Technical: Beacons for DVB-S2 Satellite Transport Layer

Discussion in 'Satellite and Space Communications' started by W5NYV, Jun 2, 2021.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: MessiPaoloni-1
ad: Left-3
ad: HRDLLC-2
ad: L-Geochron
ad: L-MFJ
ad: abrind-2
  1. W5NYV

    W5NYV Ham Member QRZ Page

    Here's a progress update on our multimedia beacon project, going on the air at 10 GHz and providing advanced digital broadband microwave opportunities for people interested in learning more about DVB-S2/X transmit and receive.



    Codenamed BeacRon, in honor of Ron Economos, who advises the work, the project transmits a signal that bridges the gap between "nothing" and the satellite downlink that we've been hard at work to deliver.

    This beacon relies upon GNU Radio, a relatively inexpensive computer, a relatively inexpensive SDR, and some relatively expensive RF hardware from DB6NT.

    This update is mainly about the transport stream, which has a lot to do with the formatting of how the data is delivered over the air.

    The traditional DATV approach is to use MPEG as the transport stream. This is the traditional video stream, like in broadcast television. We specify GSE, or Generic Stream Encapsulation. This allows any packetized data. It reduces the overhead and allows for a lot more freedom. There are so many really fun things that we can do as hams here.

    But we have to get from zero, to fun and accessible digital downlinks somehow. This beacon project is one of those steps along the way.

    We start out with MPEG as the transport stream and show the transition from constant coding and modulation flowgraph to attempting a variable coding and modulation flowgraph.

    Want to follow along or get involved?

    https://www.openresearch.institute/getting-started/
     
    WE4B and WD9EWK like this.
  2. W6RZ

    W6RZ Premium Subscriber QRZ Page

    A couple of videos I made of DVB-S2 running at negative signal to noise ratios (actual negative S/N, not contrived like FT8). This is accomplished with heavy Forward Error Correction (FEC). In this case, 1/4 rate LDPC (Low Density Parity Check). At 1/4 rate, there are three error correction bits sent for each useful data bit. The useful bitrate for this setup was 4.78577 Mbps. The signal is 12 MHz wide.

    The first video shows the QPSK constellation as the transmit power is reduced. Watch the C/N reported by the hardware decoder in the lower left hand corner and compare it to the constellation plot.

    http://www.w6rz.net/dvbs2iq.mp4

    In this video, it's the same setup except I'm decoding live video. Again, the power is reduced to a negative signal to noise ratio. The video is error free at about -2.3 dB S/N. Just to show it's real, I reduce the power another 1/4 dB and the video starts breaking up. Then I bring the power back up a 1/4 dB, and it recovers.

    http://www.w6rz.net/dvbs2.mp4

    With these low signal to noise ratios being attainable, the ground station for a future P4 satellite can use smaller dish antennas.
     
    WE4B likes this.
  3. W5SAT

    W5SAT Ham Member QRZ Page

    Hi Michelle, Thanks for sharing. I am a bit confused what you are demonstrating here.

    It LOOKS like you are simply modulating a 1.2 GHz DVB-S2 carrier being fed by a transport stream carrying audio and video PIDs and using a standard L-Band input tuner to receive it.

    I am confused on the beacon application - could you elaborate on the relationship between what was seen in the video and how that relates to a beacon, please?

    Are you actually encapsulating GSE packets in PIDs or are you transmitting a standard video transport stream with PAT, PMT, Video and Audio PIDs?

    Finally, why Variable coding?
     
  4. W6RZ

    W6RZ Premium Subscriber QRZ Page

    These are just the beginning steps in getting the beacon package working. For example, the initial efforts have explored how much processing power is required (a Raspberry Pi 4 did not have enough CPU) and what transmit capable SDRs will be adequate, yet not too costly. Also, getting receivers working is sometimes not trivial.

    The next steps will be to get the beacon on 10 GHz. At 3:47 in the video, you see some of that equipment (an up-converter and power amplifier). Appropriate power levels and signal cleanliness will need to be measured. A candidate down-converter will also need to be tested.

    For now, the beacon is transmitting video and audio in a standard MPEG-2 Transport Stream. That's because there are not many DVB-GSE receivers available on the market. Pretty much only one at "prosumer" pricing. Most likely the beacon will transmit MPEG-2 TS when it first goes on the air since this would be the easiest for other potential users/testers out there in radio land. Also, video is an excellent test format. You can see with your eyeballs if it's working correctly.

    When the beacon gets some on-air experience under its belt and some users are ready, we'll switch to DVB-GSE.

    Variable coding has to do with accommodating different size ground stations. Smaller less capable stations use a lower S/N modulation, while larger stations can use a higher S/N (and higher bit-rate) modulation. For example, if you have a 3-meter dish, you could receive high quality natural sounding digital audio with the higher bit-rate.
     
  5. W5SAT

    W5SAT Ham Member QRZ Page

    Thanks for the feedback. I'm fuzzy on the beacon's purpose. I know what a beacon is used for on a satellite but don't understand the need to have a modulated beacon carrying data/audio/video. It becomes a fixed carrier at that point and not what I would view as a traditional beacon which usually a CW with a fairly good C/N and minimul bandwidth so it can be found by an earth station. Having a fully modulated beacon that consumes more of the PA's available power must serve some purpose I am not understanding perhaps?

    This beacon would not incorporate variable coding would it? I am well versed in PSK modulation and FEC and understand satllite link budgets but not getting my head around the need for VCM for a beacon. As you well know, VCM is used to accommodate dynamic link conditions that may occurr dueing a communicaitons session. I see the value of VCM in a 2 way communications session but not in a beacon. Perhaps I am conflating the beacon and VCM?
     
  6. W6RZ

    W6RZ Premium Subscriber QRZ Page

    I think you're getting confused by the word "beacon". In this context, it's a terrestrial beacon as defined in Part 97.203.

    https://www.law.cornell.edu/cfr/text/47/97.203

    It hasn't been decided yet, but a frequency just below the satellite band at 10450 to 10500 MHz may be appropriate. This fits in with the ARRL band plan where modes >= 1 MHz are allocated 10375 to 10450 MHz.

    The purpose of the beacon is to enable satellite operators to build and test their 10 GHz station capabilities in advance of a potential P4 satellite. The format of the signal will have much of the same functionality as the proposed P4 downlink.

    We also need to advance the capabilities of current (and less expensive) receivers. As I've already noted, DVB-GSE is not well supported. VCM is also a problem.

    As for VCM, it's a little different in DVB-S2. DVB-S2 transmission is made up of frames. Each frame can have its own code rate and modulation. Each frame has an identifier called the Input Stream Identifier or ISI. A receiver uses the ISI to only demodulate the intended stream (and corresponding code rate and modulation). It's really more of a multiplexing scheme. Here's the flow graph from the video. It's setup to send three ISI's, a QPSK at 4/5 rate stream, an 8PSK at 5/6 rate stream and a 16APSK at 8/9 rate stream. A different Transport Stream video is sent on each stream.

    You're thinking more about ACM or Adaptive Coding and Modulation. That's also possible in DVB-S2, but as you said, it requires two-way communication and won't be available on the beacon.

    [​IMG]
     
    WE4B and W5SAT like this.
  7. W6RZ

    W6RZ Premium Subscriber QRZ Page

    Another purpose of the DVB-S2 beacon is to populate the 10 GHz ham band. 10.0 to 10.5 GHz will be the next target for 5G.
     
    WE4B and K6CLS like this.
  8. W5SAT

    W5SAT Ham Member QRZ Page

    Thanks, that explains a lot.
     
  9. K3RLD

    K3RLD Ham Member QRZ Page

    huh what? Glad there are people out there that understand this stuff... it's like reading swedish to me. :)
     
    KB1PVH and K4BAD like this.
  10. KK4YWN

    KK4YWN Ham Member QRZ Page

    qo-100 simulators would be a nice way to burn up some spectrum. if i had any idea how to build one i would. i have access to a mountain top or two. just mentioning it in case anyone with talent finds themself without a project ;)
     

Share This Page

ad: w5yi