Transcript downloading - IETF Tools

Next Generation 9-1-1 Standards, Transitions and
Challenges
Brian Rosen
Senior Director, Neustar
Chair, Long Term Definition WG, NENA
We’re redesigning the 9-1-1 system
• It’s a complete redesign, top to bottom
» All new network, protocols, data structures
• Moving from a telephone network to an IP network
• The project is called “Next Generation 9-1-1”
• The initial technical standards are often called “i3”
• NG9-1-1 is multimedia: voice, video, text
• NG9-1-1 handles new devices & services easily
• NG9-1-1 responds better to overload and
disasters (and deliberate attack)
NENA 08-003 Public Review
• Detailed Functional Element and External
Interface specifications (“Stage 3”) NENA 08-003
is out for public review
• This is the final technical review, and we expect to
publish 08-003 this year
• This document provides a complete technical
standard for the public safety side of the ESInet,
and the service provider interfaces to it
• The technical standards for NG9-1-1 are
colloquially called “i3”
i3 is based on IETF emergency standards
• NENA does NOT define what origination networks
do, but references IETF standards
• Single international standard for emergency
calling
• IETF “framework” and “phonebcp” (still a draft, but
on its way through IETF approval process) defines
what endpoints and networks do
Emergency Services IP Network
• The network that connects all public safety
agencies together
» Not just a 9-1-1 network
» Public Safety is broadly defined
• Managed, private, routed IP network, but not a
walled garden
• Origination networks connect to it, and it’s
connected to the global Internet
• Bottom up county-state-national network design
Some details about location
• New representation for location: PIDF-LO
» RFC4119 updated by RFC5139
» Based on SIP Presence (PIDF)
• XML data structure accommodates both a civic (street
address) or geo (lat/lon)
• Location can be passed by value (the PIDF itself) or by
reference (a URI which when “dereferenced” yields a PIDF
• New “Location Configuration Protocol” for an endpoint to
learn it’s own location
• New SIP Geolocation header to send location with the call
• Dereferencing protocols can be SIP (Presence
subscription) or HTTP (HELD protocol)
Key fundamentals
• All calls (without exception) enter the ESInet as SIP, with
location, routed by ECRF (LoST server)
» Anything that isn’t native SIP is gatewayed to SIP
outside the ESInet or on its border
• Location comes from the ACCESS NETWORK, which may
or may not be the calling network
• Since the ECRF is available from the local 9-1-1 Authority,
and calls are accepted over the Internet, costs to service
providers are minimal
• All service providers are affected, everything is changing
Compared to the existing E9-1-1 system:
• New form of location: “PIDF”
»
»
»
Carries either geo or civic
Handles location in a building
Can be sent as a “value” or a “reference”
• New form of routing database “ECRF”
»
»
»
Inherently point in polygon with geocode from civic
Driven by PSAP map (provisioning of ECRF is directly from their GIS system
Used to route calls, but also used to route to responders
• New form of routing engine “ESRP”
»
»
»
Direct location based routing (uses ECRF)
Can be multilevel (state, county, PSAP)
Policy Routing Function uses PSAP management written rules to control routing for
all normal and abnormal conditions: edit the rule, store the ruleset, and routing
changes immediately
Transition from E9-1-1 to NG9-1-1
• Starts with the current E9-1-1 system
• Ends with all PSAPs and origination networks on
ESInet
» Some origination networks could be on Legacy
Network Gateways
» Some unupgraded PSAPs could be on Legacy
PSAP Gateways
• When Transition is finished, the Selective Router,
and ALI are decommissioned
Legacy Selective Router Gateway
• Interface between an existing Selective Router/ALI
and an ESInet
• Allows calls originated on one side to terminate on
the other
• Allows calls to be transferred between PSAPs on
either side of the LSRG regardless of where the
call originated
• Handles location queries when calls transit the
LSRG
Example Call Flow
Origination network on SR to i3 PSAP
SRDB
2
5
SS7
ECRF
4
Legacy
SR
Gateway
ESRP
6
`
NG9-1-1 PSAPs
`
ALI
3
1
Selective
Router
End Office
Transition begins with ESInet and LSRG
• One PSAP is upgraded to i3
» All origination networks are still on the SR/ALI
» All calls transit the LSRG
• We then can upgrade PSAPs to i3 or move them
to LPGs
• We can upgrade origination networks to i3 or
move them to LNGs
• When the last PSAP and origination networks are
transitioned to the ESInet, the SR and ALI are
decommissioned
Transition Characteristics
• Everybody moves, one way or another
• Moves happen in any order, LSRG allows calls to
be answered regardless of who migrated when
• No fork lift upgrades required – SR/ALI don’t
change at all, LNGs and LPGs allow origination
networks and PSAPs to migrate without system
upgrades, although some data and process
upgrades are needed.
• Of course, if you migrate a PSAP via LPG, you
don’t get many NG9-1-1 benefits
Challenge #1 Education
• PSAP management education
» Lots of things change, but the basics of
answering calls and helping people don’t
» NG9-1-1 is complex, and different, but not hard
» There are very significant benefits to NG9-1-1
» It’s happening, and it’s happening now. Head in
the sand won’t work
• State 9-1-1 Authority education
» Bigger roles for states
» Not optional
Challenge #1 Education (cont)
• Origination Network Education
» It’s happening now
» It’s not optional
» Except for legacy wireline, it costs less for voice
9-1-1
• Regulator Education
» Existing regs often prohibit NG9-1-1 deployment
» New origination networks and the access
network vs origination network issues are
complex
Challenge #2 Speed of Transition
• We can’t take a decade to do transition in an area
• It’s not clear we can take more than a year
• Once transition starts, either the 9-1-1 authority
has to pay for both the old and new systems, or
the current SR/ALI operator has to continue to
operate them with declining revenue and
customers
• Neither is sustainable for several years – we have
to get transition over quickly
Challenge #3 – Existing Origination Nets
• 9-1-1 is an expense with no revenue
• 9-1-1 is regulated
• 9-1-1 was built for voice, and the network support
for 9-1-1 was built for voice. Adding support for
text/video is expensive
» See bullet 1 above
• Most existing wireless origination networks are
also access networks, operators can’t get their
head around providing location for other
origination networks riding on their raw IP packet
service
Challenge #4 Data
• Location infrastructure is all GIS based – everyone
needs a good map
• Lots of new data sources – need implementation
of mechanisms and population of databases
• Lots of new data means new processes and new
user interfaces
• Special challenge: location accuracy will often be
<1m uncertainty within several years
» GIS systems need much more accurate data
» Lines and points WILL NOT BE SUFFICIENT
Challenge #5 Regulation
• The simple part is fixing existing regs that inhibit NG9-1-1
deployment
• Access networks must supply location. Broadband
networks must supply location. Wireless networks that are
broadband networks must supply location
• How do you get a MUCH larger constellation of origination
networks to implement 9-1-1? How much is regulated vs
how much is voluntary?
• How can we get PSAPs and origination networks to move?
Who can compel what?
• What standards for critical infrastructure is required?
• Can we PLEASE have a uniform standard for PBXs (and
note that enterprises are often access network operators)
Challenge #6 Cost
• New networks, new equipment, new data = new
costs
• It’s an opportunity for vendors to increase profit
margins
• It’s also an opportunity for new vendors, business
models and relationships to lower overall costs
Next Up: Dispatch
• New data, new processes and new call sources
mean changes in dispatch systems
• This effort is already started
• Entirely new data interfaces between 9-1-1 and
dispatch (CAD) systems
• Will then affect MDTs (mobile data terminals in
responder vehicles).