GPS week

from Wikipedia, the free encyclopedia

As GPS week ( english GPS week ) is a continuous weeks counting indicated in the GPS signal, the beginning of the year 1980 for the satellite navigation system Global Positioning System was introduced (GPS) of the United States. Since the GPS signal does not transmit an absolute date, it is used by receiving devices to infer the GPS time based on the GPS time of the week (TOW) .

General

The GPS week starts on Sunday around 00:00 UTC , more precisely at 00:00 GPS time . The exact point in time is shifted compared to UTC, because in contrast to UTC no leap seconds are inserted in GPS time . In addition to the GPS week, the GPS time provides the number of seconds since the start of the GPS week (0 - 604799 seconds).

The consecutive counting of the week number has several reasons and advantages:

  1. a simple, approximate time for various GPS services
  2. the weekly counter lasts for almost 20 years and only needs 10 bits (counting 0 to 1023)
  3. The precise P-code of the GPS satellite signal has a length of 38 weeks (266 days) and can be easily divided into different segments (see P / Y-Code ).
  4. The exact time of a GPS week can be limited to the second dimension (1 week = 604,800 s), which means
  5. a simple relation to atomic time is given (constant difference).

GPS end-of-week rollover

The zero point for the GPS weekly counter is January 6, 1980 00:00 UTC (January 6, 1980 01:00 CET). Since the current GPS navigation message only transmits the 10 least significant bits of the week number, the week counter (Week Number Rollover, WNRO) returns from the binary value 1111111111 (10 bits, decimal 1023) back to 0000000000 (decimal 0). This return is referred to with the English term GPS End-of-Week Rollover . Not only GPS devices are affected by the problem, but all information systems that use GPS time, such as time servers . Usually - especially with current systems - the rollover is taken into account in the calculation by the software ( firmware ) of the device in question, while with older navigation systems there seem to be some deficits.

1999

In August 1999, reaching GPS week 1023 (and jumping to zero at UT midnight Saturday / Sunday) resulted in a situation similar to that expected four months later for the Year 2000 problem . It was seen by many as a test case for this; there were no notable mishaps with GPS receivers.

In the Central European Time Zone ( CEST ), this special “end-of-week rollover” occurred at 01:59:47 on the morning of August 22nd, when the weekly counter of GPS time jumped from 1023 to 0. Such jumps could have confused control systems and industrial plants that use GPS signals as the time standard , since these systems would have been backdated to January 6, 1980.

2019

The last, second in total, jump back from 1023 to 0 took place on April 7, 2019 at 01:59:42 CEST.

On November 3, 2019, this WNRO event also led to a disruption of the ETCS train control system in the ICE-1 fleet and thus to its sudden failure on the Munich-Berlin axis . The error also occurred in various vehicles equipped with ETCS in Switzerland, but only led to an incorrect display of the date and time on the driver's cab display.

2038

After a further 1024 weeks, there will be a return on November 20, 2038. In January of that year, another significant event is pending in information systems, which is referred to as the " year 2038 problem ".

Web links

Individual evidence

  1. ^ A b Richard B. Langley, University of New Brunswick: The GPS End-of-Week Rollover. (PDF 324 KB) November 1998, p. 40 ff. , Accessed on April 4, 2019 (English).
  2. ↑ a b heise online: Week rollover effect: Problems for older GPS devices and time servers expected. Retrieved March 24, 2019 .
  3. ↑ ETCS disruption on ICE 1 for days . In: Eisenbahn-Revue International . No. December 12 , 2019, ISSN  1421-2811 , p. 648 .
  4. ↑ TELOC . In: Association of Swiss Locomotive Drivers and Candidates (Ed.): Loco-Folio . No. 192 , 2019, ZDB -ID 2303252-2 , p. 38 ( PDF ).