Bit error ratio estimation using machine learning

1. A test and measurement system, comprising:

a machine learning system;

a test and measurement device including a port configured to connect the test and measurement device to a device under test (DUT); and

one or more processors, and a memory for storing code, which when executed by the one or more processors, causes the one or more processors to:

acquire one or more waveforms from the DUT by capturing one or more signals from the DUT and generating the one or more waveforms from the one or more signals;

transform the one or more waveforms into a composite waveform image, the composite waveform image based on at least two images from the one or more waveforms; and

send the composite waveform image to the machine learning system to obtain a bit error ratio (BER) value for the DUT.

2. The test and measurement system as claimed in claim 1 , wherein the composite waveform image comprises one of a short pattern tensor, a cyclic loop tensor, or an eye diagram overlay.

3. The test and measurement system as claimed in claim 1 , further comprising a known device connected to the DUT, the known device to transmit a known pattern along a transmission path to the DUT and receive a pattern from the DUT along a reception path.

4. The test and measurement system as claimed in claim 3 , wherein the port comprises a first channel input of the test and measurement device connected to the transmission path and a second channel input of the test and measurement device connected to the reception path.

5. The test and measurement system as claimed in claim 1 , wherein the one or more processors are further configured to execute code to cause the one or more processors to train the machine learning system, the code causing the one or more processors to:

acquire a training waveform associated with a signal, the training waveform having a BER value has previously been measured and stored;

transform the training waveform into a training composite waveform image; and

provide the training composite waveform image and the stored BER value to the machine learning system as a training data sample.

6. The test and measurement system as claimed in claim 5 , wherein the processors are further configured to repeat the code to cause the one or more processors to train the machine learning system until a sufficient number of training samples have been obtained.

7. The test and measurement system as claimed in claim 1 , wherein the one or more processors are further configured to execute code to cause the one or more processors to apply a dynamic range compression on pixels in the composite waveform image.

8. The test and measurement system as claimed in claim 7 , wherein the dynamic range compression comprises one of either a gamma correction or a log function.

9. The test and measurement system as claimed in claim 1 , wherein the one or more processors are further configured to execute code to cause the one or more processors to obtain a temperature when the waveform from the DUT is acquired and send the temperature to the machine learning system with the composite waveform image.

10. A method of determining a bit error ratio for a device under test (DUT), comprising:

acquiring one or more waveforms from the DUT by receiving one or more signals from the DUT and generating the one or more waveforms from the one or more signals;

transforming the one or more waveforms into a composite waveform image, the composite waveform image based on at least two images from the one or more waveforms; and

sending the composite waveform image to a machine learning system to obtain a bit error ratio (BER) value for the DUT.

11. The method as claimed in claim 10 , wherein transforming the one or more waveforms into the composite waveform image comprises transforming the one or more waveforms into one of a short pattern tensor, a cyclic loop tensor or an eye diagram overlay.

12. The method as claimed in claim 10 , further comprising: connecting a known device the DUT;

using the known device to transmit a known pattern along a transmission path to the DUT; and

receiving a pattern from the DUT along a reception path.

13. The method as claimed in claim 12 , further comprising:

connecting a first channel of a test and measurement device to the transmission path;

connecting a second channel of the test and measurement device to the reception path; and

comparing waveforms acquired from the first channel and the second channel.

14. The method as claimed in claim 10 , further comprising:

connecting a loop back path between a transmitter and a receiver on the DUT;

measuring the BER for that DUT;

storing the BER for that DUT;

acquiring a training waveform from the DUT;

transforming the training waveform into a training composite waveform image; and

providing the training composite waveform image and the stored BER value to the machine learning system as a training data sample.

15. The method as claimed in claim 14 , further comprising repeating the method on additional DUTs until a sufficient number of training data samples have been obtained.

16. The method as claimed in claim 10 , further comprising applying a dynamic range compression on pixels in the composite waveform image.

17. The method as claimed in claim 16 , wherein applying a dynamic range comprises applying one of a gamma correction or a log function to the values in the training composite waveform image.

18. The method as claimed in claim 10 , wherein acquiring one or more waveforms from the DUT further comprises obtaining a testing temperature.

19. The method as claimed in claim 18 , wherein transforming the one or more waveforms into the composite waveform image further comprises encoding a graphical representation of the testing temperature into the composite waveform image.