Open Research Institute Weekly Reports, 5 Aug 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Uplink Simulator Part 4

Description

The modifications mentioned at the end of Part 3 to @M17 Project VHF/UHF narrowband have been completed. A 16kbps OPUS codec has been integrated. A variety of other changes to parameterize the codebase for increased utility and flexibility for space and terrestrial microwave band use have been made.

A channel impairment block added to the GNU Radio flow-graph. This is an over-the-air demonstration of a transmitter and receiver for Opulent Voice at 905 MHz, with the baseline16kbps OPUS codec integrated.

Flow-graphs and information about the simulators can be found here: https://github.com/phase4ground/documents/tree/master/Engineering/Uplink%20Modem/Simulator

More information about Opulent Voice can be found here:
https://www.openresearch.institute/2022/07/30/opulent-voice-digital-voice-and-data-protocol-update/

Thank you to all who support the work.

A

Here we have another demo of opulent voice, this time going over actual radio frequencies using this coaxial cable here to simulate the the air um I'm transmitting from this laptop running a guinea radio flow graph. It takes input from a file that was created by the opv mod program and it modulates it into fm, because the mod program was designed to work with an fm transmitter.

A

It doesn't have that actual modulation in it, but this generates the modulation and puts it out to the pluto sdr, also some visualizers on the screen after running it through this channel model of degradation. So I have this all turned off this one's yellow because it's bypassed this one is just: has its values set to to no degradation, so we're still dealing with a fairly ideal channel?

A

Here's the pluto connected by usb to the laptop and then the output from it goes into this, which is an rtl sdr low cost receiver plugged into this raspberry pi 400, which is one of the faster, the raspberry pi family, and on this screen you can see what I'm going to do. I'm going to run this pipeline of commands on the raspberry pi. It starts with rtlfm, which is just a receiver that runs with the rtl sdr.

A

This is the channel frequency that I've chosen. That would be a good channel in the band plan for southern california for for 900 megahertz and the sample right here, 187 kilosamples per second, and then this just runs the opv d mod um with no arguments, so it just tries to receive from the air from the standard end and puts audio out on standard out, which I then pipe to a play, which is the audio player connected with alpha sound same sample rate, and this just specifies what the format is.

A

So I will run that okay, so here is some messages from the rtl sdr. This is from uh rtl fm.

A

This is the first message coming from the demodulator: it's completed its initialization sample processing and is waiting to find a signal now over here on the laptop I'm going to start the flow graph and there it's transmitting and now after some acquisition, we have started to receive the signal each one of these punctuation marks here represents a frame if it's a dot, it's a frame that was received just fine.

A

If it's an exclamation point, it means we failed to receive the sync word, but we're able to go ahead anyway and if you ever see a capital x there that means that the frame was actually lost. Let me turn up the volume, so you can hear it. Research.

B

As we all know, these mean particular things. This.

A

Is nearly easy, perfect audio coming out of the receiver.

B

Open source is a type of intellectual property.

A

This is a very minimal debug configuration. I can only see whether the frames are working or not. So let me interrupt that and run a slightly different version of this script that puts out diagnostic information as well.

A

Now, each frame actually not directly tied to frames but about the same rate as a frame is putting out one of these lines of of text. We can see here that it's in has did a carrier. Detect has been successfully received and frequency offsets are really small.

A

It's locked, the clock rate is nearly one which is about right. These three numbers represent timing to the uh to the symbol, and if everything were completely perfect and locked in like it is in simulation world, these would be constant. It would be 666 and it wouldn't be drifting, but you see it is drifting a little bit, but not very fast.

A

That's a very tolerable rate of drift- and this number here mark cost is a metric from the vertebrae decoder and as long as this stays below about 30, it's okay and so we're receiving these frames. Just fine and that's confirmed by the fact that they sound good.

A

We don't have this hooked up into a single unit for a transceiver yet, but we do have receive, transmit and receive going on generated by the the code, the c plus code and it's working over the air. And if we look at this window, this is a h-top which is showing usage.

A

You can see that there are several different processes running rtl fm this top one here is the actual opulent voice code and the total utilization here is a fraction of what's available so even on a raspberry pi we're not using up all the cpu.

A

And if you look more carefully here at the at this chart, you can see a lot of it's going to the rtl fm processing and only about 25 of one of the four cpus is being used by opulent voice processing.

A

So very scalable.

A