1. Indoor Location Accuracy for Wireless 9-1-1 NARUC 2015 Winter Committee Meeting
02/18/2015

2. Indoor Location Accuracy (ILA) for 9-1-1
Establishing a Baseline for Wireless 9-1-1
Indoor Location Challenge
Evidence of a Problem
Statistics
Live Call Data
Tarrant County, Texas, 2013
Baltimore City, Maryland, 2014 2 2

3. Wireless E9-1-1 Call/Data Flow: A Baseline
1: Person dials 9-1-1
4: E9-1-1 Center gets enhanced location
2: MSC requests routing instructions
5: PSAP queries for enhanced location
3: MSC routes call to nearest PSAP
6: PSAP dispatches assistance
Wireless Carrier
(202) 3
9-1-1
a - Tower
911
Communicator
Mobile
Switching
Center
ALI 1
b - Enhanced 5
LEC
Dispatch to Enhanced Location
PSAP 2 6 4
Voice
Data
MPC
ESRK
ESRK
PDE
CRDB
ALI
CRDB –
Call Routing Data Base
PDE -
Position Determination Entity
MPC –
Mobile Positioning Center
ESRK =
Emergency Services Routing Key
TCS E9-1-1 Center

4. Wireless E9-1-1: Indoor Location Problem
1: Smartphones with GPS
4: Failing to get an enhanced location!
2: Home “cuts the cord”
3: Calls from Indoor Locations
Wireless Carrier
(202)
9-1-1
a - Tower
911
Communicator 1
Mobile
Switching
Center
ALI 3
b – NO Enhanced
LEC
Dispatch to Tower location/verbal info
PSAP 2
Voice
Data 4
MPC
ESRK
PDE
ESRK
CRDB
ALI
CRDB –
Call Routing Data Base
PDE -
Position Determination Entity
MPC –
Mobile Positioning Center
ESRK =
Emergency Services Routing Key
TCS E9-1-1 Center

5. Problem of Indoor Location for Wireless 9-1-1
We “should” have a problem 5 5

6. More calls should be coming from indoors
40% of US population has “cut the cord”
2013 CDC study (37% of adults; 45% of children)
70% of calls come from wireless
2012 King County, WA statistic 6 6

7. Problem of Indoor Location for Wireless 9-1-1
We “should” have a problem
40% of US population has “cut the cord”
70% of calls come from wireless
We “see” a problem 7 7

8. Real-World Data Can Guide Our Testing
Actual 911 calls
Tarrant County
All carriers
August, 2013
Random?
Uniformly distributed?
Which are Indoors?
Which are Outdoors?
Color-code X/Y locations based upon Horizontal Uncertainty:
Brown = Phase I only
Green = meets stricter requirement.
Red = misses looser requirement.
Yellow = between strict/loose

9. More “bad” location fixes – due to indoors?
/2007
2013
7.5% exceeded Phase II upper bound (red)
2007
3.3% exceeded Phase II upper bound (red)
3.3%  7.5% (more calls from indoor locations?)

10. “Bad” location fixes – probably indoors…
Tarrant County, TX
9-1-1 Calls – August, 2013

11. Data Trends Reveal Indoor Challenges
Problem area seen in 2011
2007
2011
Nonexistent in 2007
Major problem area in 2011
2012
2013
Greatly improved in 2013
Improved in 2012
Goodrich Warehouse Built in 2007

12. Problem of Indoor Location for Wireless 9-1-1
We “should” have a problem
40% of US population has “cut the cord”
70% of calls come from wireless
We “see” a problem
Tarrant County, TX (Aug, 2013)
“Bad” HUNC values increased: 2007=3.3%; 2013=7.5%
Baltimore calls (Nov, 2014)
“Bad” Horizontal Uncertainty (HUNC)
carol
HUNC is a distance/range calculated by the Location Engine
Determines the range of location “error” based on Confidence value
Confidence (90% here) expresses likelihood to find device within range 12 12

13. Problem of Indoor Location for Wireless 9-1-1
We “should” have a problem
40% of US population has “cut the cord”
70% of calls come from wireless
We “see” a problem
Tarrant County, TX (Aug, 2013)
“Bad” HUNC values increased: 2007=3.3%; 2013=7.5%
Baltimore calls (Nov, 2014)
“Bad” HUNC values were significant
So how do we solve it?
carol 13 13

14. Thank you! Timothy James Lorello Senior Vice President, TCS
275 West Street
Annapolis, MD 14