World bandwidth growth over the next decade — is it viable?

Demand for bandwidth over the next decade could seriously challenge the economic viability of conventional telecommunication network architectures including modern networks utilising technologies such as IP/MPLS and ethernet.

One of the major drivers of future bandwidth growth is the growing popularity of personalised video services particularly as they migrate to high quality and higher definition video standards. Additionally, large file transfers like digital images and thin-client computing will significantly contribute to bandwidth demand. Access technology would need to migrate away from copper-based DSL towards FTTH to enable these future services, but if FTTH does develop into a mass market access technology over the next decade, bandwidth growth could be unprecedented.

The advent of ubiquitous broadband via ADSL provided an ‘always-on’ paradigm coupled with a flat-rate charging regime. Operators had little incentive to install significant increases in backhaul and core capacity because they saw little or no revenue growth due to the flat-rate charging regime. The current ADSL-driven broadband era was sustainable by extrapolation.

The future situation, however, could be very different. The potential for bandwidth growth, from a services-driven perspective, is possibly two to three orders of magnitude greater than the bandwidth from the broadband networks of today. Simply scaling the technology and architectures of today's networks won’t produce viable solutions.

Future content-rich services will continue to drive bandwidth growth in networks and increasing demand for these services will be a driver of fibre into the access network. New architectures that can enable economic deployment of FTTH need to be found but if fibre does penetrate access networks on a large scale, there is potential for huge bandwidth growth throughout the network hierarchy.

The advent of higher definition video content, together with user behavioural changes such as increasing personalisation of program material, will drive huge increases in network bandwidth. When combined with distributed storage and content systems such as peer-to-peer file sharing, sustained, busy period user bandwidths could go beyond 10 Mbps with peak rates requiring several hundred Mbps to individual users. Today’s DSL and cable modem technologies won’t be able to scale economically to meet these unprecedented demands.

FTTH will be a catalyst for massive bandwidth growth, but meeting demand for new high bandwidth services is only one of the drivers for the deployment of FTTH. Other important drivers include:

• meeting competitive threats
• reducing operational cost
• staying internationally or regionally competitive
• new revenue generation

Once a fibre-to-the-premises (FTTP) network is installed, there is no competing network technology that can outperform it in terms of technical capability. It has also been argued that operational savings alone could justify FTTP being installed.

As new high bandwidth requirements emerge, the demand for faster access speeds from end users and their dissatisfaction with slow networks could come to the fore. This demand for much faster access speeds (not necessarily huge increases in sustained bandwidth) will also increase pressure on operators to put high bandwidth FTTP networks in place. This may be especially so if there are neighbouring competing regions offering FTTP and by doing so attract inward investment, boosting local economies, employment, house prices, etc.

FTTP will enable huge levels of bandwidth growth if deployed on a large scale. By using service and traffic scenario models, it is possible to get a feel for the level of bandwidth growth that may arise in the future. The models use a range of generic services including video services, two-way communications and data services such as email, gaming, thin-client computing, etc.

Three different scenarios for world bandwidth growth were developed, the major difference between them being the levels of FTTH take-up. The optimistic scenario assumes half of broadband connections would be FTTH for developed countries by 2020, the middle scenario approximately 25% and 10% for the pessimistic scenario.

The optimistic scenario had greater take-up and usage of video services and in the later years personalised video usage was assumed to account for the majority (~80%) of entertainment video watched and had substituted the majority of off-air broadcast video viewing hours. It wasn’t assumed that overall viewing hours would increase. The middle and pessimistic scenarios assumed that 50% and 40% respectively of television viewing hours would be substituted.

The pessimistic scenario is assumed to be the result of operators not finding economic solutions to FTTH deployment and not changing the architectural structure of their networks; they therefore stay with DSL as the dominant access technology. The middle and optimistic scenarios assume that a transition from today’s networks to a new streamlined and lower cost architecture occurs, the difference between the middle and optimistic mainly being the timescale and rate of the transition.

The above discussion shows that new services and particularly personalised video delivered over the internet could drive massive increases in bandwidth demand. The advent of FTTP/H will enable this demand to be met if network operators can find ways for large-scale economic deployment. The key to this deployment is to simplify the network and remove as many expensive electronic nodes and terminations as possible. This, coupled with lower cost optical technology bought about via greater levels of component integration, could enable the true broadband future to be viable and attractive to the network operators and then bring closer a future where users are empowered to personalise their services with an ultra-fast network to deliver them and finally see an end to the worldwide wait.

*David Payne is the systems manager of the Institute of Advanced Telecommunications, Swansea University, UK.