RR v3.0 Location Data Recommendation

Status
Not open for further replies.

Dave_D

Member
Joined
Sep 30, 2005
Messages
162
Location
Incline Village, NV
Hi Lindsay,

Could you please verify for us whether or not RR v3.0 will support multiple location points per conventional system? From the Alpha 1 discussion, I'm still unclear and want to encourage it.

A single point/radius simply isn't sufficient for many applications. Take, for example, these counties:

  1. Nye County, NV. Some 245 miles across and, lovingly, "T" shaped. By the time you squeeze this sucker into a circle, you've encroached on three states, to say nothing of the thousands of square miles of non-Nye Nevada territory. If we're building an application-based system (an exclusive system), then this level of inaccuracy (gluttonous inclusiveness) undermines.

    Another important factor: Uniden scanners, for example, accept a maximum radius of 50 miles for any given location. Nye's unipoint would consequently round down, shaving off thousands of square miles; more than half of the county's area. What's more, this one unipoint (how's that for redundant?) wasn't chosen with a 50-mile radius in mind; it's almost outside of the county and the resulting circle incorporates few of Nye's populated areas. As a result, most Nye County residents would be locked out of their own county's frequencies!

  2. El Dorado County, CA. This county fits neatly into a 35-mile radius circle. Unfortunately, Uniden scanners, for example, allow for either a 30-mile or 50-mile radius. Pick one. If the scanner or application rounds the county's location radius up to 50 miles (as it should), then all of El Dorado's frequencies will be unlocked from an extra 15 miles away while other, closer county frequencies are not because they fit more snugly into their respective circles. This can get kinda weird -- nearby sites are locked out while sites much, much further away are unlocked.

    As it turns out, El Dorado is better described by two 20-mile radius circles. Problem solved. And this is typical among larger counties. Geometrically speaking, smaller circles provide better resolution and will therefore scale more accurately between scanners and applications. By contrast, when you force scanners and applications to use larger circles, your margins of error increase accordingly.
These are only two examples which demonstrate some common problems I've encountered while mapping various radio systems. But once you allow for multiple geopoints, these problems all but disappear.

Of course, polygonal boundaries would be even better -- let the applications determine the best geopoints/radii for a given area -- but multiple location points is good enough for now and a quantum leap from unipoints.

I hope you appreciate the position you're in. What you do with location data now could define the state-of-the-art for manufacturers to follow. Think big.

Thank you for hearing me out. I hope this is helpful.

Dave
 

blantonl

Founder and CEO
Staff member
Super Moderator
Joined
Dec 9, 2000
Messages
11,101
Location
San Antonio, Whitefish, New Orleans
Dave,

Here is the architectural approach that we are taking for implementation of the location data.

1. The 3.0 release will provide the ability to geocode with one location attribute set a:

- frequency subcategory
- a trunked system site
- a trunked system talkgroup category

And for inheritance purposes

- a trunked system (for default inheritance)
- a county (for default inheritance)
- an agency (for default inheritance)

So, to tackle your specific issue in the first release, you'll have to most likely define out mutiple entries for different areas of the county keeping within the constraints of how we will store location data - a single lat/long/range point entry.

Now, in a future release, the architectural approach will be to allow to types of data:

1. A lat/long/range entry with the radius range being defined in a circle -or- a square (square range being line from centerpoint to top left corner)
2. A user defined polygon that defines the outline of the entry.

Hope this answers your question and give you a good perspective of where we are going. I've also attached some screen shots of what the current 3.0 range information stuff looks like today (I'm getting ready to push out the next release of the beta soon)
 

Attachments

  • Picture 3.png
    Picture 3.png
    174.8 KB · Views: 168
  • Picture 5.png
    Picture 5.png
    130.8 KB · Views: 150

Dave_D

Member
Joined
Sep 30, 2005
Messages
162
Location
Incline Village, NV
Nice use of Google Maps API!

I think I get it now. I'm looking forward to playing with these new features in the beta. If you don't mind spoiling the suspense, which radius values will RR v3.0 support? I can use this info to begin optimizing my location data. Given the volume, you're likely to finish before I do.

Polygons?! Yes!!! You've made someone very, very happy.

Thanks a million!

Dave
 
Status
Not open for further replies.
Top