Transmitter and signal transmission method thereof

Abstract

The transmitter includes a first neural network circuit and a front-end circuit system. The first neural network circuit runs a first neural network model to perform a peak factor reduction algorithm on a baseband signal according to a peak setting parameter to generate a first signal. The front-end circuit system generates an output signal based on a second signal. The second signal is generated by performing a predistortion algorithm on the first signal to compensate for signal distortion generated by the front-end circuit system, and the peak factor reduction algorithm is used to reduce the power peak of the output signal.

Claims

1. A transmitter comprising: a first neural network circuit for running a first neural network model to perform a peak factor reduction algorithm on a baseband signal according to a peak setting parameter to generate a first signal; and a front-end circuit system for generating an output signal according to a second signal, wherein the second signal is generated by performing a predistortion algorithm on the first signal, the predistortion algorithm being used to compensate for a signal distortion generated by the front-end circuit system, and the peak factor reduction algorithm being used to reduce a power peak of the output signal.

2. The transmitter of claim 1, wherein the first neural network circuitry is further configured to run the first neural network model to execute the predistortion algorithm, thereby generating the second signal based on the first signal.

3. The transmitter of claim 1 further includes: a second neural network circuit for running a second neural network model to execute the predistortion algorithm, thereby processing the first signal to generate the second signal.

4. The transmitter of request item 1, wherein the first neural network model uses a hyperbolic tangent function to implement a hard clipping in the peak factor reduction algorithm.

5. The transmitter of claim 1, wherein the first neural network circuitry comprises: a memory circuitry for providing at least one coefficient based on the peak setting parameter; and a neural network engine circuitry for running the first neural network model based on the at least one coefficient, thereby performing the peak factor reduction algorithm on the baseband signal to generate the first signal.

6. The transmitter as requested in item 1, wherein the peak setting parameter is used to set the upper limit of the power peak.

7. The transmitter of Request 1, wherein the peak factor reduction algorithm includes out-of-band spectral shaping.

8. A signal transmission method, comprising: running a first neural network model via a first neural network circuit to perform a peak factor reduction algorithm on a baseband signal according to a peak setting parameter to generate a first signal; and generating an output signal via a front-end circuit system based on a second signal, wherein the second signal is generated by performing a predistortion algorithm on the first signal, the predistortion algorithm being used to compensate for a signal distortion generated by the front-end circuit system, and the peak factor reduction algorithm being used to reduce a power peak of the output signal.

9. The signal transmission method of claim 8 further includes: running the first neural network model by the first neural network circuit to execute the predistortion algorithm, thereby processing the first signal to generate the second signal.

10. The signal transmission method of claim 8 further includes: running a second neural network model via a second neural network circuit to execute the predistortion algorithm, thereby processing the first signal to generate the second signal.

Images