7

So the GPS Week Number rollover has been all over the news lately. I don't really know anything about GPS, but my first thought upon reading up on this topic was: why use a week-based format in the first place?

It seems like it would have been much more intuitive to use, say, the number of seconds since the GPS epoch. I realize that there's some degree of arbitrariness to all formats, and that there's usually some historical baggage that plays in, but counting weeks seems like a very unusual choice even then.

Is there some practical or historical reason for this design?

Andrea
  • 173
  • 3
  • "Why" is opion based but that's how it has been designed. http://www.geokniga.org/bookfiles/geokniga-seeber-g-satellite-geodesy-2003.pdf in chapter 7.1.5.3 Computation of Satellite Time and Satellite Coordinates. – user30184 Feb 14 '19 at 11:02
  • 3
    @ahmadhanb, this doesn't seem to be a 'general IT' question, and GPS data / metadata should be clearly a GIS question. – Simbamangu Feb 14 '19 at 11:03
  • Indeed, I chose to ask this question on the GIS Stack Exchange specifically because it concerns the GPS Navigation Message format. I'm a software developer myself, but that doesn't mean I have any insight into the design of this particular technology. – Andrea Feb 14 '19 at 11:13

1 Answers1

5

A satellite's signals (radio waves) must pass through many atmospheric interference (e.g., tropospheric and ionospheric) to reach a GPS receiver on earth, so the messages encoded in the signal must be as compact as possible. One such message is the Navigation Message, transmitted as five 300-bit sub-frames at 50 bits/sec. The Navigation Message needed to pack a lot of information in 1500 bits so every bit counts. Expressing date and time as TOW only requires 19 bits, instead of the usual 32 (e.g., Windows OS) or 64-bit (e.g., Unix OS).

https://en.wikipedia.org/wiki/GPS_signals#Navigation_message

Ralph Tee
  • 1,722
  • 12
  • 17