Signal Analysis / Simulation Team (2 People)

[MaxEllsworth]

For some reason Github won't let me assign this to two people, so this task is meant for @jasoninirio and @yixiuzhu.

We will eventually be collecting data on what normal (non-degraded) C-V2X communications looks like in order to build our machine learning model. This will involve analyzing signals emitted from both the base station and the user-end radios. There are two main programs you will want to use to analyze signal emissions: gqrx and urh (universal radio hacker).

gqrx is for a big-picture analysis of frequency emissions while urh is for a more-granular analysis.

When using gqrx, you will want to acquaint yourself with the process of manually "scanning" (scrolling) through frequencies. You can try this with FM radio or Wi-Fi; the exact frequencies don't matter, it's more so about the process. Zoom in and out. Look at the bandwidth. See if there are side channels. Things like this.

Doing the above will yield itself to letting you discover what sort of bandwidth and frequency decisions C-V2X makes later on. We know that C-V2X will eventually operate around the 5.9 GHz range. This means that when C-V2X does get up and running, you could scroll to 5.9 GHz in gqrx and start looking around.

The next step is to learn how to use uhd. While I have never used it and don't necessarily know if it is the best tool for the job, I do think that using it would give our team the ability to better understand how C-V2X sends data over a radio connection. Avoid worrying about what exactly is inside of a C-V2X packet. Instead focus on trying to understand what modulation parameters and encoding it is using. Take an adversarial approach; think about what features you would need to add to a malicious signal in order to degrade the features you find in a C-V2X signal.

The big picture here is you should practice the techniques, tactics, and procedures you will eventually use to analyze C-V2X broadcasts. Keep in mind that the direction our project is going is to have us "investigate physical-layer attacks on the availability of the C-V2X protocol."

[MaxEllsworth]

Notes from @yixiuzhu and I's meeting with Professor Carruthers on Friday, 19 November.

• For @yixiuzhu and @jasoninirio , in order to practice and develop the necessary skill set, consider generating and listening to a known signal (which you can generate), say “I want to get this bit pattern,” and then look for / at one packet of that known signal. A good sample signal would be an unmodulated FM signal broadcasting a sine wave tone.

• Find out the following regarding C-V2X (@jullianzz and @MichaelJAliberti , consider looking for these statistics as you read papers on C-V2X):

  • The number of bytes in a single packet
  • The symbol time
  • The data rate
  • The modulation scheme

• In regards to infrastructure (more so for @SamKrasnoff and @MaxEllsworth), investigate the following with the lab setup:

  • Whether the antennas for the C-V2X transmitter and the monitor are too close to each other. They should be at least a wavelength apart (approximately 6 cm for 5 GHz).
  • Calibrate TX and RX strength accordingly based on findings from the above.
  • Find out what the path loss is between the transmitter and the receiver.

• Regarding file types @jasoninirio, there is no standard format used for these types of projects. With that being said, here are some file types we've seen so far:

  • .dat : from a library called C-V2X-Virtual. Apparently the file signal_sidelink_cmw500_f5.92e9_s11.52e6_50prb_0offset_1ms.dat, according to Stefan, it is is a saved C-V2X signal
  • .raw : output from a gqrx capture
  • .complex : output from a uhd capture