Australian telecom operators are currently under challenge to provide competing service providers with open, wholesale access to their existing copper networks to support ADSL and VDSL platforms. Whilst the policy mantra from government is to deliver competitive and sustainable infrastructure, anecdotal evidence suggests that operators are managing this impact on their operations by investing heavily in regulatory resources to ensure that their position is not weakened. Industry media has enjoyed a rich source of comment and assertions from the duelling parties.

But where do existing HFC Cable Broadband networks fit in this new landscape?

Do service providers know that they can have early, open access to this existing superfast broadband infrastructure, or will they continue their narrow focus on deploying their ADSL2 in Telstra exchanges or will they wait (time unknown) for the NBN?

Existing HFC broadband networks can deliver the future now.

In the UK, through a similar New Zealand Government Broadband Quality Quarterly Reports initiative, Virgin Media boasts the average speed of their 10Mbit/s cable service is twice as fast as DSL’s 8Mbit/s service. Epitiro Technologies reported that Virgin Media 20Mbit/s plan’s average achieved throughput is 3 times faster than the advertised top speeds offered by BT’s ADSL service. Significantly, Virgin has seen 300% sales growth in its 20Mbit/s plan - 13% of new customer acqui-

sitions are signing up for 20Mbit/s speed; up from 4% in just twelve months – just as they were preparing to super-size their customers’ plan with the launch of their 50Mbit/s plan.

HFC Cable Broadband Networks use globally recognised and widely deployed technologies. Inherent in this IP-ready infrastructure are IETF, IEEE and Metro Ethernet Forum (MEF) standards with CableLabs standards compliance implicit. Increasingly, because of the shift in emphasis to fibre media, ITUT, IEEE and CableLabs are harmonising their strategic direction with channel frequency and frequency spacing. This shift will simplify product design, drive down long-term manufacturing costs and push the design performance to 10G line speeds.

HFC Cable Broadband Network operators around the world are traditionally new entrants into markets – they are the challengers and not restricted by the legacy of a copper network like many incumbent PTT’s.

The very fact that cable operators run fixed physical networks means they do not service the total market, incrementally build networks or seek government approval to build into green field estates – more often than not they fall below the regulatory radar

Constrained by geographic footprint – cable operators are challenged to extend their network reach. There has been little done so far world-wide to encourage HFC Cable Network operators to provide open access to competing Service Providers to their significant existing infrastructure.

Lord Currie, the Chairman of Ofcom (UK) noted that Virgin Media is not facing regulatory threat to ‘open’ their cable broadband network to Service Providers. Indeed, Currie noted that he could envision they might voluntarily open their network for good commercial reasons.

Our Australian cable networks are under-utilised. If a network was modernised to 1GHz bandwidth then we expect that a network could deliver up to 6.2Gbit/s of bandwidth to active users off any one node. The spare capacity presents an opportunity to open the network to service providers and for wholesale 3G & 4G backhaul transport.

In simple terms, there are two alternate methodologies to deliver open wholesale Ethernet access on cable broadband networks:

Channel Allocation is a technique where discrete 8MHz channels (or frequencies) are specifically allocated to service providers which they would in turn sell to end customers.

The Cable Operator (MSO) would assign both downstream and upstream channels to Service Providers and potential special interest users such as Government and Emergency Service organisations. The downstream channel would deliver about 55Mbit/s line speed of raw internet throughput and the upstream about 30Mbit/s. If DOCSIS 3.0 channel bonding is used the respective TCP throughput would be 222Mbit/s and 108Mbit/s respectively. As network engineers will attest, the statistical gain with channel bonding delivers massive speed capability.

The second methodology is Dynamic Bandwidth Allocation. This approach enables the cable operator to provide IP bandwidth to Service Providers on a point-to-multipoint design - it’s a ‘bitstream’ approach. The cable operator would be responsible for the provision of IP services and IP address management. The Service Provider would pay a fee to the cable operators for each end customer connected to the network and enjoy the security of a secure virtual network, using IETF Layer2 Tunnelling Protocol (L2TP) for "Native IP" based core networks, to connect their customers to their IP services platforms.

This approach positions the Cable Operator as the provider of QoS to Service Providers and the bundler of bandwidth and of prepay or post-paid billing usage plans.

Looking beyond everyday broadband and entertainment services, the cable network can be also opened to transport wholesale bandwidth (bitstream services) to mobile operators.

Cable operators can deliver open wholesale bandwidth services to 3GPP mobile operators using the existing HFC access plant, or the latest FTTP Cable PON network architecture, to backhaul IP traffic.

New internet technology with the grand name of Pseudowire Emulation Edge to Edge (PWE3) enables the Cable Operator to open the network and transport backhaul from base stations to their core network for mobile operators. An IEFT Working Group is working to refine the technology that transports services over an IETF-specified packet switched network.

As a fresh idea, the cable broadband networks could be opened to retail service providers and to 3G / 4G wireless operators to fill the massive spare capacity.