[Ground-station] Question for ORI:

Thu Apr 1 16:21:16 PDT 2021

Seconded....I was gonna write long winded stuff, but Howie summed up what I was thinking pretty well. (that won't keep me from writing more though).  What follows is a bit off topic, so feel free to skip, though IMHO is the kind of area that excites me for the hobby and is not unrelated to ideas for building our own capabilities akin to TDRSS/Iridium/Globalstar for low rate, reliable 'pings' to other Amateur spacecraft.

I've been fascinated by crosslink mission concepts for a long time now.  If GEO goals ever pan out, I would love to see an 'Amateur TDRSS' concept worked into the mission objectives (crosslink sub-sub bands worked into existing Amateur Satellite Service sub bands?).  Something APRS-like in terms of interoperation (not the underlying protocols) for things like LEO-LEO crosslinks would be exciting as well (an 'amateur Iridium' or 'Amateur Globalstar' type concept).  Would love the 'first to hear' award for some future spacecraft to go to someone that heard it via a crosslink two seconds after power up.  As far as protocols involved, not sure I have enough experience to vote one way or the other on heavy FEC narrowband vs SS for the general goal of 'reliable, low bit rate, dirt cheap, and easy pings.' (isn't this a debate that has lasted through generations of engineers?)  I do think whoever fields such a crosslink system (whatever the flight regime) would have to either agree on one or the other, or agree on something like SDR technology that could 'do both' to support a heterogeneous mix of devices.

Some more long-winded stuff about fully open source / open hardware software radio satellite transponders (feel free to skip, these are just ideas....and ideas are a dime a dozen), not completely unrelated to a crosslink style mission:

I think there is a 'technology gap' that exists.  Way back when AMSAT flew an SDX with some pretty cool features (15 years ago or there about right?).  Jump ahead to today and for a few hundred thousand bucks, you can get highly proprietary (though highly reliable) space rated SDRs (AstroSDR, Tethers Unlimited, I think GOMSpace has one that is cheaper, not sure there, etc.)........or you can harden cheaper terrestrial SDRs (AMSAT GOLF and E310) or you can 'roll your own' (Hawkeye360).  All those designs are based on Xilinx FPGAs (Xynqs with embedded ARMs) married to AD93XX ADC/DACs or similar.....but can do things like 54 MHz instantaneous bandwidths, or full DVB-S2X modulation/demodulation, etc....

I'm not knocking any of the high-end stuff.....but where is the 100 kHz to say maybe 1 MHz wide SDR based transponder?  Where is the recreation of the same type of thing that was done on ARISSat-1 (I think that's where the first SDX flew right?), but on a 1U cubesat with modern technology....maybe a stretch goal of making it reconfigurable in orbit.  I'm not talking highly sophisticated anything here, I'm talking the incremental step from an analog transponder to something that can do some amount of onboard processing that is interesting......like switching between narrowband FEC or SS (or doing both simultaneously) to support a crosslink type mission, handling doppler, or FM up / SSB down, or simple FFT based power monitoring and normalization from Uplink to downlink, or notching out uplink alligators on the downlink (re-implement G6LVB's STELLA power equalizer), or ranging experiments, etc.  ARISSat-1 I think was ahead of its time a bit, and just recreating that SDX capability with modern technology would be super cool and worth while to pursue IMHO (fully open source, open hardware as always).

Last collection of random thoughts that are hopefully relevant (mostly about software radio transponders):

1.       To bruces point about capable microcontrollers........I've wanted to build a 'simple SDX' for a long time (but lack the time, full range of skillsets required, and realize I'm probably being "blissfully ignorant" of the challenges involved....so mostly toyed with the idea when I'm bored).  Somewhere in the last year or two there was QEX article about hooking a DAC up to microcontroller that reignited this interest....and then I discovered the STM32F4 family (ARM Cortex-M4s, 180 MHz speeds, etc.):  https://www.st.com/en/microcontrollers-microprocessors/stm32f4-series.html ....way more capable than whatever flew on the SDX (which I think was like 1 MHz speed or something like that right?), and I would think fully capable of keeping up with say a 1 MHz complex sampling rate.

2.       Liquid-dsp:  https://github.com/jgaeddert/liquid-dsp.  Written in C, and from the github description:  "liquid-dsp is a free and open-source digital signal processing (DSP) library designed specifically for software-defined radios on embedded platforms. The aim is to provide a lightweight DSP library that does not rely on a myriad of external dependencies or proprietary and otherwise cumbersome frameworks."  Full disclosure, this was created by one of my VT colleagues, when he was a PhD student, and he's now back with VT as a researcher at Hume, so I might be biased......super smart dude and he still maintains the repo (also by the way, he got the JPEG compression working for us on the two VT FOX cameras, also STM32 based, when he was a Postdoc....the VT watermark in the images was his handywork).  In some short discussions with him on this topic, he seems to think getting liquid working with an STM32F4 might be super easy (though again, he's a genius in my opinion....so 'easy' with a grain of salt).

3.       The STM32F4 family also has an 'audio processing' flavor to it with built in DACs and ADCs for that sort of thing, and I think some kind of coprocessor built in (maybe for something like FFTs, been a while since I looked into this) so I'm wondering if some of that can be leveraged (akin to the funcube showing up as a 'sound card' and IQ samples pumped over stereo L/R channels).....not sure if this really makes sense though, but maybe.....if not, then external DACs/ADCs.

4.       Combine 1-3 and I feel like that's a pretty solid core device for some simple, cheap SDR transponder experimentation.

5.       I feel like there are also some examples out there for useful technology, like DIY softrock SDRs or something like that.....but not for space based transponders.

6.       Suitable up/down converters with traditional good ole analog RF engineering would be needed (maybe that's obvious).

Making a fully open hardware / open source transponder, even if it lacks the super high end capabilities of existing commercial stuff I think would be super cool and worthwhile to pursue.  Keeping the signal processing in the C/C++ domain (as opposed to say FPGA wizardry...not saying we shouldn't also pursue FPGA based designs in general, just not the design I'm currently running my mouth about) might encourage more development support and volunteers since it would be a bit more accessible to folks.  Experiments in novel crosslink techniques (including modulation schemes), spectrum monitoring, illegal transmitter hunting (talk about policing the bands...), etc. could all be 'on the table' for the mission concepts........and when not trying that stuff......turn it over for normal transponder use with novel modes......and I haven't even gotten into low cost digital phased array concepts (for spacecraft) I think would be cool to pursue as well.

-Zach, KJ4QLP

--
Research Associate
Aerospace & Ocean Systems Lab
Ted & Karyn Hume Center for National Security & Technology
Virginia Polytechnic Institute & State University
Work Phone: 540-231-4174
Cell Phone: 540-808-6305

From: Ground-Station <ground-station-bounces at lists.openresearch.institute> On Behalf Of Howie DeFelice via Ground-Station
Sent: Wednesday, March 31, 2021 7:40 PM
To: Bruce Perens <bruce at perens.com>; Douglas Quagliana <dquagliana at gmail.com>
Cc: Michelle Thompson via Ground-Station <ground-station at lists.openresearch.institute>
Subject: Re: [Ground-station] Question for ORI:

I'm not a patent expert but it appears the first one listed covers the generation of spread spectrum "chirps" using a fractional N synthesizer. The other patent covers the  data formatting and modulation scheme. Assuming we have to stay away from those two aspects, we could still use chirp spread spectrum, just not generated the same way. The biggest advantage to chirped spread spectrum from a satellite operations perspective is that it's inherently resistant to doppler issues. As long as the signal is in the receiver bandwidth and you can detect the direction of the chirp you can decode the signal.  Using a chirp spread spectrum physical layer into a an adapted 802.16 mesh network configuration could provide a way to have satellite augmented ground networks (or vice versa) without having to have a planned constellation of satellites. If every new LEO carried the transponder, the network would automatically form and grow.  When satellites are visible to each other traffic would also be repeated satellite to satellite. If the frequency plan was compatible to QO-100 transatlantic relays could be possible into the QO-100 footprint.

Howie AB2S

From: Bruce Perens via Ground-Station<mailto:ground-station at lists.openresearch.institute>
Sent: Wednesday, March 31, 2021 7:13 PM
To: Douglas Quagliana<mailto:dquagliana at gmail.com>
Cc: Michelle Thompson via Ground-Station<mailto:ground-station at lists.openresearch.institute>
Subject: Re: [Ground-station] Question for ORI:

On Wed, Mar 31, 2021 at 3:43 PM Douglas Quagliana <dquagliana at gmail.com<mailto:dquagliana at gmail.com>> wrote:
Bruce writes:
> someone more skilled than me would be sitting down to make an open data link implementation built on some cheap microprocessor

If I understand what you're saying, Phil has already written several downlink schemes.

Yes, but I am not aware of Phil addressing this particular application, which is dirt-cheap SS data links between so-far-terrestrial embedded microprocessors with a link budget to go miles at the lowest data rate and long life on small batteries. I hold out some hope that the functionality of their chip can be duplicated with a relatively small number of discrete components and a cheap microprocessor. Certainly we have ones that can do significant DSP in the $4 range these days, and they idle at microamps drain.

    Thanks

    Bruce

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openresearch.institute/pipermail/ground-station-openresearch.institute/attachments/20210401/274909ea/attachment-0001.html>