Modeling Incident Scenarios on the PS
Broadband Network
March 26, 2014
John Facella, P.E., C.Eng.
Senior VP, RCC Consultants
1
Agenda
• User Needs Assessments
• Capacity is Different from Coverage
• Summary
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
2
User Needs Assessments - not Easy
• Law, Fire EMS use data differently
• For decades Law has been a consistent user of data: NICIC
lookups, field records, CAD silent dispatch, etc.
• Fire & EMS have not used data much in the past, but that is
changing (e.g. EPCRs)
• NPSTC and PSAC have put together a 30 page list of applications
that might be used by each service
• Many PS users have difficulty in imagining future uses of data as
opposed to current
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
3
Identifying Future Uses of Data
can be Hard…
“Sonny, how can I tell
you how often I’ll use
the new
superhighway? I don’t
know what a
superhighway is!”
Photo: www.ar15.com
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
4
Some Example Data Uses by Service
• Law
• Fire
– Real time video
from commercial
bldgs
– Arrest process
check list
– AFIS Fingerprint
scanner
– Airborne platform
video sharing
– Officer cam
– ALPR
– Personnel
accountability
– Vehicle
extrication info
based on license
plate
– GIS info for
wildland fires
– Wildland
personnel
tracking
– In-building
personnel
tracking
• EMS
– ePCR from CAD
to EMT to
hospital
– Geo based illness
or pandemic info
– 2 way video
hospital –scene
– 12 lead + O2 sat
+ EtCO2
– ‘Just in Time’
video refresher
enroute
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
5
Dimensioning the App Usage Profile
• Numbers of responders, locations
• Numbers of mobile, portable, and vehicular modem wireless devices
• Current apps in use, by responder type & incident type
• Expected future apps, by responder type & incident type
• GIS geo-overlay of incident history
• Multiple other factors
• Create probabilistic data demand by location, time of day, etc.
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
6
Needs Assessment:
On-Line or Desktop Application
2. Add Wireless Data Apps
1. Begin Survey
3. Add Usage
Patterns,
Number of Users by
Department
Expected Data Rates
and Latencies
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
7
Agenda
• User Needs Assessments
• Capacity is Different from Coverage
• Summary
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
8
Capacity vs. Coverage
• For FirstNet to succeed, it has to provide PS what cellular carriers
cannot do today: improvements in coverage, capacity, reliability,
security, & local control
• Most of the industry discussion has been about coverage
• Capacity is equally important, even in rural areas
– Examples: 9/11 Shanksville PA; Horrific PA Amish school house massacre
– Although rural areas may have less responders; as a result incident commanders
may therefore be more dependent on multiple video feeds
• Designing for coverage
Designing for Capacity
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
9
Issues Impacting “Delivered Capacity”
• “Delivered Capacity” is bandwidth that a PS user has available at
her location. It is affected by:
– User location to cell center (LTE ‘gear shifts’ data speed between 64QAM, 16
QAM, and QPSK)
– User location to adjacent cell (interference)
– Antenna sectorization
– MIMO use
– User’s prioritization
– User’s app bandwidth and prioritization
– App’s FEC and jitter, latency, packet loss sensitivities
– Nature of user’s traffic, and adjacent users’ traffic: bursty, variable rate,
continuous, random (stochastic)
– Adjacent users in same sector and their apps/priorities
– Total cell site backhaul capacity vs. total bandwidth demands from all sectors of
that cell
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
10
Issues Impacting “Delivered Capacity”
• Sidebar - Operational Implications:
– Delivered capacity to a particular user will change over time depending
on the changes to the factors in the previous slide
• Completely different experience from voice trunking
• For example this will result in inexplicable changes to video quality during an incident
– Depending on the incident’s needs, It may be necessary to make onthe-fly changes to user and application priorities in real time
Photo: monkeymucker.blogspot.com
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
11
Issues Impacting “Delivered Capacity”
• LTE assigns Resource Blocks to users, and reserves some for
channel signaling
• LTE systems reuse the same frequency blocks at every cell site, so
adjacent cell sites can interfere with each other and reduce capacity
• How much capacity is needed for a single UL video feed?
– 0.2 Mbps to 6.1 Mbps depending on codec, frame rate, subject movement
speed, lighting, video usage (situational awareness vs. evidential)
• But capacity (in megabits/second) falls off rapidly away from the
center of the cell
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
12
Data Capacity vs. Distance to Cell Site
12 Video Feeds
(1 Mbps each)
At 2 KM
(1.3 mi) away
Uplink Mbps vs. Distance to Cell
Antenna
LTE, 10 MHz x 10 MHz
35
30
25
Data speed
Mbps
20
Only One
Video Feed
At 10 KM
(6.5 mi) away
15
10
5
0
0
1
1
2
2
4
3
6
4
8
5
10
6
12
7
KM
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
• From Tait presentation at
2013 Radio Club of America
Tech Symposium
13
Issues Impacting “Delivered Capacity”
• Multiple authors have shown that incidents may require large
amounts of bandwidth – up to 6.3 Mbps UL
– Andy Seybold’s Cornerstone project in 2011*:
•
•
Barricaded Hostage: peak 6.3 Mbps UL; peak 19.2 Mbps DL
Bomb incident: peak 4.4 Mbps UL; peak 13.5 Mbps DL
– Tait** Traffic Accident Scenario: 4.8 Mbps UL
• Designing a system that delivers at least 6.3 Mbps will require more
sites than just providing coverage with minimal capacity
– However as sites get closer together, interference from nearby cells begins to
become an increasing factor
– Bringing in deployable cell sites into an incident creates additional variables
– LTE networks in future will be self-optimizing
* www.andyseybold.com
** 2013 Radio Club of America
March 26, 2014, “Modeling and Simulation of PS LTE Networks” Tech Symposium
14
Example Sectorized LTE Coverage & Traffic at Aggregation Site
COVERAGE REQUIREMENTS
• Coverage to a PC-Mounted USB device or
handheld device for 95% area coverage
with 95% reliability with the following inbuilding requirements:
– 6 dB statewide,
– 10 dB in the cities of Milford and Georgetown,
and
– 18 dB in the cities of Dover, Wilmington,
Newark and Rehoboth.
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
15
Example Statewide Coverage and Datarate Analysis
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
16
Agenda
• User Needs Assessments
• Capacity is Different from Coverage
• Summary
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
17
Summary
• User data needs not easy to survey
– Future data needs even harder
• PS LTE systems must deliver both coverage and
capacity
– Designing for capacity is different
– Software tools are available to do this
• A single incident may need peak UL speeds of 6.3 Mbps
• “Delivered Capacity” to a user dependent on multiple
factors
– Data speeds to a user will vary over time
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
18
THANK YOU!
QUESTIONS?
jfacella@rcc.com
March 26, 2014, “Modeling and Simulation of PS LTE Networks”
19