Harry Potter and the Broadbandius Confundus curse

Maybe it’s because it’s the 20th anniversary of the publishing of the first Harry Potter book, but whatever the reason there seems to be a lot of magic in the air at the moment. Both the Tories and Labour have their own magic money trees and Ed Vaizey seems to have been hit with a Broadbandius Confundus curse whereby he believes the simple act of choosing “superfast broadband” will magically deliver superfast speeds.

As Hermione Granger will tell you without dashing off to the library to look it up, the effects of the Broadbandius Confundus curse is to forget all about the laws of physics we mere muggles have to contend with, and that 10Mbps download speed and 2Mbps upload speed is sufficient to squeeze our daily digital lives through.

Hogwarts students who study the effects of curses and spells on the non-wizarding world should study the effects of this common curse by watching this video:


Fortunately, there is a counter-spell which makes you almost impervious to the effects of the Broadbandius Confundus curse. It comprises reading and understanding twenty simple points, so simple that even Crabb and Goyle are reputed to have fully understood them:

  1. “Superfast broadband” is a made-up term by BT’s marketing team. It doesn’t really exist.
  2. What it comprises is a bunch of technologies, running signaling systems (variants of the DSL family) over copper wire
  3. Copper wire is a good conductor of electrical signals, but it also has intrinsic resistivity to these electrical signals
  4. These signals are alternating current, so the type of resistivity involved is called impedance
  5. It is impedance that causes most of the loss of signal strength with distance - the longer the copper wire, the weaker the signal gets and the slower your broadband. The number of people signing up for the service is nothing to do with it
  6. So Ed Vaizey is correct in that anyone can sign-up to the service, however this does not mean they’ll get “superfast’ speeds. In fact there is nothing super or fast about superfast broadband
  7. Indeed if, like my Dad, you’re at the end of a long piece of copper wire, you will not get “superfast broadband” - the most he gets is about 6Mbps download and 2Mbps upload. Yet he’s CHOSEN, as Ed puts it, to sign up for superfast broadband - it’s those pesky laws of muggle physics getting in the way
  8. Dad lives in Surrey. 15 miles way from Diagon Alley, central London, in case you’re wondering
  9. Where they don’t have their own network, VirginMedia resells BT’s services - as do a lot of broadband suppliers - so often it’s the same stuff, with a different brand...
  10. Virgin’s own network uses a different signaling system (DOCSIS) over a different type of copper wire - in this case it’s “COAX cable”, the type of cable you use to connect an aerial to a TV. Unlike with BT, no one else gets to sell this service, they don’t have what the industry calls a “wholesale” option
  11. The mix of DOCSIS and copper coax cable is slightly better at handling high electrical frequencies, so it provides faster broadband than BT’s set up, and Virgin have been doing some smart things in their network to further boost their broadband speeds (port bonding). But even Virgin are tying themselves up in knots as on their posters they talk about fibre optic broadband and show a picture of a copper coax cable
  12. Neither of these systems (BT’s copper wire + xDSL and Virgin’s copper coax cable + DOCSIS) is FIBRE OPTIC broadband because they both use copper wires to connect to our houses
  13. This make a huge difference as laser light through fibres obey different laws of muggle physics from electrical signals over copper wire - even a few feet of copper makes a massive difference
  14. Laser light through fibre is to all intents and purposes impervious to distance - at least as far as the distance to connect houses to a telecoms network is concerned - it is after all used for trans-ocean undersea cables
  15. Because of this, a real fibre optic network (or pure fibre or full fibre) is a much fairer network - and this is the really important thing – as if by magic everyone receives pretty much the same speed irrespective of where they live - up a mountain or in the middle of London
  16. A side effect, and a very welcome side effect, is that the right kind of fibre optic network can deliver 1000Mbps symmetrical (upload and download) or MORE (10Gbps for example). If it were a broomstick, it would be a Nimbus 2001, as compared to Ed’s Cleansweep 1.
  17. Why is this important - because it’s a higher quality service - try downloading a Mac or Windows system upgrade (4GB or more) on a feeble 10Mbps service and see how you get on, or when working from home taking part in a multi-party Skype call on a 1Mbps uplink
  18. It’s for these reasons that copper-based broadband services should not be promoted as “fibre optic”, as it’s a) incorrect and b) misleading
  19. But not all fibre optic networks are the same. In the UK, Gigaclear, Hyperoptic, B4RN are among the very few that can deliver 1000Mbps (1Gbps) symmetrical services today. Where BT do provide real fibre optic services, which is as rare as a unicorn, they provide up to 300Mbps down and 30Mbps up
  20. We have to have a Brexit angle on this - the FTTH (Fibre To The Home) Council of Europe publish stats of each member state’s fibre optic broadband penetration - remember this is where the optical fibres terminate at the house and there’s no copper used - guess which country isn’t even on the charts? Now that we’re leaving the EU, and our soon-to-be-former-partners are zooming away with REAL fibre optic broadband, we’ll be left in the uncompetitive copper slow lane… unless you live within reach of the service providers I mentioned earlier

So there you have it. The counter-spell to Broadbandius Confundus curse which disillusions those affected so that they see copper broadband for what it is:

A damp squib.

The Broadband Manifestos

With the impending general election, I thought it might be instructive to see what the various parties were offering with respect to broadband in the UK, which might influence your vote one way or another. Or not, as the case may be. 


We will deliver universal superfast broadband availability by 2022. Labour will improve mobile internet coverage and expand provision of free public wi-fi in city centres and on public transport. We will improve 4G coverage and invest to ensure all urban areas, as well as major roads and railways, have uninterrupted 5G coverage. On day one we will instruct the National Infrastructure Commission to report on how to roll out ‘ultrafast’ 300Mbps across the within the next decade. 

My Views On Labour's Broadband Proposals

In my opinion this is vague and woolly. It's not inspiring. There's nothing super nor fast about "superfast broadband", which is a made up marketing term anyway, and what does "availability" mean? Uninterrupted 5G coverage - what does "uninterrupted" mean? 

The promise to instruct yet another report to be written on how to roll out fibre to the home (when we already know how to do this), with such a low target of 300Mbps (is this symmetrical or asymmetric? What's the upload speed, Jeremy?) misses the mark and, based on what the rest of the EU is doing (read my post here), isn't competitive enough.

The 300Mbps makes me think this is an allusion to BT's GPON-based optical infrastructure, so this is simply a "more of the same" proposal.

Marks out of 10: 2, as they really haven't tried very hard


We will ensure that consumers and businesses have access to the digital infrastructure they need to succeed. By the end of this year, 19 out of 20 premises will have access to superfast broadband and our Universal Service Obligation will ensure that by 2020 every home and every business in Britain has access to high speed broadband. We will work to provide gigaspeed connectivity to as many businesses and homes as possible. We will introduce a full-fibre connection voucher for companies across the country by 2018 and by 2022 we will have major fibre spines in over a hundred towns and cities, with ten million premises connected to full-fibre and a clear path to national coverage over the next decade. 

We have similar ambitions for mobile phone coverage. By 2022 we will extend mobile coverage further to 95 per cent geographic coverage of the UK. By the same date, all major roads and main line trains will enjoy full and uninterrupted mobile phone signal, alongside guaranteed WiFi internet service on all such trains. We will continue to release more spectrum from public sector use to allow greater private sector access and begin the roll-out of a new 5G network, providing gigaspeed connection to your smart phone. We plan to have the majority of the population covered by a 5G signal by 2027.

My Views On The Conservative Party's Broadband Proposals

"We will ensure that consumers and businesses have access to the digital infrastructure they need to succeed" is an encouraging start... but then it runs out of puff as quickly as a superfast broadband connection does. 

The pledge "19 out of 20 premises will have access to superfast broadband" is pretty meaningless as "access" isn't defined and no service performance parameters are mentioned. As "superfast broadband" is electrical signalling over copper wire, this isn't surprising.

The USO is for an underwhelming 10Mbps download speed only - this is not "high speed". This is barely adequate now, let alone by 2020. It might be irksome to the pro-Brexit faction of the Tories, but the EU is lightyears ahead already.

A voucher to be introduced for full-fibre connection for businesses means they are promising a voucher, and not the connection - is this weasely words? Again, nothing about speed of connection or what will actually be delivered. And this doesn't apply to the consumers referred to in the opening sentence - remember this: "We will ensure that consumers and businesses have access to the digital infrastructure they need to succeed"?

Towns and cities already have fibre optic "spines", so I don't think this is anything new or startling.

5G "providing "gigaspeeds" to your smartphone". According to the recently published technical specification by the ITU, 5G will deliver in dense urban areas, TARGET experienced data rates of:

  • downlink data rate of 100Mbps and 
  • uplink at 50Mbps 

“Experienced” means real world data flows in terms of bits delivered over Layer 3. These are "megabit" speeds, not "gigabit" speeds, or "gigaspeeds" as the Tories refer to it.

Unless they are referring to the somewhat theoretical, under ideal conditions etc, "peak data rates*", the minimum requirements for which are:

  • Downlink peak data rate of 20Gbps
  • Uplink peak data rate of 10Gbps

If they are referring to this, then there's a distinct whiff of "smoke and mirrors" here. But then it's a manifesto, after all... 

Unlike manifestos, technical specs are full of details and the ITU document I'm referring to can be found here but before you rush off to read it, here's an extract: "Peak data rate is the maximum achievable data rate under ideal conditions (in bit/s), which is the received data bits assuming error-free conditions assignable to a single mobile station, when all assignable radio resources for the corresponding link direction are utilized (i.e., excluding radio resources that are used for physical layer synchronization, reference signals or pilots, guard bands and guard times)."

Marks out of 10: 4 as they made a bit of an effort, but it's basically the status quo

Liberal  Democrats

Invest to ensure that broadband connections and services to be provided before 2020 have a speed of 2Gbps or more, with fibre to the premises (FTTP) as standard and unlimited usage by 2020 across the whole of the UK. SMEs should be prioritised in the roll-out of hyperfast broadband. 

Ensure that every property in the UK is provided, by 2022, with a superfast broadband connection with a download speed of 30Mbps, an upload speed of 6Mbps, and an unlimited usage cap. 

Invest £2 billion in innovative solutions to ensure the provision of high-speed broadband across the rural UK, working with local authorities and providing grants to help areas replicate the success of existing community-led projects. 

My Views On The LibDem's Broadband Proposals

This floats my boat as it's the real deal - they actually seem to get it. The marketing term "Hyperfast" is often over-hyped, but is usually regarded as 1000Mbps (1Gbps) symmetrical. If this proposal sounds familiar, it's because it's the same as the House of Lords' proposed amendments to the recent Digital Economy Bill which I wrote about here.

The insistence on 2Gbps speeds forces through the infrastructure switch from electrical signalling over copper wires to optical signalling through glass fibres all the way through to our houses a.k.a fibre optic broadband or "full-fibre" or fibre to the home (FTTH). The 30Mbps/6Mbps USO is smart too, because that's really hard to deliver in rural areas using electrical signalling which would need to also be upgraded.

They are the only party that's referenced the special case of rural broadband "working with local authorities and providing grants to help areas replicate the success of existing community-led projects". This reference to "existing community-led projects" must be the remarkable B4RN project which you can read about here, but other funding options exist for rural broadband such as Gigaclear here.

Rural areas are only "special" because those pesky laws of physics of electrical signalling over copper wires simply don't allow for decent service provision where the population is geographically dispersed i.e. where the lengths of copper wires are longest, the effect of impedance is greatest and the end user experience the worstest.

However the LibDems don't mention cellular or mobile broadband using 4G or 5G, they don't mention the poor coverage we all experience on trains and how this really needs to improve.

Marks out of 10: 7 as they've used their imagination, would have been a higher score if they'd mentioned mobile.

The Green Party

I rummaged around on their website but couldn't find a 2017 manifesto, probably my bad. They do however have 10 pledges, but they don't mention broadband. This is disappointing as a 2Gbps fibre optic broadband network (which uses less energy than a copper+electrical network) such as proposed by the LibDems would make working from home a practical proposition for a lot of people, and would help with rural regeneration. Plenty of bandwidth for hi-def multi-party video conferencing and services we can barely imagine today, all of which would obviate the need to travel by (electric) car.

Marks out of 10: 0


If the only thing that matters to you is broadband internet access, then the LibDems are the party of choice. Of course you might say they're unelectable so they can say anything they like, but at least they've come up with an imaginative proposal. 

Broadly speaking, there really isn't much difference between Tory and Labour on broadband, as they both seem to be saying "plus ça change", so you'll have to make your voting decision between these two on something different

House of Commons ushers in a new Dark Age

Digital Economy Bill 

Laser light through optical fibres provides high speed broadband that's largely unaffected by distance. This is why this technology is used for the ocean-straddling networks that joins us all together.

Yesterday in the House of Commons, UKplc got short changed. At the moment of leaving the EU, when we need every competitive advantage over our soon-to-be-former colleagues in the world's largest trading block, the House voted down amendments proposed by the Lords which would have meant we'd have had "full fibre" broadband delivering target speeds of AT LEAST Two Gigabits per second (2000Mbps) by 2020.

This of course would necessitate Fibre to the Home (FTTH), replacing the old and unfit for purpose copper network that was deployed by the GPO under the last Universal Service Obligation (USO) for low bandwidth analogue telephone services.

The House of Lords proposed this time around that there should be a USO of 30Mbps download with 6Mbps upload. This may sound pretty feeble compared to 2000Mbps, but given the incontrovertible laws of physics, to deliver it on copper wires would require significant investment, especially as the minimum COMMITTED information rate would be 10Mbps.

Ying tong ping pong iddle I po*

After a 90 minute discussion, the changes that the Lords proposed were killed off in the interests of saving time as no one appeared to have the heart for a game of Parliamentary ping pong on this important national infrastructure issue.

So what have we ended up with?

Quite possibly the worst result possible.

An utterly unambitious and pretty useless minimum 10Mbps download, nothing about upload, and no Universal Service Obligation.

In other words, nothing changes. We're now stuck for the foreseeable future with electrical signalling for our broadband.

The European Enlightenment

Meanwhile, the rest of the EU is zooming ahead with the lighting up of what we call "full-fibre" broadband (to distinguish it from the copper-based "fibre broadband" so prevalent here).

Full-fibre broadband is where the optical fibres terminate at your house, which is why the rest of the world calls it FTTH, Fibre to the Home. It's this technology that delivers 1000Mbps+ speeds and, over the short distances involved in delivering to your home, there's practically no drop in speed caused by signal loss. Which means:

  • You get pretty much the speed as promised
  • Everyone in the country gets the same speeds irrespective of where they live

The FTTH Council of Europe regularly updates and publishes data on how member states are doing. Here's the latest update from earlier this year:

As you can see, despite the valiant efforts of Gigaclear, Hyperoptic and B4RN, with less than 1% fibre optic broadband penetration we're not even on the chart.

And the 2017 Digital Economy Bill does nothing to change this.

Maybe the Commons should rename this bill to "Digital Bill - Economy class"?

What we could have had - the Lords' prayer


“(2B) The universal service order must specify that the target for broadband connections and services to be provided before 2020 must have— 

  1. (a) speeds of 2 gigabits or more; 
  2. (b) fibre to the premises (FTTP) as a minimum standard; 
  3. (c) appropriate measures to ensure that internet speed levels are not affected by high contention ratios; 
  4. (d) appropriate measures to ensure service providers run low latency networks. 
  5. (2BA) The universal service order must specify as soon as reasonably practicable that, by 2020, the following will be available in every household in the United Kingdom— 
  6. (a) download speeds of 30 megabits per second; 
  7. (b) upload speeds of 6 megabits per second; 
  8. (c) fast response times; 
  9. (d) committed information rates of 10 megabits per second; 
  10. (e) an unlimited usage cap. 
  11. (2BB) In meeting the obligations set out in subsection (1), internet service providers have a duty to ensure that their networks offer at least the minimum standards specified in subsection (2BA) to every household in areas of low population density, before deploying their networks in urban areas. 

Life in the slow lane

A short while ago PM David Cameron put in place a Universal Service Obligation (USO) whereby UK citizens had the "legal right to request a broadband connection capable of delivering a minimum speed of 10Mbps by 2020."

Note that this isn't the same as the right to RECEIVE such a service, and in any case 10Mbps is already too little too late. And this refers only to the download speed, the upload speed is generally 1/10th the download speed based on the DSL technology widely deployed, and 1Mbps at best is pretty feeble.

Meanwhile, this is what's proposed in the rest of the EU: 

"The Commission hopes that the ECC will help Europe meet its target of providing 1Gbps (1000Mbps) broadband to schools, hospitals and large businesses, and a minimum of 100Mbps for all households – which need to be upgradeable to 1Gbps – by 2025."

Where's Wally? The UK is conspicuous by its absence... From FTTH Council Europe FTTH Panorama webinar 6.4.2017

As we are leaving, we won't be subject to this, and Ofcom don't seem to be on the same page as the European Commission when it comes to broadband, preferring instead to tinker with organisational change, rather than infrastructure change.

Because those pesky laws of physics means that to deliver a "minimum of 100Mbps to ALL households" requires swapping copper wires for optical fibres (the case of VirginMedia is different as, where they have their own infrastructure, they already use a different tech).

So our continental competitors and former EU colleagues will soon be in the optical fibre fast lane, with all the competitive benefits that provides, while the UK, or potentially what's left of the UK, will be limping along on copper wires.

If you're not into telecoms, this would be like Europe having motorways and the UK having only leafy country lanes.

Which would be somewhat... slow, don't you think?


ASA to (finally) review the marketing of broadband services

One of my bugbears is the way that copper-based broadband has been marketed as fibre optic broadband.

There's a massive difference between the two, but when I raised this with the ASA over a decade ago, I was told by them that the amount of copper in the network compared to the amount of fibre is so small it makes no difference.

However - the laws of physics of an alternating current over copper wire (BT and BT's resellers) or over copper coaxial cable (VirginMedia) are quite different from the laws of physics of a laser beam through glass fibre (Gigaclear, Hyperoptic, B4RN).

Which is why fibre optics are used for undersea trans-ocean cables... 

The common starting point for broadband offered on "full-fibre" services is 1000Mbps symmetrical i.e. the same speed uploading or downloading, but is future proofed and much faster speeds can be offered by a simple upgrade.

A typical BRN customer's Speed test

A typical BRN customer's Speed test

But more importantly, this also means that you get pretty much the same speed irrespective of where you live - on top of a mountain or in the middle of London.

In my view there should be a clear distinction in the marketing of services based on copper and those that are pure or "full" fibre and this review is long overdue.