Device and method for receiving data in a radio frequency transmission

Patent metadata

Field	Value
Publication number	US-11616516-B2
Application number	US-202217664987-A
Country	US
Kind code	B2
Filing date	May 25, 2022
Priority date	Jun 19, 2020
Publication date	Mar 28, 2023
Grant date	Mar 28, 2023

Abstract

According to one aspect, an embodiment radio frequency receiver device comprises an input interface configured to receive a radio frequency signal of a given type and convert same into an electric signal, a detector configured to detect at least one voltage level in the electric signal, a pulse generator configured to generate at least one pulse train representative of the voltage levels detected, and a processing unit configured to determine the type of the radio frequency signal from the at least one pulse train.

First claim

What is claimed is:

1. A method for receiving data in a radio frequency transmission, the method comprising:
receiving a radio frequency signal of a specific type;
converting the radio frequency signal into an electric signal;
detecting, at least once, at least one voltage level in the electric signal;
generating at least one input pulse train representative of each detection;
generating, by an artificial neural network implemented on a processing unit, from the at least one input pulse train, at least one corrected pulse train comprising pulses representative of the specific type of the radio frequency signal; and
determining the specific type of the radio frequency signal from the at least one corrected pulse train.

2. The method according to claim 1 , wherein the artificial neural network comprises a succession of neuron layers, and the generating the at least one corrected pulse train further comprises:
retrieving, by an input neuron layer, the at least one input pulse train;
converting, by at least one hidden convolutional and/or propagation neuron layer, the retrieved at least one input pulse train into the at least one corrected pulse train; and
transmitting, by an output neuron layer, the at least one corrected pulse train.

3. The method according to claim 2 , further comprising each neuron layer applying respective weights to its respective inputs to generate its respective outputs.

4. The method according to claim 3 , further comprising, during an artificial neural network learning phase, adjusting the weights by executing the artificial neural network using already-classified pulse trains from a reference database as input data.

5. The method according to claim 2 , wherein the converting comprises:
recognizing, by the at least one hidden convolutional and/or propagation neuron layer, pulses representative of symbols in the retrieved at least one input pulse train; and
eliminating, by the at least one hidden convolutional and/or propagation neuron layer, pulses in the retrieved at least one input pulse train resulting from noise or transmission errors.

6. The method according to claim 2 , wherein a nature of the radio frequency signal is identified based on the at least one corrected pulse train.

7. The method according to claim 2 , further comprising decoding the radio frequency signal from the at least one corrected pulse train, as a function of the determined specific type of the radio frequency signal.

8. The method according to claim 1 , further comprising, after determining the specific type of the radio frequency signal, sampling then processing the electric signal in accordance with the determined specific type of the radio frequency signal.

9. The method according to claim 1 , wherein the radio frequency signal is selected from among a Wi-Fi signal, a Bluetooth signal, or a Long Term Evolution (LTE) signal.

10. A radio frequency receiver device comprising:
an input interface configured to:
receive a radio frequency signal of a specific type; and
convert the radio frequency signal into an electric signal;

a detector configured to detect at least one voltage level in the electric signal;
a pulse generator configured to generate input pulse trains representative of the voltage levels detected; and
a processing unit configured to:
implement an artificial neural network to generate, from the input pulse trains, corrected pulse trains comprising pulses representative of the specific type of the radio frequency signal; and
determine the specific type of the radio frequency signal from the corrected pulse trains.

11. The radio frequency receiver device according to claim 10 , wherein the artificial neural network comprises a succession of neuron layers including:
an input neuron layer for retrieving the input pulse trains;
at least one hidden convolutional and/or propagation neuron layer for converting the retrieved input pulse trains into the corrected pulse trains; and
an output neuron layer for transmitting the corrected pulse trains.

12. The radio frequency receiver device according to claim 11 , wherein each neuron layer comprises respective weights used to generate its respective outputs from its respective inputs.

13. The radio frequency receiver device according to claim 12 , wherein the weights are pre-trained by executing the artificial neural network using already-classified pulse trains from a reference database as input data.

14. The radio frequency receiver device according to claim 11 , wherein the at least one hidden convolutional and/or propagation neuron layer is configured to:
recognize pulses representative of symbols in the retrieved input pulse trains; and
eliminate pulses in the retrieved input pulse trains resulting from noise or transmission errors.

15. The radio frequency receiver device according to claim 11 , further comprising a decoder configured to identify the specific type of the radio frequency signal from the corrected pulse trains.

16. The radio frequency receiver device according to claim 15 , wherein the decoder is configured to decode the radio frequency signal from the corrected pulse trains, as a function of the determined specific type of the radio frequency signal.

17. The radio frequency receiver device according to claim 10 , wherein the detector comprises at least one comparator configured to detect the at least one voltage level of the electric signal by comparing the electric signal with at least one threshold.

18. The radio frequency receiver device according to claim 10 , further comprising a signal processor configured to, after a nature of the radio frequency signal has been determined, sample and then process the electric signal in accordance with the determined specific type of the radio frequency signal.

19. An object comprising:
a radio frequency receiver device comprising:
an input interface configured to:
receive a radio frequency signal of a specific type; and
convert the radio frequency signal into an electric signal;

a detector configured to detect at least one voltage level in the electric signal;
a pulse generator configured to generate input pulse trains representative of the voltage levels detected; and
a processing unit configured to:
implement an artificial neural network to generate, from the input pulse trains, corrected pulse trains comprising pulses representative of the specific type of the radio frequency signal; and
determine the specific type of the radio frequency signal from the corrected pulse trains.

20. The object according to claim 19 , wherein the artificial neural network comprises a succession of neuron layers including:
an input neuron layer for retrieving the input pulse trains;
at least one hidden convolutional and/or propagation neuron layer for converting the retrieved input pulse trains into the corrected pulse trains; and
an output neuron layer for transmitting the corrected pulse trains.

21. The object according to claim 19 , wherein the object is coupled to a network, and is configured to receive data from a remote server or appliance via the radio frequency signal.

Assignees

• St Microelectronics Sa

Inventors

• Dautriche Pierre
• Engels Sylvain

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• H04B1/16

  Circuits · CPC title

• H04B1/0007 Primary

  wherein the AD/DA conversion occurs at radiofrequency or intermediate frequency stage · CPC title

• H04B1/0475

  with means for limiting noise, interference or distortion (H04B1/0483 takes precedence) · CPC title

• G06N3/084

  Backpropagation, e.g. using gradient descent · CPC title