Re: [110attendees] E911 position accuracy question

[Alexandre PETRESCU <[e-mail]>]

Le 08/03/2021 à 15:24, Robert Moskowitz a écrit :
> Anyone here knowledgeable about E911 altitude accuracy and how  > achieved? > > FAA has decided to copy FCC on altitude accuracy, but 
for a different > use case. > > As I recall, FCC wanted EMTs to go to 
the right floor in a building > to find the emergency. > > FAA is using 
this to find the unmanned aircraft (UA) operator for > whatever reason. 
Is this the UA operator on the 3rd floor balcony or > the 4th? Please do 
not start a privacy discuss on this. See my > drip-privacy draft first. 
 > > Thus this is really the Ground Control Station (GCS) location, not 
 > the UA! > > FAA want 3M but will settle for 15' at 95% of the time. I 
think EASA > also wants similar numbers. > > Our research is this is 
VERY expensive for cheap GCS (GPS altitude > accuracy is 13M @ 95% 
according to gps.gov). Well what if the GCS is > a smartphone; it has 
(supposedly) E911... But how? Is this > information available to general 
apps (UA command&control app)?

I am not sure what you mean by 'GCS' and 'E911'.

I can tell that the availability of accurate localization data in GPS 
chips is not a matter of their price.  Actually along the years they 
struggled a lot to make highest accuracy possible in even the cheapest 
and smallest devices.  In the cases of smartphones, the GNSS (including 
GPS) localization modules are often part of the 4G/5G cellular modems, 
or 'SoC's (system on a chip).  These modules do provide very high 
accuracy, in the range of the meter accuracy, when possible.

That possibility has to do with the visibility to the satellite. That 
has to do, ultimately, with the receiver antenna size and obstacles 
shape and placement between smartphone and satellite.

If I remember correctly, there are two parameters available almost in 
all GPS receivers: HDOP and VDOP (horizontal and vertical 'dillution of 
precision' for 2D and altitude respectively).  These range between 1 and 
6, 1 being the best possible accuracy.  I.e. if a drone has a good 
antenna on its top (not belly) and is flying normally (i.e. not flying 
under a bridge, or in a canyon) there should easily be a comfortable 
HDOP and VDOP of 1 and meter-level accuracy.

At IETF there is this RFC 7459 about Uncertainty and Confidence about 
localization data.

When organisations request a certain level of precision from GNSS data, 
often they might (just might) be influenced by various intervening 
intermediaries that promote their particular enhancements to 
localization technologies.  RTK, EGNOS, WAAS and more come to mind; they 
rely on some base stations situated on the ground.  But often times, the 
best accuracy can be obtained simply with proper antenna size and 
positioning.

Other times,simple  GNSS signal repeaters could be more advantageous 
than RTK, EGNOS, WAAS.  For example a tunnel in the Czech republic 
(Prague!) recently announce plans to equip with such GNSS signal 
repeaters instead of ground stations.  Maybe the same can be planned for 
drones that fly in difficult to reach satellite areas.

In automobile technologies some times organisations require sat-based 
centimeter-level precision for a use-case such as self-parking.   
However, in that range (cm-level, centimeter), other technologies could 
be more appropriate than satellites.

What I see as difficulty with drones and their position is the 
difference in time: a certain lag migh tbe noticed between where the 
drone is now (very accurately) and what a person might see as the 
position on the screen.  That lag could be in the order of several 
seconds.  That could be a problem.  During these seconds many bad things 
can happen.  To give the human time to react (needs one second between 
brain gives order and arm make action) then one needs a very high 
performance (very low latency) communication technology.  For example, 
the bluetooth that is used to stream high quality stereo sound might not 
be the best appropriate.  But yes, it is better than no identification 
at all.

Alex

>  > Please let me know if you have knowledge you can share. > > Bob >