Campaigning for better Connectivity and Universal Broadband Access!

Heres one improvement to the proposed Copyright Infringement code.

 

Progress on the Digital Economy bill will be dependent  on the skill of law makers to offload the Copyright Infringements code of practice to Industry and Ofcom to sort out.  This looked to be the original plan of the Digital Britain report and in observing snippets of the webcast from the House of the Lords,  it was good to see many of the amendments were being withdrawn on the basis that they will be covered in the industry code.  It was good to see that the much fought over amendment 138 in EU telcoms package being used to draw a boundary around the limits of the code.

 

The problem in writing a bill,  before the industry code is written and agreed,  is that the bill will list a whole series of things a code should look like based on the imagination of the good Lordships (Whitty, Lucas, deMauley, Earl of Errol, to name four) and their advisers, whom are a very erudite bunch and are quite brilliant at finding all sorts of peculiarities.  Given the very imperfect nature (its a kludge) of the internets naming and addressing system, their list of tabled amendments will be worth a special edition of Hansard.  They have to guess, because industry has given them so little to work with,  but it could produce something unworkable and impractical to implement.  The evidence gathering, the letter writing, another letter,  a means to appeal and as yet undisclosed technical measures, the costs are mounting,  and the behaviour we are seeking to change can itself  change in shape through encryption and other measures.

 

Left to itself industry would or ought to come back with something different, a fix, best agreed within industry and consumer groups.  It is built on the assumption that copyright infringement problem and network congestion are related. Tackle the congestion issue in an open and transparent way and the copyright issue is mitigated.  ISP like to claim their 'up to' speeds while  placing their fair use clauses on their online help pages. The latter allude to limits of the system without explaining the actual resources allocated too each broadband service.  If as anecdotally told a high proportion of traffic is p-2-p and and a proportion of this is illegal copyright as alleged then adopting the ISPs 'fair use' policies to tackle copyright infringement would be a mutually beneficial line of attack.

 

Fair use policies form part of a users contract,  and would normally stipulate that ISP is free to take measures to protect its network for the benefit of all if an individuals activity is over using a resource that needs to be shared.  In this particular case the Fair Use Policy could be adjusted to specify in traffic terms per month the amount of peer to peer traffic above which that particular data stream (application specific) is then limited to say 5kbps.   As it stands many ISPs limit this traffic type at busy hours.  It's not a hammer,  it is quite surgical and and the ISPs have tools to do it.  It would serve another purpose.

 

Current Broadband services lack transparency in terms of resources applied per customer.  Developing and exposing 'fair use' policies would lead to better informed customers.  Our connectivity is shared so has limits,  managing congestion is essential element of delivering customer service.  The fact the industry is also dealing with copyright issues would be a benefit to an industry which needs to become more transparent in its practices.  Industry have a ready way of dealing with the problem in a generic way,  the costs appear when you begin dealing with individual cases.  Is their a way to match a CIR (copyright infringement report) and match it with a congestion report and trigger a clause in a fair use policy? Yes, win, win surely,  the biggest offenders are dealt with and all those costs are kept at bay.

 

Earl of Erroll did pose the question as what volume of records in a CIR would trigger an action.  The answer ought to be the number that causes a breech in his ISPs fair use policy.  The action then is not a letter, but an automated action to slow down one particular data stream,  which provides relief to anybody else using that connection.  It's not cutting anybody off,  but eases congestion and mitigates the copyright problem while keeping the lawyers out of the process.

 

 

 

 

 

 

 

 

AT&T has made a submission to an FCC consultation where it requests a sunset date for the legacy phone service.  AT&T would like to divert the increasing costs of sustaining the PSTN service into delivering more Broadband! The AT&T filing states 700,000 customers cancel their phone line each month and move to mobile or a broadband service supporting VOIP.  The cost of maintaining the legacy network is having to be divided between fewer and fewer customers.  AT&T maintains that 50% of operators investment capital is going into maintaining a legacy service.  The cynics will suggest AT&T is seeking to shift the cost of legal obligations to support unprofitable customers and an unprofitable service back to the federal authorities,  while it picks the customers it wishes to serve with fiber and mobile services.  Maybe,  but the delivery of more and more services is dependent on good connectivity and an intelligent device, and fewer on an aging legacy voice services.

 

Drawing parallels with the UK are always tricky, as the differences in circumstances mitigate against a direct comparison.  The UK plans;  that's BT plans for next generation access is centered on delivering Fibre to the Cabinet (FTTC) for 40% of homes and Fiber to the Home available to all new builds.  FTTC demands you retain your existing phone line and phone service. The Ofcom consultation on Next Generation Networks notified us of another four years of call conveyance charges and the change in focus by BT from migrating PSTN customers to an all new IP based phone service to delivering FTTC and FTTH in a £1.5bn programme.  The former was due to deliver a substantial part of the £1bn annual operational savings from the 21C programme,  the status of which is now less clear.  Expectations need to be somewhat restrained given the allocation of costs for copper between PSTN and Broadband are allocated heavily towards PSTN.

 

Although the last mile copper will be need to be maintained and powered it would be interesting to determine how much money could be saved if the obligations to run a PSTN service were removed on the understanding that the money freed put would be invested in extending NGA.  The PSTN obligations  could be replaced by an obligation on Openreach and partners to deliver a data transport network (fixed and mobile) capable of supporting real time services.  It is important the obligation is on Openreach and not BT Group as they are a essentially a facilities provider to all operators and do not offer end customer services.

 

The spectrum modernisation programme being consulted upon currently makes no reference to the possibility of replacing the PSTN service with a more convenient mobile service.  Ian Livingston has been quoted in the Guardian as seeing the process as a gift to mobile operators!  For me, spectrum is just another path for data and thus licensing for the specific use of a Mobile service is fundamentally wasteful.  The opportunity must exist for BIS/Ofcom and BT to trade spectrum for a PSTN sunset date where costs savings could be diverted to extending NGA investment - That £1bn pa of operational cost savings must still be there to be released if legacy costs can be cut. This could be an insurance policy against the Digital Britain levy being caught by an early election.

 

Tying spectrum to an inidividual vendor and service is sub-optimal for customers but it may release additional investment.  Given this is unlikely to occur and a little sub-optimal, BBBritain will persist with a request for specific spectrum to be allocated for home cell use so users can get a better mobile experience and get more from our Broadband connections.  Customers day to day use of their computing devices and the available connectivity remains a long way ahead of both the regulatory definitions and operator propositions.

 

Update 6/1/2010- On December 22nd BIS issued an extention to the consultation on spectrum modernisation emphasising that spectrum can be shared.  This provides a good deal of latitude.

 

Dec 15th.  My wife and children has been asking me what I would like for Christmas. There asking alone is a present beyond measure.  I have, however written a letter to Father Christmas.  It is a most peculiar request.  Can I have some radio spectrum Santa?  I do not want it for myself,  we can all share it.  I just need the spectrum allocated so we can improve Digital Britain for everyone.  There is feel good family movie here somewhere,  but stop that thought there!  The workings of the electromagnetic spectrum whose mathematical workings were first revealed by Maxwell is a good deal more magical than the TV content delivered on that medium,  so we should be careful how we use it and share it.

 

In a follow up to the blog on femtocells I have submitted some ideas to the BIS consultation on its intended direction to Ofcom for the Digital Dividend.   You can download it from here.  

 

In the submission I am seeking clarification which is a polite way of asking for specific spectrum to be assigned so our home or broadband services can work more seamlessly with our mobile access.  The intention is to begin to treat our mobile and fixed data transport as a single high speed fabric engineered to support critical services of the future.  It is an extra dimension to the Digital Britain universal service commitment moving us towards  'connectivity' as opposed to discussing fixed and mobile. 

 

The BIS consultation on the Digital Dividend finishes on January 8th.  It is a one in ten year opportunity to comment on the use of a natural resource every bit (yes, bit!) as important as fossil fuels.  If you agree with the sentiment please drop BIS or Ofcom a one line note of support or better still feel free to take this effort,  improve it and re-submit.

 

Happy Christmas.

 

 

 

 

 

 

 

Developing a Universal Broadband Service Obligation

and a national Data Transport network.

(From best effort browsing to quality home working.)

The move from a Universal Broadband Service Commitment to a Universal Broadband Service Obligation could be driven by a simple policy goal of demanding that the UK data transport infrastructure supports a quality home working experience on a national basis.

If the policy makers were to define an Obligation as opposed to a commitment based on todays broadband infrastructure, how could it be written? It could not be an 'up to' speed or an Ofcom average. The obligation would be expressed as the delivery of a browsing experience in the peak hour needing 30Kbps per user keeping packet loss kept at about circa 3% of packets delivered and ensuring the networks were never more than 95% loaded so they do fail due to congestion. This, amidst all the speed claims and the propaganda this is our lot  from the perspective of describing what data transport exists and what sort of an obligation could be agreed at todays 'affordable' prices.

If the policy goal was set to move from an undeclared best effort browsing experience which supports lots of downloading in off peak periods to a goal of delivering quality home working (for instance), then our connectivity would need to support full video conferencing, virtual private networking, streaming, and file sharing. Excluded is the need provide a full multi-channel TV experience. The goal is simple to state, but in stating it we need to accept that we are reconfiguring a whole industry to achieve that goal. Such a goal needs declaring at some point if the Government is to be successful in transforming the delivery of key public services, care , heath and education. Tax collection can be done on the network equivalent of wet string, but delivering services requiring human interaction demands more than a best effort browsing experience.

It is essential to note that this will cost more to deliver but the underlying data transport is then capable of support legacy voice services. The affordability argument goes away when you add the cost of todays affordable broadband experience to the cost of an affordable fixed voice service and the cost of an affordable mobile broadband and mobile voice service and treat them as single bill item.  This is what convergence means.  A well configured data transport service is disruptive in nature. Todays connectivity is just a taster.

Moving from a simple browsing outcome to a quality home working outcome could be done in a networking and engineering sense pretty quickly, but the disruption to the existing status quo would be immense.

We can use the 2Mbps headline number,  only because it means little.  Quality home working would demand access to fixed and mobile communications. It would demand a much lower packet loss rate .1% loss so video and multi-media calling could be supported. The termination device will need to double as a mini-mobile base station. A more generous peak hour allowance is needed. The customer would have additional transparency of network parameters

Put simply the underlying data transport would be re-tuned to allow permit it to do more of what we expect. This would include using those resources to support existing voice services.

On its own it would suggest the cost would be higher, but if done correctly any increase would come from removing the costs of legacy services and using that money to increase the resources assigned to each persons connectivity.

Setting what is a straight forward goal, demands you start describing the UK's data transport infrastructure. These are just some of the characteristics needed.

At its simplest a national data transport infrastructure will deliver each1 bit packet of data to and from anywhere in the UK in 20 milliseconds or less. The volume of data carried and its peak hour loading will be the subject of a long debate, but whatever the peak load and its distribution be it 1 terabit or 3 terabits, the national data transport infrastructure should deliver 99.9% of its load in less than 100 milliseconds anywhere in the UK.

Low loss rates and statistical predictable outcomes permits key service delivery to be planned.

The data transport infrastructure consists of fixed and mobile components in a single converged fabric, using the same protocols, and engineered to make the most efficient use of the electromagnetic spectrum. The does mean the users Broadband connections should act as mini base stations, leaving the macrocells for what they are best at which is facilitating connectivity for fast moving vehicles.

This will mean a change to spectrum auctions, or at least should do. Five mobile operators having exclusive use of spectrum where 70% of the data is coming from devices which are stationary is not a good use of radio spectrum.  My home has more bandwidth than a mobile base station,  and my home if treated as a mini-sell,  which is already is for other frequencies is perfectly placed for my unconnected neigbours to access services they need.

If the base station as broadband end point is accepted as it should, there exists the means of converging fixed and mobile services into a single proposition and product set.

It changes dramatically the current regulatory divisions of fixed and mobile services. It proclaims the end of voice and data. It is a common data transport infrastructure whose job is the transfer of bits.

It demands a move away from cost recovery based on call conveyance and the creation of a regime bases on bits transferred.

On the one hand it suggests that infrastructure providers jointly plan a national data transport fabric, perhaps two fabrics , where the network operations of existing providers merge a create choice of no more that two fabrics. Openreach, BTW,and two mobile operators make one network group, Virgin Media, C&W, the remaining two mobile operators, and the network operations of Sky and Carphone networks forms the other group.  This is from an network planning perspective.  High speeds and cheaper bandwidth can only be achieved if fibre is efficiently deployed, rather than 5 sets of infrastructure in urban areas and none elsewhere,  where all that is needed is a choice of two.

The move by mobile by network operators to share networks is just a realisation that building five networks is an inefficient thing to do. It is also an appalling waste of radio spectrum when compared with using mini-base stations at the end of broadband connections.

The network providers would be regulated to provide world class data transport. They would be motivated and rewarded for transporting bits. The more bits they carry, the more they earn, the latter interconnect replaces the existing call conveyance regime which recovers historical costs.

There would need to be a clear demarcation between data transport and those selling services over that transport. All retail services providers would have equal and transparent access to services provided by the transport carriers.

The is a short summary. The ongoing focus on speed rather than outcomes is a real concern. By focusing on an outcome, you get a much better debate on how to deliver the outcome, the policy changes  and the business case needed to support the improvement in connectivity. The investment in our connectivity is a function of the outcomes we aim for.

Those like Kip Meek (BSG), Carphone Warehouse (Andrew Heaney), even BT say they cannot see the need for high speed access on a national basis. Reconciling that with view those demanding fiber everywhere is hard to explain,  but both are looking at two different things. Data Transport infrastructure configured to deliver a best effort browsing experience of 30Kbps is only fit for the purpose stated and no other, so it does lack ambition. Those demanding fibre roll-outs are less guilty, but never the less need to specify the outcomes we expect.  Without a significant re-configuration of TV broadcast delivery then the best case to support investment is the need to support key service delivery. In this context the delivery of a quality home working experience acts as useful proxy to define what connectivity is needed for a whole range of critical services.  The investment in next generation access will be a by-product of this change.

 

The challenge for incumbents (Fixed and Mobile operators) is that in order to get at increased revenue streams from Broadband  they will need to sacrifice their legacy voice and messaging revenues.  They like Ofcoms competition policies are becoming victims of their own success.  In delivering and popularising first generation best effort access to the internet and services hosted on the world wide web, it is apparent to users that notion of fixed and mobile is dated and the notion of voice and data is equally out of date.