A Cubic Metric for Inhibiting Signals in Deep Neural Networks

Abstract

The invention discloses a method for suppressing a signal cube metric by a deep neural network, aiming to provide a method for suppressing CM with good CM suppression characteristics and signal distortion performance, better compromise performance, and a short time consumption. The present invention is realized through the following technical scheme: After the signal is mapped to the constellation of the receiving end of the OFDM system, the signal is converted into the OFDM multi-channel signal through the first serial-to-parallel conversion unit and sent to the deep neural network The unit uses the neural network parameters obtained by training to suppress the cubic metric CM of the input signal, and then performs an inverse fast Fourier transform on the signal output by the deep neural network unit through the IFFT module to obtain a parallel time-domain signal, which is transmitted to the second parallel The serial conversion unit converts it into a time-domain signal, and the obtained time-domain signal is converted into an analog signal by a digital-to-analog converter DAC and sent to the PA port of the power amplifier for transmission.

Description

technical field [0001] The invention belongs to the technical field of wireless communication, and in particular relates to a method for reducing the cubic metric of an OFDM signal by using a deep learning neural network [0002] technical background [0003] Excessive signal envelope fluctuation is a major problem in Orthogonal Frequency Division Multiplexing (OFDM). OFDM systems have two performance indicators to measure signal envelope fluctuations: peak-to-average power ratio PAPR and cubic metric (CM). Compared with Power Peak-to-Average Ratio (PAPR), cubic metric can more accurately predict the power backoff of PA, so it is considered to be a more effective metric to measure the envelope variation of Orthogonal Frequency Division Multiplexing (OFDM) signals. In the amplifier circuit, the cubic nonlinear component of the amplification gain is the main cause of ACLR, that is to say, this cubic term is the cause of channel distortion, third harmonic, and in-band interferen...

AI Technical Summary

Problems solved by technology

[0009] The purpose of the present invention is to optimize the problem of excessive time overhead of the algorithm when OFDM signals are used in traditional schemes, and provide a depth with good CM suppression characteristics and signal distortion performance, better compromise performance, and very short time consumption A Cubic Metric Method for Inhibiting Signals in Neural Networks

Images