4 ms·

I’m currently reading this book about the history of the DSN. https://www.amazon.com/Uplink-Downlink-History-Space-Network-1957-1997/dp/1494740613 https://www.

I’m currently reading this book about the history of the DSN. https://www.amazon.com/Uplink-Downlink-History-Space-Network-1957-1997/dp/1494740613 https://www.amazon.com/Uplink-Downlink-History-Space-Network... Fascinating stuff. I wonder how they will manage the priority of getting that data back over the next two years? Timeshareing on the DSN has been a problem since the early 60’s.

dogma1138 8y ago
I’m not sure they can even get all of the data back. The probe has limited data storage capacity and even more limited bandwidth to send it back to earth. I’m guessing they have a protocol to prioritize data from different instruments as well as prioritize data purging if there is an opportunity to look at something more interesting before everything could be sent back home. It can be a heartbreaking process for some of the teams for sure. My best bet is that they get some initial indication of what data was collected then prioritize based on what they think is the most interesting or valuable from what was collected. They were still retrieving data form the original Pluto flyby as this one was planned I’m not sure if they even got all the data from pluto yet.
- jessriedel 8y ago
  > purging if there is an opportunity to look at something more interesting before everything could be sent back home. This is not a risk as, with high probability, there is nothing else on their trajectory they will be able to look at for the 18-month downlink time. > I’m not sure if they even got all the data from pluto yet. They did. It's true that they have limited data storage space, but they have had plenty of time to download the entire data set from Pluto. https://futurism.com/it-took-15-months-but-all-of-new-horizons-data-has-finally-been-downloaded/ https://futurism.com/it-took-15-months-but-all-of-new-horizo... I believe the major constraint is how much data they can take during the short flyby time, not storage space or downlink time.
  - dogma1138 8y ago
    Well it sorta both yes the flybys at this point are so short that the amount of data they collect is limited but they still collect GB’s of data and the up line data rate will likely drop even further as the probe goes further. It’s still astonishing to me that they can get 1kbp/s with s 1.5m dish antenna pointed at a target 0.29281 arc minutes in size. This is the equivalent of pissing into your toilet form orbit. But heck now the FCC can claim that broadband coverage has been expanded up to the Kuiper Belt.
- tabtab 8y ago
  Re: I'm guessing they have a protocol to prioritize data from different instruments as well as prioritize data purging if there is an opportunity to look at something more interesting before everything could be sent back home. I suspect if a new nearby target were discovered, they may have to send back just thumbnails and summarized/compressed data, review the summaries back on Earth, and only request the more promising full images and data snippets. I just hope they already have such compression/summarizing software onboard because sending significant new software that far can be a bear. I wonder if they planned for such a "rush job" scenario?
  - dogma1138 8y ago
    It’s main CPU is a 12mhz MIPS processor I don’t think they do any complex software analysis on board.
    - tabtab 8y ago
      Summarizing doesn't necessarily require complex algorithms. For example, a 64x64 pixel image can be averaged into a 16x16 image, and/or a highly compressed JPEG. And data streams can maybe average 1-second intervals into 30-second intervals. You can get a little fancier by only averaging intervals that don't vary much, using standard statistical algorithms (including "cheap" approximations to save CPU). A lot of times cameras and instruments will be pointing at empty space because they error on the side of capturing unknown petributions/objects. If you are not looking for faint moons etc., you can mostly ignore those "dark areas", at least in a "data emergency". And remember, the summarized versions are used to narrow down the set of full data sets returned. Thus, if something looks unusual in the summaries, the full set of that segment is requested from the probe. If there is a need to rush, it's a matter of narrowing down which parts of the data you want to send back in full.
      - dogma1138 8y ago
        The instruments are turned off until they are needed they cannot be used when the space craft is spinning which is required for it to transmit data to earth which is what it does most of the time the flyby lasts minutes at the most, the data takes over a year to transmit. Also even if it didn’t need to be spun the instruments would be turned off to save power, the probe has no instruments that could detect any objects without a preplanned flyby.
- thiagotomei 8y ago
  Even if they cannot get all the data back, I suppose they could do a preliminary analysis onboard and decide what to send back. It is, in a very broad sense, the same problem we have at the LHC. There is just too much data taken for the available storage and bandwidth.

dogma1138 8y ago

I’m not sure they can even get all of the data back. The probe has limited data storage capacity and even more limited bandwidth to send it back to earth. I’m guessing they have a protocol to prioritize data from different instruments as well as prioritize data purging if there is an opportunity to look at something more interesting before everything could be sent back home. It can be a heartbreaking process for some of the teams for sure. My best bet is that they get some initial indication of what data was collected then prioritize based on what they think is the most interesting or valuable from what was collected. They were still retrieving data form the original Pluto flyby as this one was planned I’m not sure if they even got all the data from pluto yet.

jessriedel 8y ago
> purging if there is an opportunity to look at something more interesting before everything could be sent back home. This is not a risk as, with high probability, there is nothing else on their trajectory they will be able to look at for the 18-month downlink time. > I’m not sure if they even got all the data from pluto yet. They did. It's true that they have limited data storage space, but they have had plenty of time to download the entire data set from Pluto. https://futurism.com/it-took-15-months-but-all-of-new-horizons-data-has-finally-been-downloaded/ https://futurism.com/it-took-15-months-but-all-of-new-horizo... I believe the major constraint is how much data they can take during the short flyby time, not storage space or downlink time.
- dogma1138 8y ago
  Well it sorta both yes the flybys at this point are so short that the amount of data they collect is limited but they still collect GB’s of data and the up line data rate will likely drop even further as the probe goes further. It’s still astonishing to me that they can get 1kbp/s with s 1.5m dish antenna pointed at a target 0.29281 arc minutes in size. This is the equivalent of pissing into your toilet form orbit. But heck now the FCC can claim that broadband coverage has been expanded up to the Kuiper Belt.
tabtab 8y ago
Re: I'm guessing they have a protocol to prioritize data from different instruments as well as prioritize data purging if there is an opportunity to look at something more interesting before everything could be sent back home. I suspect if a new nearby target were discovered, they may have to send back just thumbnails and summarized/compressed data, review the summaries back on Earth, and only request the more promising full images and data snippets. I just hope they already have such compression/summarizing software onboard because sending significant new software that far can be a bear. I wonder if they planned for such a "rush job" scenario?
- dogma1138 8y ago
  It’s main CPU is a 12mhz MIPS processor I don’t think they do any complex software analysis on board.
  - tabtab 8y ago
    Summarizing doesn't necessarily require complex algorithms. For example, a 64x64 pixel image can be averaged into a 16x16 image, and/or a highly compressed JPEG. And data streams can maybe average 1-second intervals into 30-second intervals. You can get a little fancier by only averaging intervals that don't vary much, using standard statistical algorithms (including "cheap" approximations to save CPU). A lot of times cameras and instruments will be pointing at empty space because they error on the side of capturing unknown petributions/objects. If you are not looking for faint moons etc., you can mostly ignore those "dark areas", at least in a "data emergency". And remember, the summarized versions are used to narrow down the set of full data sets returned. Thus, if something looks unusual in the summaries, the full set of that segment is requested from the probe. If there is a need to rush, it's a matter of narrowing down which parts of the data you want to send back in full.
    - dogma1138 8y ago
      The instruments are turned off until they are needed they cannot be used when the space craft is spinning which is required for it to transmit data to earth which is what it does most of the time the flyby lasts minutes at the most, the data takes over a year to transmit. Also even if it didn’t need to be spun the instruments would be turned off to save power, the probe has no instruments that could detect any objects without a preplanned flyby.
thiagotomei 8y ago
Even if they cannot get all the data back, I suppose they could do a preliminary analysis onboard and decide what to send back. It is, in a very broad sense, the same problem we have at the LHC. There is just too much data taken for the available storage and bandwidth.