Post processing GPS data

Hello,

I'm new here with questions regarding post processing of data.

I have a GPS puck style unit BU-252S4 interfaced over USB to my computer. I would like to map the position of survey markers positioned in a heavy wooded area. I notice that the output changes quite a bit, my guess is that this is a multipath issue.

Since the software I wrote to use it is custom, what would be the best way to post process data to eliminate error and get a better fix on the markers?

I am thinking about doing the following:
- Take 20 samples and average
- Take another 40 samples and eliminate those with > 4 arc sec error
With those samples take the difference from the average / 20
and add them to the average. (filter)

The numbers could be changed increasing/decreasing the filter Q.

Does anyone have a better suggestion?

Thanks -Tom

 

On a related question, does anyone have information about
the NMEA sentence $PSRF156?
My unit outputs this now and again. Not sure what or why.

 

Thank you for your reply N-N

If I am correct about multipath being the issue, it makes no sense to use signal strength to weight fixes. So I think average and filter is what I am stuck with. Could even do this over longer time to get fixes with different sat positions.

RE. $PSRF156, why would the GPS send this if I didn't send the '114 to begin with? It just seems to output at random times. Perhaps longer than an hour then...

-Tom