Thursday, January 21, 2010

Estimating future bandwidth requirements

The Broadband Stakeholder Group's Predicting UK Future Residential Bandwidth Requrements study from May 2006 reported that:
"By 2008, the bandwidth demand for the most bandwidth intensive households could reach 18Mbps downstream and 3Mbps upstream. By 2012, the bandwidth demand for the most bandwidth intensive households could reach 23Mbps downstream and 14Mbps upstream."
The two most bandwidth intensive household types from the eleven identified were "two adults, young couple" and "two adults, with children". A key point made later in the study is that "a significant percentage of UK homes may not be able to receive such downstream speeds using current technology/networks." Bandwidth requirements were estimated from first principles and industry data, shown on the following chart:

Ante meridiem usage by the two adults, two children household is shown below:

While post meridiem usage is shown here:

Interesting to see educational use getting a mention. Perhaps a similar exercise could be undertaken to identify school broadband usage and consequential future bandwidth requirements? However, in comparison, CED Magazine (Communications, Engineering & Design) had this to say in January 2008:
"...analysis shows that by 2012 the power user will require over 100 Mbps of downstream bandwidth and the average user will require over 40 Mbps of downstream bandwidth. (A power user is defined as a first adopter of new technologies and would expect the best bandwidth. The average user is defined as a late adopter of new technologies and would be satisfied with standard bandwidth.)"
Their calculation, based on bandwidth per application and number of digital streams, including the number of connected devices (TVs, phones, computers, digital video recorders) per household, produced the following graph:

...showing usage significantly higher than the BSG's estimates. A 2006 study (Quantifying the Broadband Access Bandwidth Demands of Typical Home Users) by the Australian Centre for Advanced Internet Architectures concurs with these higher estimates, but also flags that requirements could be higher still, depending on the approach taken by ISPs:
"We take a closer look at the probable demand for bandwidth, presumed to motivate the deployment of new broadband access technologies. We construct plausible estimates of what a ‘typical home’ might need, given two different service provision scenarios - managed bandwidth and best-effort statistical multiplexing. We estimate a household of five people requires between 58 and 113Mbit/sec if bandwidth is managed on a per-application basis. If statistical multiplexing is used, the same family’s bandwidth requirements jump to between low hundreds of Mbit/sec to almost a Gbit/sec."
To explain managed bandwidth versus statistical multiplexing:
"Our estimates in the previous section make a key simplifying assumption - the home broadband service provides ideal isolation between traffic belonging to different classes of application. In other words, we have assumed that some form of quality of service mechanisms are deployed to ensure bursts of packets from one application do not temporarily ‘starve’ other applications (such as HD channels being rendered in real-time). This can only really be achieved if the ISP deploys and manages, moderately sophisticated queuing and traffic classification in their routers.
A more likely situation is that an ISP will utilise statistical multiplexing to ‘manage’ the interactions between packets belonging to different application flows. Statistical multiplexing manages through benign neglect and over-provisioning. Enterprise IP network operators often use a rule-of-thumb that the link speed should be 2 to 5 times the average capacity. In this case our pessimistic estimate would require something in the order of one Gbit/sec at the link layer to support the proposed mix of application traffic."
I would imagine that this second scenario remains the most likely, though I guess it must all come down to cost - is it still cheaper and easier to over-provision than to manage your network?

Finally, Connected Planet offer this in relation to the "how much bandwidth?" question:
"Bandwidth planning is nothing more than ensuring that your network can adequately support the dynamic service requirements of your customers...How can I build a network without a definitive amount of bandwidth? Well, the answer is – do the best that you can...So, how much bandwidth is enough? Rather than a definitive number, the real answer is more along the lines of “as much as you can deliver” given the circumstances in front of you. At a minimum, this means meeting and anticipating the demands of your customers better than your competition does. Ultimately, the broadband service provider of choice will be the one who delivers the most salient and valued services from the consumers’ perspective, over the most efficient network that can best throttle to meet customers’ varying demands. So enough about how much is enough. Build what you can – but keep in mind that there will be a direct correlation between the network speed you can deliver and the capacity you have to dynamically deliver what the customer needs."
Wise words I think.

Thursday, January 07, 2010

Next Generation Fund consultation begins

The Department for Business, Innovation and Skills (BIS) has today launched a consultation on how the Next Generation Fund (aka the 50p landline levy) described in the Digital Britain final report should be deployed. This follows on from a Treasury consultation lauched in December about how the levy itself should be implemented.

On a positive note, it's good that local authorities get a couple of mentions as important delivery partners, but no recognition of the potential to extend the reach of existing education networks (as provided by RBCs, JANET, local authorities etc) to support currently un- or underserved areas. A shame, as is the fact that the author clearly did more homework on NGA's potential to support healthcare rather than education: the examples given are much more detailed.

Some specific comments:

68. Whilst in time, next generation mobile and satellite broadband services may start to match the headline speeds of fibre-based connections, we do not believe the overall service levels will initially be of the same quality. Inherent problems with backhaul capacity, contention in the network and latency will mean an inferior service compared to a fibre-based connection.

...backhaul and contention will continue to be issues regardless of what technology is used in the last mile? A satellite connection with high latency has limitations sure, but it's it's still much better than dial-up or no connection at all (how long before these two equate to the same thing I wonder).

69. Coupled with the fact that the Fund will be created by a landline duty on fixed lines, we believe it is right that solutions provided by the Fund should be fixed line, fibre connections only.

...yeah right. Obviously this is the same logic that ensures our road tax is used solely to maintain the road network. This seems like a thinly disguised ruse to ensure the money goes to the incumbents to me? Surely ruling out a set of technologies from the outset is stupid in such a fast-moving area - talk about stifling innovation!? This completely contradicts para 60 on page 19, which states that "Government’s approach is, as far as possible, to remain technology neutral and focus more on the services, applications and content that we would expect to be delivered through an NGA network, through specifying service quality criteria rather than the underlying technology".

70. In addition, the European Commission State Aid Guidelines for intervention in NGA markets state that “NGA networks are mainly fibre-based or advanced upgraded cable networks that are intended to replace in whole or to a large extent the existing copper-based broadband networks or current cable networks. ”

...I'm sure the EU folks that came up with this definition would be appalled to see it used as a get out of jail free card in this context. Also a bit rich if you don't specify what minimum bandwidth constitues NGA access?

81. The decision on which approach to use will be based on which approach has the potential for attracting the most bidders, and providing value for money for the landline duty payer, along with factors such as speed of deployment and the technology solutions posed in any bid.

..."value for money for the landline duty payer"...hmmm. I'm sure that the 35% or so of households that don't currently have broadband (many of whom have chosen not to have it, and are quite within their rights to continue to not to want it) will be very interested in the value for money that ensues from their £6 per year contribution.

85. An “inside out” approach has the benefit of deploying infrastructure closer to where the market is delivering, and therefore it is likely that the business case will stronger in these areas (primarily due to density of population and ease of deployment), with the subsidy provided by the Next Generation Fund. The economic impact of NGA roll out is also likely to be greater in these areas.

...but what about the social impact on delivering NGA where the market won't ever deliver? The areas that are worst served by a market led approach to NGA are the areas that stand to benefit the most in terms of improving people's quality of life?

All in all, not sure whether this represents progress or not. We certainly still seem to be an awfully long way behind Australia, New Zealand, USA etc?

Monday, January 04, 2010

US broadband policy: $183 Million to expand broadband access in seventeen states

On 17th December Vice President Biden announced an initial $183 million investment in eighteen broadband projects (many of which mention schools alongside other "anchor institutions"), benefiting seventeen states:
"The projects receiving funds today are the first in the $7.2 billion program – $4.7 billion through the Department of Commerce’s National Telecommunications and Information Administration (NTIA) and $2.5 through the Department of Agriculture’s Rural Utilities Service (RUS) – being implemented under the Recovery Act to expand broadband access and adoption across the country.  The awards are designed to help underserved – and often hard-hit – communities overcome the distance and technology barrier by expanding connectivity between educational institutions, enabling remote medical consultations and attracting new businesses – as well as the jobs that come with them.  They are part of an over $100 billion investment in science, technology and innovation the Administration is making through the Recovery Act to lay a new foundation for economic growth."
At the same time, the National Economic Council released Recovery Act Investment in Broadband: Leveraging Federal Dollars to Create Jobs and Connect America, which found that Recovery Act investments in broadband will "create tens of thousands of jobs in the near term and expand opportunities and economic development in communities that would otherwise be left behind in the new knowledge-based economy".

...which provides an interesting counterpoint to BSG CEO Antony Walker's comments at a recent Westminster e-forum that there is no case for broadband as an interventionist, Keynesian mechanism to stimulate economic recovery. Somebody forgot to tell that to our friends across the pond.