describing short-term memory effects, and is not good for long-term memory effects. Describes poorly behaved problems, and is good at describing power amplifier nonlinearities and memory effects

Images