Merge SEA features

[aromanielloNTIA]

• Dependencies:
  • Updates its_preselector 2.1.0 -> 3.0.0
  • Adds psutil>=5.9.4
  • Adds ray>=2.0.0
• Adds NASCTN SEA data product action, along with a corresponding test YAML config
• Updates Y-Factor calibration:
  • In the action, calculates IIR filter ENBW on action initialization instead of during action runtime, if using IIR filtering
  • Suppresses pointless warnings when configuring sigan
  • Calibration now expects preselector temperature values to be returned in Celsius rather than Fahrenheit.
• Metadata
  • Proactively adds new SigMF extension "ntia-nasctn-sea" with version "v0.2" ahead of future specification of this extension
  • Adds a set_num_channels method to the SigMF builder to support setting the global `core:num_channels" field (note: this is not part of the v0.0.1 SigMF spec. This is part of a larger fragmentation described in NTIA/SigMF lags gnuradio/SigMF #29 )
  • Adds an optional extra_entries parameter to set_capture in the SigMF builder. This is out-of-spec, but used as a placeholder to add metadata to captures before the ntia-nasctn-sea extension is fleshed out.
• Adds a convert_celsius_to_fahrenheit method to signal_processing/unit_conversion.py

[aromanielloNTIA]

A few potentially 'breaking' changes are included in this:

• Metadata for all actions will now include a ntia-nasctn-sea SigMF extension listing
• There are 2 new dependencies and an updated dependency with a major version update
• Actions now require that a preselector has rf_paths in the sensor definition, which could cause a situation where previously recognized preselectors are no longer recognized. This is not an issue for any real-world use case.
• probably most important Preselectors will need to be reconfigured to report temperatures in Celsius instead of Fahrenheit. This is already done for all seadog sensors.

Due mostly to the last bullet point, I think a major version increment makes sense. Open to discourse here.

[jhazentia]

Looks good to me