1

The Development of VoIP

in Fixed Networks

A Report for DTI

John Horrocks
Horrocks Technology Limited

with

David Lewin

Peter Hall

Ovum Limited

27 February 2001

1

Table of contents

Executive Summary

Objectives

Networks

Services

The market

Telephony services types

Terminals

Access

Networks

Traffic

Conclusions

1Introduction

1.1Terminology

1.2Acronyms

2IP Technology and networks

2.1The Internet protocol

2.2The commercial drivers behind IP

2.3IP networks and their state of development

2.3.1The public Internet

2.3.2Managed IP networks

2.3.3Comparison

2.3.4ATM

3Networking issues

3.1Identification systems

3.2Routeing

3.2.1Routeing in circuit switched networks

3.2.2Routeing in the public Internet

3.2.3Routeing in Managed IP networks

3.3Protocols and signalling

3.3.1Session Initiation Protocol (SIP)

3.3.2H.323

3.3.3H.248 & Megaco

3.3.4BICC Bearer Independent Call Control

3.3.5Tiphon

3.3.6Proprietary protocols

3.3.7Comparison of protocol stacks

3.4Next generation network architecture

3.4.1Introduction

3.4.2Softswitches

3.4.3Transport structures (ATM, IP and MPLS)

3.5Network boundary devices

3.5.1Firewalls

3.5.2NATS

4VoIP services

4.1Service types

4.1.1Categorisation

4.1.2State of development of services

4.1.3Wholesale services

4.1.4Related services

4.2Service provision

4.3Terminal types and availability

4.4Access configurations and user installations

4.4.1Analogue access

4.4.2ISDN

4.4.3xDSL

4.4.4ADSL

4.4.5VoDSL

4.4.6Cable modems

4.5The home of the future

5The Retail Market

5.1The current UK telecommunications retail market

5.2Internet access

5.3Commercial models

5.4Implications for new forms of telephony

5.5Differences between the UK and the USA

5.6Conclusion

6Forecasts

6.1Telephony service provision

6.2Telephony services types

6.3Terminals

6.4Access

6.5Networks

6.5.1Network types

6.5.2Network use

6.6Residential traffic forecasts

6.6.1Qualitative

6.6.2Quantitative projections to 2005

6.7Summary

7Conclusion

Contacts

1

Executive Summary

Objectives

Technology and liberalisation have led to many changes in telecommunications during the last decade but greater changes are in prospect for this new decade.

The Internet Protocol is recognised as the common transport system for networks of the future. The public Internet with its email and world wide web information service have become part of everyday life although few people had used either eight years ago. The process of moving voice services onto IP is beginning.

This report has two objectives:

  • to outline how networks will change over the next 5-7 years with respect to voice communications
  • to estimate the proportions of voice communications of residential and small business users that will be carried on IP in fixed networks

Networks

With IP based networks, there is a clear separation between access, transport and service provision. Two types of IP transport networks are used to provide services:

  • The public Internet, which was conceived from the computer data communications culture and is an “open” network of interconnected networks run by different parties and available without restriction for the creation of services through functionality at the edge of the network,
  • Managed IP networks, which are essentially closed and support only specific services that are created by the network operator or an authorised service provider.

Figure 1 summarises the differences between the Internet and managed IP networks.

Figure 1: Comparison between the Internet and managed IP networks

Internet / Managed IP
Approach / Open = unrestricted access and use / Closed = restricted access and use
Functionality and service creation / At network edges by users or independent third parties / Within or at the edge of networks but only where enabled by the operator
Charging / Subscription or traffic volume only / Can support usage based charging
Connectivity at the IP level / Full interconnection / Hardly any at present
Addressing / Public global system / Internal private addressing possible
Quality of service / Not managed / Managed to support defined levels

Whilst there are distinct differences between Internet and Managed IP networks, there is a widespread misunderstanding that the Internet cannot be used to provide a delay intolerant service such as telephony because of its unpredictable quality of service. However service providers can and do add a layer of management[1] to the Internet to improve quality of service, and a significant proportion of international traffic to some destinations is already carried on the Internet by wholesale services that are used by many large and well known operators.

IP is not the only new network technology. ATM (Asynchronous Transfer Mode) is a link layer “small packet” technology that can either be used underneath IP or be used to provide virtual circuits to carry media without involving IP. Many of the new systems designed for telcos who want a next generation PSTN are ATM based and so the report includes these ATM developments.

Services

We focus on two main service types:

  • Public telephony, which uses only E.164 numbering.
  • Internet named telephony[2], which uses only Internet naming and so far has not yet developed into a full any-any service and whose quality is more unpredictable and may be significantly lower.

We use these service distinctions based on identification because they relate to users and are independent of terminal types which will change. The Internet trade-jargon, however, uses distinctions based on terminal types:

  • Phone-Phone: a bypass service which uses ordinary telephones and the PSTN for access and termination and the Internet for the long distance or international part of a call. It includes calling card services.
  • PC-Phone: where a call can be made to a traditional telephone from a PC and where the call will be carried most of the way on the Internet and handed to a circuit switched network operator at the terminating end.
  • PC-PC: where a call will be made entirely on the Internet.

Although at present:

  • Phone – Phone = Public telephony
  • PC – Phone = Public telephony
  • PC – PC = Internet named telephony

These relationships will change and become more complex since in the future:

  • Phones may support Internet naming as well as E.164.
  • public telephony will also be supported on PCs via IP based networks.

There are four main categories of company that are providing services:

a)Traditional telcos with directly connected customers and a significant proportion of their own infrastructure

b)Telcos who work mainly or wholly as indirectly connected operators (their customers use carrier selection, carrier pre-selection or 2-stage call set up) and use leased lines for transmission

c)Internet Telephony Service Providers, who predominantly use the Public Internet and have little infrastructure under their own control

d)Internet Service Providers who are adding voice related service features to their access services.

Figure 2 shows the relationship between the service categories and the development routes being followed by the different players. The terms “PC-PC” etc have been added but apply only with the current state of their relationship to services.

Figure 2: Development routes

The current situation is that:

  • The market for public telephony from a PC is growing and selling primarily on low prices for bypass traffic
  • Internet named telephony from PC to PC is starting as a free function provided by ISPs and portals. Usage is driven by both cost savings and functionality (mainly the presence feature[3]).
  • Both the Internet and managed IP networks are being used for public telephony bypass services
  • Very few telcos have yet migrated onto managed IP networks

The market

Twenty years ago the market was characterised by:

  • Most traffic being telephony
  • Nearly all revenue being from telephony
  • Calls being charged per minute
  • Prices being strongly dependent on distance
  • International calls subsidising national and local calls

The growth of the Internet has changed the market completely by providing hugely growing data traffic funded by subscription and partly by advertising. Within a few years the market will be characterised by:

  • Most traffic being data
  • Significantly reducing revenue from telephony
  • Prices being distance independent and fixed network call charges being replaced by subscription (mobile calls will probably continue to be charged per minute because of the relatively high economic cost of radio capacity)

There are two drivers for users to change their service providers:

  • Cost savings
  • Functionality and ease of use

When prices were high, cost savings were the main issue for users and this created a strong bypass market for calls, but as prices drop, cost savings will be less of an issue leaving functionality and ease of use as the main drivers for users.

Users currently pay both access and call charges. Call related costs and charges are reducing and billing and administration are becoming an increasing proportion of costs. These changes are creating pressure for calls to be included in subscriptions.

We think that the combination of technology advances towards multi-service solutions and price reductions will drive a process of combination in service provision in the retail market. As prices reduce consumers will find it easier to buy everything from one service provider. Figure 3 shows our conclusions. The separate market segments of:

  • Circuit switched access line paid by subscription
  • Internet access paid by subscription
  • Fixed telephone calls paid according to use

will be replaced by:

  • Access line paid by subscription but including Internet access
  • Basic services including email, fixed voice calls and data paid by subscription with some funding by advertising. These services will probably be combined with the access line
  • Premium services paid separately, probably on a pay-as-you-go basis using electronic payment

Figure 3: Changes in retail market segments

We expect that telephony service provision will go through a period of fragmentation as new players enter the market providing bypass services that compete on price. This sector will grow and then decline as price becomes less of an issue and calls begin to be charged by subscription. Some bypass operators may be able to continue by reselling services or bundling services with non-telecommunications services such as power.

This period of fragmentation will be followed by consolidation as the volume of IP based traffic grows, as call related charges for basic telephony disappear and the pressure for simplification and combination takes over. Figure 4 shows these developments.

Figure 4: Voice service provision

It should be noted that this figure is not indicating the technology used. For example changes from circuit switching to ATM or IP will take place within the services provided by the telcos.

Telephony services types

Public telephony based on E.164 numbering is universal. Although many people consider numbers to be less memorable and user friendly than Internet names, they have the advantage of being used across all languages, alphabets and cultures. We think that E.164 numbering will continue as the basis for any-any public telephony for the indefinite future. In other words we think that for the foreseeable future it is unlikely that telephony users will cease to have E.164 numbers.

We think that Internet named telephony will grow rapidly as a service for informal closed user groups, but that the any-any capability that would make it a public service will not develop until the second half of the decade because of addressing and standards issues.

In the long term we think that public telephony and Internet named telephony will exist in parallel.

Terminals

Currently almost all the residential market for telephony terminals is a combination of analogue wired telephones and cordless telephones. VoIP is supported only from PCs where users have special software and use a headset.

As VoIP develops, we expect that PCs will be fitted with conventional style handsets. This will either be through the addition of sockets for analogue telephones or through new phones with USB interfaces. We think that USB handsets with matching non standard software in the PC are the more likely solution. These development may coincide with the introduction of better support for VoIP in Windows.

The next development will be the introduction of a standard self standing IP based telephone using either Ethernet or USB as the interface. This development will be driven by the move towards an integrated digital home system and the introduction of new network termination points for both public and Internet named telephony. This development will occur later than the USB handsets for PCs because it is more dependent on standards. Figure 5 shows these developments.

Figure 5: Trends in terminal use

Access

Nearly all residential users have analogue access lines; few have ISDN. BT is wholesaling ADSL access to its own ISP (BTInternet) and also to other ISPs.

We expect that the prices for ADSL access will fall to below £20/month within two years and that user demand will be high. We are not so bullish about local loop unbundling which is starting very slowly and where the experience from the USA is that there can be significant operational difficulties. This may mean that BT operates most of the ADSL and resells the access to other ISPs.

The next step for residential users is likely to be the presentation of services via a single standardised IP based network termination point (Ethernet and USB seem to be candidates at present) The introduction of this interface will contribute to growth in IP based telephones. We expect that this new form of network termination point will start around 2005. We think that the development of standards will be the determining factor in the timescale. Figure 6 summarises these developments.

Figure 6: Access developments

Networks

There are three main types of packet network used for voice:

  • Public Internet backbones.
  • Managed IP networks
  • PSTN replacement networks.

The same underlying transmission system can support all three network types and some operators are running or planning to run all three types with partitioning based on MPLS.

The two major technological developments will be:

  • Improved quality of service in managed IP networks through further development and implementation of MPLS and the introduction of special quality control protocols that give voice packets priority over other traffic.
  • Further refinement and implementation of SIP, which is expected to be the main call control protocol for voice and other media services.

We think that these networks will develop in the following ways:

  • The public Internet as a “best efforts” network will not change much.
  • Operators of Internet backbones will add QoS features as soon as practicable in order to be able to increase their revenue by offering service level agreements. Improved QoS be introduced from 2003
  • Network interconnection at the call control level using SIP over IP will not start before about 2005 because of the need for further standardisation.
  • Nearly all procurement of new PSTN capacity will use PSTN replacement networks based largely or wholly on ATM. For straightforward PSTN applications, operators will not be in a hurry to replace their local circuit switches, which can continue to be used at marginal cost. Where customers want advanced services, they will use an overlay of managed IP networks.
  • In the very long term, the managed IP and PSTN replacement networks may merge and SIP may become the dominant protocol for call control.

We think that BT and the cable operators will start to offer VoIP services from their managed IP networks to the SME and residential markets starting from 2001/2. They will address the SME market with broadband and VPN services delivered over ADSL and VoDSL. These services will include voice but initially may not have all the features required for PSTN substitution.

For PSTN replacement, we think that BT and the cable operators will introduce an overlay of packet based technology in very limited areas, starting from around 2004. This overlay will grow gradually according to need and some circuit switches may still be in use after 2010.

Traffic

Figure 7 shows qualitatively how we expect public and Internet named telephony from PCs to grow. The relative proportions of public and Internet named telephony are impossible to predict with confidence. Internet named telephony will grow more slowly initially as there is a square law effect based on the probability of both parties using PCs, but since most residential call minutes are traffic between regular correspondents (family and friends) the potential for growth is enormous.

Figure 7: Growth of voice traffic from PCs

Figure 8 shows the quantitative projections.

Figure 8: Projections of residential minutes

Figure 9 draws these various forecasts together into a single table for the near, medium and long term future.

Figure 9: Future scenarios

Period / Short term
2001-2005 / Medium term
2006-2010 / Long term
2011 onwards
Services / Public telephony (E.164) universal
Internet named telephony grows but only for informal groups / Public telephony (E.164) universal
Internet named telephony becomes an any-any service / Both exist alongside each other
Service provision / Fragmentation / Consolidation and battle between ISPs and telcos / Impossible to predict
Terminals / Analogue unaffected
Growth phase for telephony from PCs / Analogue terminals start to decline
Growth phase for standard IP telephones and integrated home systems / Integrated home systems
Access / Separate analogue and ADSL NTPs / Analogue access declines
Growth phase for new IP based NTP / Standard IP based NTP
Networks / Growth phase for bypass and ITSPs
Growth of wholesale services
Growth of global IP managed networks without interconnection / Replacement of circuit switched networks with SIP on IP or BICC on ATM. Media carried direct on ATM in many networks
IP based interconnection implemented / Slow migration to all-IP as SIP gradually replaces BICC and IP is used without ATM and SDH

Conclusions

The provision of voice related services over fixed telecommunications networks will change significantly in the next decade with increasing use of packet technology. The main conclusions about the way in which networks will develop are: