P717

1
P717 P805 SIRTE

• Study for Internet Roaming Throughout Europe
• Franco Guadagni - Telecom Italia / CSELT
• guadagni_at_cselt.it

2
AGENDA

• Scope
• Objectives
• Technical approach
• P717 results
• P805 expected results

3
Scope

• Narrow, focused scope
• No IP number roaming
• No terminal roaming
• No mobile IP issues
• Simple ISP roaming addressed
• Avoid international modem calls when abroad

4
Model

• Multiple ISPs in each country
• Problem similar to GSM roaming
• Same model for roaming solution
• Based on bilateral agreements between parties
• No central clearing point
• Distributed solution Scaleable and robust

5
Existing solutions

• Problem already addressed, solutions exist.
• Single ISP solution
• based on global ISPs
• limited choice
• cost problems only partially solved
• Trusted Third Party solution (e.g.. I-PASS)
• based on centralized architecture
• need to pay the clearing agency

6
AAA how does it work
NAS
AAA server
C/N interface e.g. PPP
C/A interface e.g. RADIUS
PSTN/ISDN network
IP network
7
roaming how should it work
AAAS (ISP R)
NAS (ISP R)
AAAS (ISP H)
N/A
C/N
A-A interface e.g based on HTTP
PSTN/ISDN network
IP network
8
P717 Objectives, results

• study current situation
• report on existing solutions
• define architectural model for EURESCOM roaming
• identify requirements
• draw possible alternatives for a EURESCOM
  solution
• identify work to be done on the different topics
• experiment international PSTN and ISDN Internet
  access between project partners

9
requirements

• terminal-network interface
• should work for PSTN and ISDN
• should work for most common devices and configs
• network-network interface (A-A protocol)
• should allow transport of all necessary
  parameters
• should be secure (encryption, mutual validation)
• should run over IP
• Compatible with existing third party solutions!

10
P805 objectives, priorities

• concentrate on A-A protocol definition
• choice between RADIUS, HTTP, LDAP
• define protocol parameters and procedures
• implement a demonstrator of the selected solution
  for the A-A protocol
• implement a simple solution for the user roaming
  software (to demonstrate feasibility)

11
Timetable

• spring choose protocol
• summer detailed protocol specification
• fall implement prototype solution, refine
  protocol definition
• winter demonstrate SIRTE roaming solution
  between two or more project partners.

12
Were not alone!

• Work together with IETF Roamops group
• input our ideas and proposals to the IETF
  standardisation process

13
We shouldnt say

• Were too late
• if we keep saying this, it will always be true
• Its not worth the effort
• True only if the Internet market stops growing
• Solutions are already available
• SIRTE and commercial solutions are not mutually
  exclusive
• the more roaming users we have, the more money we
  give to the clearing agency

14
We need

• Commitment, specific know-how and enthusiasm of
  project partners
• Strong liaisons with IETF group working on
  roaming
• Strong liaisons with ISPs within partners

15
What if we succeed?

• If we design a working solution for A-A protocol,
  theres still much more to do
• set up a framework to make real ISP cooperation
  possible (e.g. POP common distributed database)
• design a nice and simple client for easy roaming
  access
• define a technical framework for ISP to ISP
  agreements
• Theres still room for cooperation projects!