Wednesday, May 04, 2005

WTF is the NGN? 

The problem with the Next Generation Network (NGN) is not that there exists no definition what it is, it is the other way round: there are too many definitions and viewpoints, so everybody has his own opinion. This faith of course is depending what an individual thinks that the "Existing Generation Network" is. Bellheads living in a circuit-switching (TDM) world basically think the NGN is a "packet-based" network (because they like to be generic and want to avoid the term IP, because it is to "specific").

In turn the term NGN was not well understood (and if, not very much liked) by the netheads at the IETF, because they obviously consider "packed-based" as the "Existing Generation Network", not as NGN.

Living in both worlds, I personally consider (and have used it in this way intrisically in my blog) the term NGN as opposite to the "public" Internet, so the NGN is an IP-based network or application in a "walled garden". This separation can be in the transport layer (e.g. by having no or only a very controlled connectivity with the "public" address- or namespace), or on the application (or service) layer (e.g. Vonage or Skype), or both (e.g Yahoo!BB or the IMS NGN). This viewpoint is of course influenced by my work in ETSI TISPAN on IMS and also by my work in IETF on ENUM.

So I was very interested to visit the ITU-T Workshop on NGN in collaboration with the IETF and to hear the different viewpoints, especially what the IETF has to say on the NGN approach from ITU-T and ETSI.

The progam started after the welcome by Mr. Houlin Zhao (director of the ITU TSB) with a presentation from Brian Carpenter (IETF chair), stating the position of the IETF in general and citing from RFC3935 "The Mission of the IETF", 2004:
The goal of the IETF is to make the Internet work better.
and defining the Internet:
A large, heterogeneous collection of interconnected systems that can be used for communication of many different types between any interested parties connected to it. The term includes both the “core Internet” (ISP networks) and “edge Internet” (corporate and private networks, often connected via firewalls, NAT boxes, application layer gateways and similar devices). The Internet is a truly global network, reaching into just about every country in the world. The IETF community wants the Internet to succeed because we believe that the existence of the Internet, and its influence on economics, communication, and education, will help us to build a better human society.
He also set the scene of the IETF view in relation to the ITU-T NGN view:
We believe that the requirements and services described in the ITU-T NGN effort represent an important plan for future network activities, but
  • they are just one plan
  • we may disagree about the expected viability of
some design choices, based on collected experience in developing and deploying the Internet.
Next on the schedule was Session1: Requirements and Functional Architecture

For didactical reasons I start with the second presentation from Dave Meyer: A Brief Overview of the IETF and the Internet Architecture: Past, Present and Future.

The core statements of his presentation regarding the NGN where related to the End-to-End principle of the Internet:
... the end-to-end principle is perhaps the most fundamental and least
understood of the Internet’s architectural principles…:
  • Nothing should be done in the network that can be efficiently done in an end-system.
  • A function that must be performed at a higher layer should not also be performed at a lower layer (without a good reason)
He did not say it, but especially the first bullet point means: keep the network simple and stupid.

He then made a tour-de-force through the principles of the Internet architecture, for each point stating past, presence and future areas of activity:
Simplicity, Multiplexing, Transparency, Universal connectivity, Immediate Delivery, Subnet Heterogenity, Common Bearer Service, Connectionless Network, Minimal Dependency,Global Addressing, Regions, Mobility, Protocol Layering, Distributed Control, Global routing computation, Security, Network Resource Allocation
The conclusion was:
The Internet is evolving to support diverse and emerging requirements sets
  • Including many of the requirements that are beginning to be specified by projects such as the ITU’s NGN
  • Note this kind of “evolvability” is a fundamental property of the IA (deriving from its “minimalist semantics”)
  • So where are the architectural misalignments?
  • And where can we work together?
Ok, he knew the misalignments already, because Keith Knightson, Rapporteur ITU-T SG13 Q.3 -Architecture gave his presentation "Basic NGN Architecture - Principles & Issues" first.

Keith started with the Next Generation Network (NGN) definition from ITU-T Recommendation Y.2001:
A packet-based network able to provide telecommunication services and able to make use of multiple broadband, QoS-enabled transport technologies and in which service-related functions are independent from underlying transport-related technologies.

It enables unfettered (unrestricted - for non-native readers) access for users to networks and to competing service providers and/or services of their choice. It supports generalized mobility which will allow consistent and ubiquitous provision of services to users.
Ok, any net-head may sign this on a first look, but behold, WTF is meant by "services" and the "provision" of these?. "Service" is similar to "NGN", too may different definitions around.

One had to wait until the morning session 5 "Security" on the second day to get the simple answer from the IETF on this by Jon Peterson on slide 2 of his presentation:
On the Internet, telephony is an application
– Not necessarily a service, no service must be provided
But what is the understanding of ITU on "services? Let's have a look again at the presentation of Keith:

So on slide 4 (sorry, I have still problems putting pictures in the blog - so please look at the presentation itself) we have the horizontally-integrated network from ITU-T Rec. Y.2011 featuring two layers - the transport layer and the services layer. In the services layer we have telephone services, video services and data services (WWW, e-mail, etc.).

So phone and video is NOT a data service and WWW is a service? Interesting.

I do not want to comment on the rest of the slides (although it s tempting), just four of them:

Slide 8, the Basic Functional Architecture, which is basically a functional and very telltale view of slide 4:

First: so much about simplicity in the transport network: 10 "functions" including the customer "function" (whatever that is) - no further comment

Second: what about the "unfettered" access of users to networks (plural) and WTF are User-Network-Interfaces (UNI) and Network-Network-Interfaces (NNI) on the Internet? Note also very carefully that the UNI and NNI are at the transport layer and NOT at the service layer. Service interworking, third-party services and user access to services in "other" networks are only via the control point of the NNI (Session Border Controller).

And then my favorite - slide 13 (sic!) - the component/subsystem viewpoint - containing also the beloved IMS NGN et. al. Although I never really understood it - I have to admit I never really tried, because it gives me a headache - I see now two additional layers: the NASS/RACS layer and the applications layer (separated from the subsystem (services?) layer. I have to admit that this figure shows (different from slide 8) very tiny lines (NNIs?) to other networks on most layers (interestingly not on the RACS layer).

I do now want to comment in detail on back-up slide 27 - Generic Functional Architecture - enjoy yourself, just a last word on the IMS architecture itself (slide 28):

The major IMS network elements include (from here):

  • Access Gateway (AG): This network element provides an interface between the radio network (Access Network) and the IP-based network.
  • Access Network (AN): This is the radio portion of the network.
  • Breakout Gateway Control Function (BGCF): Controls resources allocation to IP sessions
  • Call Session Control Function (CSCF): Provides control and routing function for IP sessions.
  • Foreign Agent (FA): Advertises itself to mobile stations in serving area. Provides registration information to Home Agent. Forwards packets from mobile to Home Agent.
  • Home Agent (HA): Tracks current Foreign Agent serving the mobile. Forwards packets to current FA.
  • Home Subscriber Server (HSS): Can take the place of a HLR in all-IP network. Contains AAA function and other databases
  • Media Gateway (MGW): Provides interface for bearer traffic between IP and PSTN
  • Media Gateway Control Function (MGCF): Provides signaling interoperability between IP and PSTN domains – SIP to ISUP and vice versa
  • Policy Decision Function (PDF): As IP networks, unlike TDM networks, assign network bandwidth and resources in real-time, the PDF’s role is to assign resources according to demand and QoS requirements.
  • Position Determining Entity (PDE): While some mobiles can determine position independently, the PDE can provide assistance by way of location determination algorithms
  • SIP Application Server: Represents a platform for SIP application development and operation.
So this is the NGN idea of a simple network - at least 12 different network elements with approx. 22 different interfaces.

At the end of his presentation Keith raised some very good questions:
  • A recognition in both organizations that some new approaches may be required?
  • What kind of changes are required and/or foreseen?
  • Could/should the expertise of the ITU-T and the IETF be jointly applied to create the principles and architectural framework for next generation networks?
  • Do the ITU-T and IETF have different views on the architecture of next generation networks?
  • Can a number of differences (if any) be identified and catalogued for study?
  • Can ITU-T and IETF collaborate with a view to eliminating or (at least minimizing) any differences wherever possible?
My resumee is:

Within most of the topics discussed in the other sessions of the workshop (Nomadicity & Mobility, QoS, Control & Signalling Capabilities, Network Management and Security) at least the aim and goals where in common.

Identifiers, Naming, Numbering and Addressing where curiously NOT on the agenda AT ALL - maybe because this is not an issue at the IETF and ITU and ETSI do not get it yet that they will have a problem here - they still live in the E.164 name space.

But the detailed solutions within all these topics will require a common understanding of the basic architecture and there is still a long way to go.and the ITU, ETSI et. al. will have to change their position more then the IETF.

I agree that there are too many definitions and viewpoints to what the Next Generation Network (NGN)is.And that it depends on what an person thinks that the "Existing Generation Network" is. Bellheads living in a circuit-switching (TDM) world basically think the NGN is a "packet-based" network is confusing if not general.I wanna know also what NGN is .Im trying to educate myself to voip and the industry and technology and terms.But i find it all to confusing ,everyone has thier own definition that is different than the next.
Post a Comment

This page is powered by Blogger. Isn't yours?