[Namaste-dev] Re: Tuesday Challenge For June 3, 2008

Paul Williamson kb5mu at amsat.org
Tue Jun 3 23:00:07 PDT 2008


At 6:21 PM -0400 6/3/08, Tom Clark, K3IO wrote:
>The cost of a suitable GPS antenna+RX module would be less than $30.

$30 is a lot to add to the Bill of Materials for a project that's supposed to be affordable and is already doomed to be painfully too expensive for user comfort.

>I would presume that the engine is packaged as a part of the s/c [you mean groundstation] antenna

Would a $30 GPS be able to survive the transmit power at the groundstation antenna?

>We can assume that the s/c geocentric 3D position will be known at a sub-km level simply by making use of Kepler.

Is this assumption good for an Eagle orbit as well as an Intelsat orbit? For the whole orbit? Do we need to broadcast Keps from the spacecraft in order to make open-loop predictions this good? Will government Keps do or will we need command stations to generate better ones?

>I doubt that your one-way ranging scheme will work.

Who said it was one-way? I envision a cooperative ranging scheme where the spacecraft searches out a special probe signal sent by the ground station and transmits its estimate of the probe's time offset from the spacecraft's nominal time. This gives the ground station exactly what it needs to know to pre-correct timing. It doesn't give absolute time or position, but we don't need those things.

Once you have the one measurement from the spacecraft, you can track by simply watching the downlink timing. This works until the combined drift of the spacecraft's clock and your local clock gets too big, at which time you can re-range with the spacecraft. Or, if that turns out to require re-ranging too frequently, the spacecraft can send user-specific timing corrections along with the recently-used-channel information it's already going to send as overhead on the downlink. Then an ordinary uplink transmission is all you need to lock in on timing again.

This does add (software) complexity to both the spacecraft and the ground station, but it doesn't add any new hardware. The turnaround time requirements for the spacecraft's processing can be fairly relaxed, compared to its regular signal-acquisition search (which must be completed at PTT speeds for voice), and the ranging need not be repeated very often per user, so the overall processing load should be modest compared to other loads on the spacecraft signal processing. Or so I guess -- real numbers will have to be run to confirm or refute this.

We also need real numbers about spacecraft DSP loading in order to decide whether we even need to do this at all. The conceptually simple solution is to do no frame timing at all, and let the spacecraft search out each uplink transmission. Only if that turns out to be too burdensome do we need any trickery with timing. We're not talking about TDMA here, just about timing the uplink to ease the acquisition search burden on the spacecraft. Multiple access is still provided by FDMA.

> I'd argue just the opposite,  that in the grand scheme of things, GPS is the OBVIOUS & SIMPLE solution.

You may be completely right about that, but OBVIOUS & SIMPLE is not the opposite of LESS EXPENSIVE. We should investigate the possibility of other appropriate tools in the toolbox before reaching for the GPS hammer, which we know will add cost to each ground station.

73  -Paul
kb5mu at amsat.org


More information about the Namaste-dev mailing list