I recently carried out some tests on a UBLOX 8 (M8030 based) to see the variation in hot fix times over a 24 hour period. The project involved an Arduino waking up from sleep every 10 minutes and waiting till the GPS got a new fix, transmitting the fix and fix time information over a LoRa link and then going back to sleep. The GPS hot fix times were collected on a remote receiver which recorded the data on an SD card in CSV format so the data could be graphed. The result for a 24 hour period was;
The overall performance was OK, an average of 11.63 seconds fix time, so a considerable power saving over running the GPS continuously.
Although a lot of the fixes are under 5 seconds (good!) there is clearly a cyclic nature, approximatly once per hour the fix time is 32 seconds, which is very close to the cold fix time.
Is this variation in hot fix time normal, and if not is there anything that can be done to reduce the hot fix time average ?