diginomica has not been alone in covering the growing mismatch between BT’s service delivery capabilities and the expanding needs of business in an increasingly digitalised world.
Many places in the UK (with a couple being `home territories’ for some diginomica contributors) have very slow Internet (with 7Mbit downloads being typical).
But even BT’s `superfast’ broadband, at around 20Mbit/sec, is still positively moribund when compared to services available in many other places around the world, especially to mobile services consumers in the Far East.
The calls for action from BT have been many, and over some time as well, but the lack of extensive real competition, particularly for last mile alternatives, has put the company under little pressure to extract any digits and get on with it. It is still difficult to get any type of cable service available in some areas, but expanding that faces the same return on investment issues that BT faces in laying fibre optics to the street, let alone individual properties.
So the extra performance of fibre optic cabling remains, of itself, a premium service. The practical upshot of this is that many businesses in the UK paid hefty premiums to get a level of bandwidth that individual consumers consider their minimum service level on their smartphone in many Far East locations.
But, many will say, those Far East services are for consumer mobile: that is a completely different ball-game to providing internet services for businesses. For now that is still true. But the inflection point is approaching, and getting nearer. Soon it will not be true.
That inflection point will be in the form of 5G wireless communications, and it is an area that semiconductor ‘chipzilla’ Intel is starting to eyeball with some real interest, as a chat with the company’s General Manager for 5G products and services, Rob Topol, at the recent Intel Developer Forum in San Francisco indicated.
This is the next stage of mobile services development that could……might just…….. take the market away from not just BT, but all providers of fixed line services.
Yes, fixed line services the world over are by no means dead, yet. But their executioner is now sharpening its axe and studying the position of the chopping block.
The ‘hot air’ starts to solidify
5G has been coming for some time, but the announcement by Huawei earlier this year of an intermediate 4.5G service offering 1Gbit bandwidth on network that would allow operators to start their 5G infrastructure investments with an immediate return on the investment, indicated that the 5G ball is now very much in play and should be part of current planning for both operators and business users.
Operators could start offering pretty fast premium services that would pay for the introduction of at least some of the network infrastructure needed to also then run 5G. That would mean many of them will be broadly 5G-ready when 5G comes along in around 2020.
The question then becomes whether businesses can at last move from their reliance on wired internet services? Does 5G offer the next real step forward – all wireless/mobile capabilities for all aspects of business and consumer use? Topol says:
The only comment I have on 4.5G is that we do support it. It is an evolution of 4G and therefore re-uses some of the existing infrastructure. But 4.5G is also a platform for 5G in as much as this is about getting every inch of spectral efficiency out of 4G, and it will require new infrastructure deployments.
Topol sees future implementations of 5G falling into three main categories – narrow band applications in areas such as IoT where communicating with thousands of sensors are the key requirement. The middle track, where most business and consumer users will reside, where faster broadband offering higher data rates will be important.
The third category is also going to be attractive to many businesses. The millimetre wave band, where frequencies of up to 30 GHz can provide very high levels of data traffic. There is, of course, a downside to this, which is that signals at these frequencies do not propagate well, which therefore means they are strictly line-of-sight transmissions without any ability to pass through objects put in their way.
Letting go of cables
Despite this, the mix of capabilities available with 5G means, in Topol’s estimation, that many users will move their entire service requirement away from fixed cable-based services and onto all-wireless infrastructures.
This is because there is no way of defeating the reality that cabling the last mile from local exchange to individual properties is rarely going to provide a healthy RoI. Topol says:
We see 5G being used to provide something like a microcell in a neighbourhood that can use the short range line-of-sight capabilities of millimetre wave frequencies. It will be like having a big wireless pipe into the home.
In practice, therefore, the most common 5G infrastructure will be fibre optics to a microcell – whether that is in a housing neighbourhood for consumer users or to smaller microcells in an office campus environment for business users. Large enterprises could then have microcells in individual offices.
The other factor 5G brings here is the ability to mix different transmission environments together. It would be possible to have high capacity fibre optics to a large, millimetre wave microcell that then fed cells that used lower frequencies with better signal propagation capabilities to service individual users and systems in a typical office environment.
One important contributor here is this ability to re-use existing technologies to fulfil user needs. For example, one of the key established technologies being pursued is re-purposing existing TV frequencies. The idea is to aggregate them together to provide fatter bandwidth for users, explains Topol:
Most of the lower frequencies in the available spectrum are already accounted for, often assigned to old TV services that are no longer used. It is not a simple task to aggregate them together, but they are still there, available, and not being used for anything.
This approach can then be coupled with fat-pipe fibre optic cabling to regional nodes and millimetre wave microcells delivering line of sight fat-pipes to individual locations. The final leg could use local microcells that aggregate lower frequencies into fat-pipe deliveries that propagate through buildings to end users and services.
Intel has also developed services such as FlexRAN, an open-standards Radio Access Network system, that can manage shifting workloads onto resources most appropriate for the task. It can make decisions on what level of processing work is undertaken at the core of the network and what work goes out to the edge and the local cells, all of which will have full processing capabilities available.
Recently there was also the introduction by Intel of Silicon Photonics, small devices that integrate the electrons of compute resources and the photons of fibre optic communications systems. This is likely to be one of the key bits of `glue’ that holds 5G together as a practical deliverable.
But what will it do for the business?
The main question for businesses, of course, is whether 5G can be a solid, reliable alternative to a wired internet, and will it increase the diversity of services that are available to them? Topol argues:
They need to think about it this way. First of all the data component is growing at an unprecedented rate and there is no stopping it. Therefore, it is going to break the existing networks of today, and we are reaching a point where services will start to degrade. So users need to do one of two things.
One is to look for additional efficiencies that make use of every inch of available capacity and maximise every resource they have. But even here they will still need to expand capacity. The second is where 5G comes in using different modulation schemes, sub-carriers, frequency or time-based separation etc.
The wired infrastructure is still very important to Intel and the world, but the law of diminishing returns applies when trying to reach every corner of the earth. In some markets that will mean turning to wireless, and some emerging markets will jump straight to it.
This is why Intel is making sure it covers all these technology tracks by trying to give operators more options. They can then look at the demographic and geographic issues they have to deal with.
The business interest here could have long term implications, especially as the over-riding push is for businesses to transform themselves with increasing levels of digitalisation. The inevitable partner of that is the need for faster, richer broadband capabilities with the fattest bandwidth pipes it is possible to get.
That points to variations on a 5G theme, where combinations of service provision mixing fibre-optic cabling where it exists through to millimetre wave microcells and aggregating microcells that deliver good bandwidth services to the individual wherever they are working can be provided more economically than fully cabled infrastructure.
So while the hot corporate news of the moment involves the exploitation or otherwise of national taxation regulations, the long term story for many enterprises is likely to be about which countries are selected as future locations for business, based on the fatness of the internet `pipe’ that can be provided. And the chances are that these may well not be in Europe – and on current evidence, certainly not here in the UK.
Technologies come and technologies go, and with the approach of 5G wireless communications, the days of the cabled internet, at least out towards the edge of the network are now numbered, even for the largest of enterprise users. So the days of the cabled internet infrastructure services companies will also come into question. There are some that might say this can’t come a moment too soon.
Image credit - Freeimages.co.uk