Open Research Institute Weekly Reports, 12 Oct 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: 20211012 FPGA Standup

Description

Open Research Institute FPGA stand-up meeting for 12 October 2021.

A

A

Welcome to the open research institute, fpga stand-up meeting for the 12th of october 2021, and what we will do today is go through what we did last week, what we're going to be doing next week on fpga, primarily for the indian demo that we're very interested in seeing happen.

A

Whether or not you have any roadblocks or need any resources, and I will hand it off to to mr williamson to be the first speaker.

B

Well, okay, I didn't do anything fpga related this week we did facilitate a demo of the um the dbvs ii downlink beacon at the san bernardino microwave society and that went well, but no actual fpgas were harmed and I'm making it up. Demo.

A

A lot of discussion about them, though yeah that was that went really well, and we did get an offer of mountaintop space in southern california for installing the beacon, which definitely could involve some some fpga work. If we, if we pressed forward- and we had a lot of discussions about the channels and the format and uh getting people um making it as accessible as possible, but also pressing forward with gse and and making sure that that gets uh kind of the spotlight so plenty going on there.

A

We have a another demo coming up in um tomorrow, actually wednesday at the defcon 619 858 group, and that will be a completely different crowd and I'm hoping to get at least as good feedback that we got from from sbms um yeah. So we have some questions about gse versus mpeg and transport stream, and whether or not we can maybe field two channels at the same time, so that people with traditional or restricted receivers can use the beacon with mpeg but also provide a test, a way for them to test and transition to gse.

A

So all right, so hey uh aaron! You have the floor.

C

Hello, um so it feels like I've been playing catch up with andre I've been chasing down the platform he's developing on, so that we have a comment platform that we can do our work on um so from last week I did create an open cpi uh support project.

A

C

The light fury, um it still needs a little bit clean up of the fpga programming scripts, but it uses one pcie lane and it which should be, should be more than enough, and I also tested this. um The fpga itself is on a thunderbolt 3 adapter, which is designed for nvme ssds, but it works for the fpga as well. So I can connect to my laptop and I can send and receive data to and from the fpga.

C

um So I and I, in addition to that, I began implementing the encoder on open as an open cpi, hdl worker. um I did track down how the axi tid config parameters were being set and I believe, there's. I have a path forward now of being able to um consume the encoder inside an hd hdl worker for open cpi uh the next for next week.

C

I I plan on getting the light, fury, support project out the doors that can be used by others and then continue working on implementing the encoder and then I have no roadblocks and I don't need any any additional resources at this time.

A

All right, thank you. So much all right to lock. You have the floor.

D

Oh great, thank you so uh so regarding the authentication or the this work that what we've been doing so so I started studying the m17 protocol uh trying to see if I could implement their grc files or the cpp libraries that which they have so that by understanding it better, I try to make sure that we can do the modifications that we plan, like the adding of the authentication or testing how adorant this modulation scheme of m17 is to any kind of interference pattern.

D

uh With the with the uh experience I gained from the radio resilience competition, uh where I got some pretty good uh insights on how it would work so and also to add in the other security measures we thought of so so I'm there at starting the m17, the libraries and how we could implement that.

D

I think there is some uh confusion about how they use their uh error correction course, in that there, uh the kelloy based error, correction code, or I think there is other error correction code, so I'm just trying to understand how they implemented it.

D

And apart from all of this work, I'm trying to finally see how we could put up a paper for october 30th, the symposium, uh that's where I think I have put across a basic uh skeleton of how the paper would look like like covering the prior art and our suggested strategies and all so. So. If you two guys, I think could see that in the channel that we have think.

D

We all could agree on that and see how we could best uh try to publish that paper or how at least we could try to target for what date it was possible. That's all.

A

Okay, yeah, I think that's doable. If you have a draft or questions or anything like that, then then email is probably best and we're happy to help. However, we can and sure yeah if it, uh the for the through the event on the 30th of october it'd, be entirely appropriate, uh even if it's a kind of a pre-print or preview and then in terms of error, correction coding, there is error, correction coding for for m17.

A

So if there's any confusion or or anything that needs to be asked understood that isn't clear from their protocol, uh then we can certainly help with with that uh either on on slack or via email. And if it's something that we go, oh um well, that's that's interesting! Then we can ask them directly too uh so so yeah. If there any any specific questions, we can start working on them right away.

A

They do. They do have a different error: correction coding scheme for the for the uplink, so the uplink and the downlink in terms of error correction coding are completely separated. You know, so you know because that's the system that we have uh so so don't don't worry too much yeah. If it, if it looks completely different than the downlink, that's on purpose and and you know it it, it's good, that's good news.

A

Let's see all right andre, you have the floor.

B

Hey: hey hello, there. Can you hear me yeah? Yes, I can hear you. Okay,.

E

Cool um yeah, so last week um yeah, I spent getting the my my local setup up and running and I got yeah the good news is it works. The bad news is I. I realized that the engine is swapping thing.

E

I might want to fix that because essentially, if I'm a user- and I you know- I'm dropping in the the box and I'm connecting- I really don't want to care about the end. I just want the thing to work and I think it's not that difficult. It's just annoying as in having to go back and fix, but anyways.

E

So there is that um and I'm probably going to start looking at testing the gs, gs, encoding.

E

Thomas put some links to some reference implementations. I can use a yeah and start looking at other stuff in his um very helpful block diagram. So.

B

A

Yeah I saw that it's all uh very great stuff, um so yeah. If, if any of you all are not on slack, then then please do uh and yeah check those out. It's good things and yeah the the big indian versus little indian thing um well we're not alone issue. So I had nothing but empathy. When I read that yeah- and I agree with your um your instinct- is to well you know whatever: whatever we need to do to make it easier to use and adopt is something that we can do.

A

Then we we, we should all right. Do you need anything or any roadblocks.

E

um No! No, uh not at this point north.

A

Okay, very good, all right so yeah for action items for me over the next week. There's a lot. um If we find out anything, we will find out plenty from tomorrow night from presenting the work and talking about the indian demo.

A

um So we got a set of opinions and some offers of real assistance from san bernardino microwave society, and we will get a different perspective from the defcon group tomorrow night um and then all of that will be write it up and and send it out and see if there's a if there, if there's anything that that actually reaches as far back as the fpga work um yeah, that's, I think that's it from here. Anybody have any final comments before we close.

A

All right very good, I I.

E

A

Okay, thank you yeah. If you any questions, please uh email, slack whatever and we'll see you in the lab and on the internet.

B

A

Right see you soon.