Transcript Presentation of Tom Kershaw, VP Verisign

Four Related Acronyms:
ENUM, SIP, DMX and 2Pac
How They All Come Together….
Tom Kershaw
Vice President, VoIP
VeriSign
Yet Another Take On ENUM…(aka Agenda)
+ ENUM Defined for the 34,567th time
+ ENUM the Theory
+ “Peering,” Traffic Sharing, Killing the PSTN etc
▪
The SIP Way
▪ The ENUM Way
+ The IETF View v. The Carrier View v. the End User View
+ Does QoS Matter?
+ Does Network Interoperability Matter?
+ Why Do Gangsta Rappers Love ENUM?
2
What Is ENUM?
+ ENUM is a protocol
+
Born in the IETF
+
Simple Concept: Use DNS to resolve addresses for VoIP
+
Approved, Done, and Nothing Controversial
+ If We Just Talked About ENUM the Protocol, this would be a
very short conference
+ ENUM is a Political Movement
+
Ownership of Addresses
+
National Sovereignty
+
Tom Cruise and Katie Holmes
+
Global Disarmament
+
Etc.
+ There is a strong need to separate the protocol/implementation
issues from the public policy issues
3
Some Initial Comments on ENUM
+ Private (Carrier) ENUM v. Public (User) ENUM
+
Debates, Controversy, Confusion
+ The Key Points:
+
Carrier and User ENUM are different and should have different
structures
+
Carrier and User ENUM are consistent and can co-exist
peacefully
+ There is no clear agreement on what ENUM is for:
+
The wonderful world of the Internet
+
The wonderful world of the PSTN
+
The alleged convergence of these two things
+
OR….Something totally different
+ SIP Peering is 100x bigger than ENUM right now…
+ ….(which is like Luxembourg calling Lichtenstein small)
4
Current State of ENUM
+ Public ENUM trials and “production” environments
+
Austria, Australia, Korea are leading
+
Volume is very small
+
Driven by the Internet Community
+
Dependent on users actually caring
+ Public ENUM Regulatory Bodies
+
U.S., Japan
+
Driven by the PTTs
+
User involvement is little to none
+ Private ENUM efforts
+
Cable
+
Mobile Operators
5
Drivers for ENUM
+ The Driver Matters – Results are Different
+ Internet Community Driven
+ PTT Driven
+ Mobile Content Driven
+ Our View: What is the Goal of ENUM – To Drive IP-to-IP
Communications that goes beyond traditional voice
+ People assume that VoIP operators and users are driving ENUM –
but they are not
6
Who Cares About ENUM?
What are you
talking about?
I don’t care
about ENUM!
7
I love
ENUM! I
have all
of his
CDs!
VoIP and ENUM
+
+
+
+
+
+
+
+
+
+
+
+
8
ENUM is not relevant to VoIP yet
Volumes are too small
Japan Case
10 million VoIP endpoints
10% x 10% = 5% of calls are IP to IP
Benefits of the query with a 5% resolution rate is questionable
ENUM matters only when you can drive res rates above 25%
Enterprise Verticals
Communities of Interest
Peered Private-Public ENUM structures
Cable Efforts are Likely to be Key Drivers
IE – we have to drive volume and drive resolution rates up
collectively rather than pursuing our own private interests
+ Conclusion: VoIP operators and users do
not care about ENUM at present
+ But there is someone who does care
about ENUM…..
9
Who Cares About ENUM?
I Love ENUM!
ENUM is great!
It makes me
money.
10
ENUM and Mobile Content
+ ENUM is currently driven by mobile content
+ 50 Cent makes more money off of ENUM than all the VoIP operators
combined
+ When a user downloads a ringtone, it is sent to the destination
MMSC using SMPP
+ SMPP requires a mailto: address
+ ENUM is used to discover the mailto: address of the destination
+ This application leads to some perverse results
11
+
how to you map the phone number to the correct mailto:
+
what if the number is ported?
+
what is the number is issued under an MVNO?
Business/Regulatory State of the “Roots”
+ Tier 0:
▪ Only one database controlled by RIPE NCC and ITU (policy
only)
▪ Contains participating country codes.
▪ Delegation would be at the NPA level for the US
+ Tier I:
▪ Several valid country specific public trials – Austria is leading
▪ U.S. has decided to issue a tender for CC1, split into to
administrative domains
▪ Lot’s of Boring Trials Going on Now
+ Tier II:
▪ A Few Interesting Trials Underway
▪ Every Carrier and Cooperative will have a Root
▪ VoIP Tier IIs brag about 500K users; Mobile will be in the 50
Millions soon
There are actually many ways to do this….
12
Current Issues With ENUM
+ Very few VoIP platforms support ENUM today
+ Nobody has figured out how to make money from ENUM yet
+ Nothing in ENUM you can’t do with SIP
+ Huge political issues over data ownership
▪
Two camps:
▪ I want the data and I want everything to be free
+ ENUM solves only a small part of the problem
▪
Where you are is easy – how to get to you in a secure, reliable matter is
another issue
+ Mobile Content application is creating a critical mass in ENUM that
is not necessarily consistent with the VoIP application
13
ENUM: Missing Pieces
+ I Know the Destination Domain of the Called Party
+
I Can Now Query the Destination to Find the IP Address
+
+
+
+
+
But:
What QoS Rules are Associated with the Destination
What Protocol/Variations are Available at the Destination
What Network Path to Take
What Security Policies/Keys Are Needed
+ ENUM provides the information, but assumes the network will be
able to figure it out.
+ Reality: It Won’t (at least not yet)
14
Private Peering: Real World Example
IN NAPTR 10 10 "u" "E2U+sip"
“!^.*$!sip:[email protected]!”
Private Root
Private ENUM
Enterprise
Location Server
Call
Control
15
Call
Control
IN NAPTR 10 10 "u" “E2U+mailto"
“!^.*$!mailto:[email protected]
Private IP Backbone
Call
Control
Call
Control
QoS: Perceived if Not Real
+ Many Carriers & Enterprises utilize MPLS for Real-Time Transport
▪
Connection oriented traffic engineering with bandwidth protection
▪ Quality-of-Service mechanisms (e.g. voice prioritization)
▪ Secure MPLS Tunnels/MPLS Virtual Private Networking
+ Problem: No Exit
NRD
▪
MPLS protects the on-net traffic
▪ There is no way off
▪ Firewalls are never touched
Federated Extranet
(Domain Bridging NAP)
THIG
Redundant carrier-grade
THIGs utilized by one or
more federation members
Internet
Gateway
VoIP
Gateway
MPLS CORE
INTERNET
SITE A
Internet/External connectivity is a
completely separate connection
PSTN
SITE B
Signaling
Bearer
16
MPLS and ENUM
Federated Extranet (Domain
Bridging NAP)
DNS
SS7
INTERNET
NRD
ENUM
PSTN
THIG
MPLS Carrier A
MPLS Carrier B
DA
Corporation A
(MPLS VPN A)
17
Corporation B
(MPLS VPN B)
Signaling
Bearer
Corporation B
(MPLS VPN B)
Corporation A
(MPLS VPN A)
Public and Private: A Real Example
IN NAPTR 10 10 "u" "E2U+sip"
“!^.*$!sip:[email protected]!”
Austrian Public Root
IN NAPTR 10 10 "u" “E2U+mailto"
“!^.*$!mailto:[email protected]
VeriSign Private
Root
Private IP Backbone
Public IP Backbone
Company 2
Company 1
18
ENUM with SIP Functionality
EREG
IN NAPTR 10 10 "u" "E2U+sip"
“!^.*$!sip:[email protected]!”
Device Resources
Tier 1 ENUM
Tier 2
Location
Server/Registrar
Option 1
Call
Control
Call
Control
IN NAPTR 10 10 "u" “E2U+mailto"
“!^.*$!mailto:[email protected]
Option 2
ENUM
Call
Control
Option 3
Perimeter Security and Interop Resources
19
Call
Control
ENUM Issues to Be Resolved
+ Critical Mass (the Network Problem)
+ Application developers
+ Public or private directories
+ Update rate
+ One or many - providers, databases, …
+ Regulatory and policy issues
+ New identifiers
+ Coverage
+ PSTN Service Logic
20
Conclusions
+ ENUM is currently a mess
+ Private, Public, Mobile applications are uncoordinated and there is
mass confusion
+ Keep the end goal in mind – creating a public IP infrastructure for
applications (voice, video, IM, gaming, etc)
+ Opt-Out of Opt-In
+ First to 30 million wins
+ Anyone doing Private ENUM that is not peering is being shortsighted
+ And finally…
21
ENUM is the Preferred Protocol of
Gangsta Rap!
Thank You