12

Wireless Radio Access Network Hardware

12.1 Historical Context

The next five chapters review the technology and commercial dynamics of network hardware. This first chapter looks at wireless radio access network hardware, but in practice the topic has to be broader. A modern radio base station in a mobile broadband network has a similar amount of software code to a digital switch a few years ago and has an ability to make decisions on admission control that would have once upon a time been taken by the central switch. This is useful because many of these decisions have to be taken so quickly that by the time the signalling had travelled to the centre and back again there would be too much decision-time latency.

Given that signalling travels more or less at the speed of light or 300 million metres per second this might seem odd, but in practice a six-kilometre round trip adds 20 microseconds to a delay budget and several trips might be needed to make a decision. Add a few million clock cycles to the process and it can be seen that distance becomes a nuisance, particularly for decisions that have to be made every few milliseconds.

Additionally, the once-clear distinction between computing networks and telecommunications networks irrespective of whether those networks are fibre, cable, copper or wireless or a mix of all four is disappearing, or has already disappeared. The process started some time ago.

Immediately after the Second World War, a decision was taken to capitalise on the work that Alan Turing had done at Bletchley Park on the Colossus code-breaking machine and build a general-purpose electronic computer.

The work was done at the National Physical Laboratory and was based on Alan Turing's conceptual discovery of the Universal Turing Machine, a computer that is not structured to carry out particular tasks. The result was a computer, the Automatic Computing Engine1 that at the time was probably the fastest in the world.

Thirty years later Seymour Cray delivered a Cray 1A supercomputer2 to the Aldermaston Atomic Weapons Authority where it still remained in 1990, the last operating Cray 1A in the world. The machine cost £8 million pounds and had a Freon cooling system and an early form of vector processing that allowed the computer to achieve then unrivalled operating speeds.

Both of these machines can be traced intellectually back to the 1832 Babbage Difference Engine – the ancient ancestor of today's modern computing industry and therefore by default the ancient ancestor of today's modern telecommunications industry.

In fact, it might be intellectually more rigorous to describe these not as computing or telecommunications networks but as information networks, networks that have the potential to transform the world in terms of social and economic progress, education and health and possibly even environmental sustainability.

There are, however, fundamental differences between computing and communication in terms of cost and investment return that need to be factored into these global ambitions. It may also be that present telecommunications investment in developed economies will not result in networks that are more cost efficient or energy efficient than the networks being replaced.

This is not to say that these networks will be socially and economically unprofitable and/or that investment is not justified, but rather that the returns might take longer than anticipated and be more broadly based in terms of social rather than purely economic return, networks that become the connection engines of the information age, delivering benefits that with political will can be evenly distributed across developed, emerging and survival economies, a net gain for all.

12.2 From Difference Engine to Connection Engine

Babbage's design in the 1830s embodied almost all the conceptual concepts of the modern electronic computer. The Babbage engine never quite worked in practice due to tolerancing problems and Babbage's tumultuous temper but it was intended to be a general-purpose machine that could be programmed with punched cards to perform almost any function. Figure 12.1 illustrates the complexity of this device.

Similarly, the Automatic Computing Engine and Cray Super Computer were described as general-purpose machines but were applied to a very specific purpose, both were used to model and analyse the behaviour of atomic explosions.

Modern cellular networks are similar in that they have had a general purpose, providing person to person connectivity but are now evolving into being broader multipurpose delivery platforms. However, at heart a cellular network is still a connection engine.

Moore's law has powered the cellular industry to the point where a cellular or fixed line switch that would fill up a large room in the 1980s can now be fitted in a box that can shipped by lorry or air lifted to deliver instant emergency communication into disaster zones.

Although voice communication has to date been the predominant application, these networks are now beginning to do far more both in terms of imaging and video and data.

We have come to assume that each generation of computer is faster and more cost and energy efficient than its predecessor. Similarly, it might be assumed that each new generation of cellular network provides faster connection speeds at lower cost with ever better energy efficiency and that this provides the basis for an improved financial return. Both of these assumptions are unproven.

The first practical automatic telephone exchange based on electromechanical switching was introduced by Almon Strowger in 1888 and was first introduced to the UK in 1912. Strowger exchanges similar to the one illustrated in Figure 12.2 remained in use until the mid-1990s, proving the point that telecoms equipment when fit for purpose can have a remarkably long life cycle, particularly when compared to modern-day computing life cycles.

The development of electronic switching in the 1950s proved to be a surprisingly painful and expensive experience with practical deployment only being realised through the 1960s in the US and rather later in Europe and the rest of the world. These exchanges, however, remain today at the heart of every telecommunications network including every cellular network in the world and are completely fit for purpose for the job that they were intended to do, setting up, maintaining and clearing down telephone calls. Figure 12.3 shows an AXE digital switch manufactured by Ericsson.

We sometimes forget that exchanges have also been efficient at generating and distributing power and for many people the idea that an old-fashioned fixed phone still works in a local power outage provides a curious comfort. However, there is now a general assumption that telecommunication networks including cellular networks are moving away from centralised switching and transitioning to distributed architectures with IP-based addressing and prioritisation.

12.3 IP Network Efficiency Constraints

There is, however, no compelling evidence that this transition will either be cost efficient, energy efficient or directly environmentally efficient. IP networks were and are considered to be more resilient than traditional networks based on centralised switch architectures. As mentioned in the Introduction, they bear a close resemblance to the postal service, where packets can and often do travel by many and various routes from origin to destination.

As with postal systems, packets can be stored (buffered) during the delivery process and as with postal systems packets can be lost or have to be deliberately discarded if the storage or buffer space overflows. This results in a variable customer experience. Post offices traditionally manage this by introducing differential pricing, first- and second-class stamps and guaranteed or signed for delivery. IP networks manage this by introducing multiple service levels based on multilevel packet prioritisation. In theory, this increases the multiplexing efficiency of the network, trading variable delay against network utilisation efficiency and trading delivery bandwidth efficiency against memory bandwidth occupancy. However, this requires routers to read and interpret IP addresses together with the priority labels embedded with the address. This information then provides the basis for a complex buffering and routing decision. This introduces variable delay into the delivery budget. The overall delay and the variability of the delay can be reduced by using hardware accelerators and high-speed memory.

These are costly energy-absorbent devices.

Even more frustratingly, just as in the traditional postal system, many of the packets delivered over the network are unwanted and have negative value – they annoy the receiver. These unwanted packets get in the way of latency-sensitive end-to-end services including voice communication. In a cellular network these costs are multiplied in several ways.

IP address overhead has to be accommodated within the RF link budget and the unwanted and unneeded negative value packets cost money to deliver. Wireless marketing teams have decided that users want and need to be offered high peak data rates even if this compromises average data throughput. High peak data rates, can only be achieved by using higher-level modulation schemes that are inherently noise sensitive and perform poorly most of the time resulting in user frustration and packet send retries that absorb network bandwidth and battery and network power. The higher-order modulation schemes require more linear and therefore less power efficient radio-frequency transmitters. The adaptation schemes needed to manage and mitigate these effects in turn introduce variable delay and add to the overall energy budget.

At this point you might question why the fixed and mobile telecommunications industry remains so set on IP network implementation. The idea that LTE networks will be more cost efficient and power efficient than the networks they are replacing is naïve. The notion that they can be profitable is not.

12.4 Telecoms – The Tobacco Industry of the Twentyfirst Century?

Telecommunications networks including cellular networks now provide access to instant information on a scale that would have been hard to imagine thirty years ago. Information is addictive and develops dependency, two core ingredients of a successful business model. Even better, the capital and energy costs of storing and managing the information in thousands of server farms around the world are paid for by other third parties.

For 150 years (from 1850 to the year 2000), the telecommunications industry has made money out of people's desire and need to interrelate to one another – the telegraphy and telephony age. Telegraphy and telephony provided a basis for social and economic progress based on social, economic and business efficiency, profitability with a social purpose and associated emotional value.

Information networks extend that model by helping us to interrelate with the physical and intellectual world around us. Earlier chapters discussed the role of the super phone and the interrelationship with the physical world around us. The next ten chapters shift this focus to this interrelationship between information networks and the intellectual world.

Information can have emotional value. It can also be misused, but in general it has to be assumed that informed decisions and actions are better than uninformed decisions and actions. Broader and more efficient access to information should at least theoretically enable wiser more broadly collective decision making, the world should become more intellectually efficient.

However, the tobacco analogy would suggest there may be longer-terms costs attached that are not presently visible and might only become evident in the distant future. It has taken Europeans 500 years to wake up to the health risks of tobacco but if the world was to worry about unknowable long-term risk then the world would never move on. In the meantime, the telecommunications industry appears to have reinvented itself more by accident than intent. The economic implications are, however, neither simple nor straightforward.

12.5 Amortisation Time Scales

Electromechanical switches were expensive to develop and slow to be deployed. In the UK depressed wage rates after the First World War meant that it was more economic to use human operators to set up a call. The financial returns from the investment were substantial but were based on an extended deployment cycle. The development cost of the Strowger exchanges still in use in the 1990s had been amortised over more than 100 years and the capital costs would have typically been amortised over forty years or more.

It would be surprising if the electronic switches installed over the past thirty years did not have a similar long tail in terms of investment return. Old technology takes a surprisingly long time to die. The counterargument is that the computing revolution has changed the economic rule book and that R and D and capital amortisation now has to be achieved in months not years and certainly not over fifty years or more. The fact that a Cray computer was still being used fifteen years after it was delivered is considered curious.

But communications is not computing.

There are similarities, both are dependent on hardware and software but this is where the similarity ends. The mechanics of connection are more complex than the mechanics of calculation. Telecommunication networks are dependent on signalling systems and protocols that took decades to develop and that absorbed thousands of man-years of international standardisation effort. SS7 signalling is one example.

Cellular networks also have to develop and standardise the radio access part of the network. Adding IP protocol to fixed or wireless networks introduces additional cost and complexity. It is no surprise therefore that the companies that have enjoyed the most success in the telecommunications industry and that are presently the most resilient are those that have managed to support long-term investments that realise long-term returns bearing in mind that the definition of long term in this context is thirty years or more. In Europe, that includes companies that have ownership and shareholder structures that give them a measure of protection against institutional shareholders focused on shorter-term returns.

Telecommunications companies in command and control economies similarly can find it easier to justify thirty- to fifty-year investment returns. Whether US companies can still do this must be open to question.

12.6 Roads and Railways and the Power and Water Economy – The Justification of Long-Term Returns

If we are defining telecommunications as just another transport system along side roads and railways and power and water utilities then it is reasonable to assume that the model of the last one hundred and fifty years will carry on for the next 150 years. Change occurs over decades rather than days and the return on investment is on a similar time scale.

There seems to be a growing assumption that universal broadband connectivity irrespective of how it is delivered will drive future economic growth. In common with other nation states, the US and the UK have ambitious plans in place to roll out next-generation networks that will allow us to work rest and play at a new level of efficiency and intensity, assuming that this is want we want or need to do. The release of the Digital Britain4 report in 2009 by the UK government articulated the ambition but was nonspecific on where the funding would be found.

12.6.1 Historical Precedents – Return on Infrastructure Investment Time Scales

There are many past examples where infrastructure investment has been considered as critical to achieving economic, social and political progress.

When technology innovation has coincided with an ability to raise capital and an appetite for risk then the change can be dizzyingly fast. The canals and railways of Victorian Britain were an extreme example. Over the next one hundred and fifty years, telegraphy, telephony and telecommunications grew out of a similar mix of invention combined with private and public investment. The return on that investment can be very long term. Passenger numbers on the railways for example are presently increasing in the UK. Most train journeys cross a nineteenth-century bridge or embankment.

Investment benefit time scales of this order are difficult to reconcile with conventional investment horizons. Railway planning is best done over thirty- to fifty-year time scales and telecommunications should arguably be the same.

Conveniently thirty to fifty years is coincident with the average working and earning life span, a fact that makes or should make infrastructure investment interesting and relevant to pension-fund investors. Australian pension funds have embraced this opportunity with some enthusiasm but most other countries remain focused on shorter-term returns.

12.7 Telecommunications and Economic Theory

Economic theory when applied to telecommunications unearths other puzzling contradictions and apparently underexploited opportunities. For example, it could be argued that speed equates to value. This of course depends on the fiscal cost of achieving the speed. This may include environmental cost – Concorde comes to mind. It also depends on the additional value achievable from additional speed. Canal journeys may take longer but may be more efficient and in some instances the net gain over faster alternatives, for example trains and lorries, may be greater. If the commodity being transported is increasing in value over time, the longer it takes to get there the more it will be worth.

But surely the canal analogy cannot be applied to the economic theory of telecommunications? The benefits of broader-band telecommunications must surely outweigh investment cost? Well, that depends on the economic scale of the benefit and the time scale required for a return on the investment. If the cost of delivering broadband to rural areas could be amortised over 50 years, effectively what happens with telegraph poles, then almost certainly a net gain will be achieved. If the cost has to be amortised over five years then almost certainly it will not.

This explains the present lack of appetite for private sector broadband investment. It also implies a need for a longer-term plan.

12.7.1 Telecom Politics in Malaysia – An Example of Long-Term Planning and Mixed Results

In 1991 the Prime Minister of Malaysia Dr Mahathir bin Mohamad produced a social and economic plan that would make Malaysia a self-sufficient industrialised nation by 2020 based on an annual growth of 7% between 1990 and 2020.

The building of a multimedia supercorridor stretching from Kuala Lumpur to Penang was promoted as a key enabler in the industrialisation process, alongside ambitious plans to upgrade the roads, railways and utility sector (power and water), transforming Malaysia into an Asian economic tiger.

The multibillion dollar telecoms plan was formalised in 1998 as a connectivity corridor 15 kilometres wide and 50 kilometres long, a Malaysian Silicon Valley to include two smart cities that would have cutting-edge communication and facilitate ‘electronic government’, the later being a touch sinister given Dr Mahathir's propensity to imprison political opponents.

Dr Mahathir resigned in 2002 after 20 years in office blaming the failure of the plan on the failure of the Malay community to adapt to an industrialising society and later stating that the education system had failed to deliver as well. In practice, Malaysia had been achieving year-on-year growth of 8% from 1987 up until the currency crisis of 1997, so blaming the Malay population might have been disingenuous though they may well have had well-founded doubts about the overall aims of the plan. Additionally, the policy was credible only as a result of positive Asian investment sentiment, which proved in the case of Malaysia to be unsustainable at least in the shorter term.

The policy did, however, achieve a noticeable if not radical cultural shift that included greater emphasis on electronics engineering education and a gender-neutral approach to engineering recruitment, an example that the UK and other countries could usefully follow.

12.7.2 The UK as an Example

Harold Wilson attempted something similar after the 1964 British election. The White Heat of Technology would rescue Britain from industrial decline. The British political process more or less guaranteed that this ambition would be frustrated. Autocracy may have been the missing ingredient here.

Forty five years on the question we seem to be asking ourselves is a reworking of the Wilson mantra – will the White Heat of Telecommunications Technology rescue Britain from industrial decline. After all, if Britain can be set on the road to recovery simply by putting in some broadband investment then just think what could happen elsewhere in the world. The answer of course is that investment in broadband connectivity is only part of the solution.

There is a parallel need to invest in the education of a new generation of telecommunications engineers and a need to invest in the R and D and manufacturing capabilities needed to bring new telecommunication technologies and techniques to market. These are needed in order to reduce costs and increase added value. This is needed to reduce return on investment time scales – a virtuous circle within a virtuous cycle. The same principles almost certainly apply to other national economies. The only difference is that it is challenging to create a nationally specific source of telecommunications engineering expertise from a virtually zero start point.

12.7.3 The Advantage of an Autocracy?

Taking Malaysia as a reference example, Dr Mahathir's political power base and initial lack of democratic accountability allowed for a number of ambitious projects to be undertaken of which telecommunications investment was one, but given that these were substantially financed by inward investment implied that they would be vulnerable to shifts in international investment sentiment that in turn would prompt political change and so it was. The pattern has been repeated in many countries since, both in the Middle East and in Europe (Portugal and Ireland).

Similarly, China is creating a potentially world-beating home-grown telecommunications R and D and manufacturing capability at a speed that is probably only possible in a closely coupled command economy. India could potentially do the same though political and industrial policy making in India is traditionally elliptic, which tends to slow things down.

At a recent conference in Cambridge one of the speakers5 with experience of selling into China and India pointed out that regulatory policy does not always align with political policy. In China, as an example, the political system remains communist, but the regulatory system is capitalist.

12.7.4 UK, US and ROW (Rest of the World) Implications

What does this mean for the telecommunications industry in the UK, US and rest of the world? It seems unlikely that either the UK or US will adopt autocracy as an acceptable method for accelerating technology investment, though given the state of present UK politics you can never be too sure. Liberal democracies are not immune to the occasional miscarriage of justice but a whole scale abandonment of the democratic process might be a step too far. The end does not always justify the means to the end.

Collaborative initiatives that bring common interest countries together are one alternative, but recessions tend to increase national self-interest – the car industry being a present stark example.

So we are forced back to seeing what we already have and whether we can make better use of what we have. Theoretically, the UK should have a legacy advantage. In radio communications for example, the UK was one of the first countries to implement digital mobile GSM networks, the first country in the world to implement digital broadcasting with the launch of DAB in 1995 and DVB-T in 1997, and one of the first countries to introduce digital two-way radio. The UK presently hosts the world's largest TETRA-based public safety network. The UK will be one of the first countries to implement DVB -T2 broadcasting and has a leadership role in LTE standardisation. A UK company directly supports nearly 300 million cellular subscribers internationally and has a track record of technology innovation that has delivered global competitive advantage. We have several UK companies that are world leaders, or at least potential world leaders, in radio and telecommunications silicon. The UK undertakes internationally acclaimed fibre optic research. The UK is surprisingly good at designing and making and occasionally launching low-cost satellites. Depressingly, this is probably not enough.

It is remarkable how easily national technology resource, the engineering collateral of a nation, can be squandered as a result of political paralysis. Political inefficiency may be the price that we pay for living in a democracy and if so we should just accept the fact that free will is not free but comes with a cost attached. Even discounting the direct and indirect ‘cost of democratic governance’, developed nation states have high running costs that are a consequence of quality of life expectations compounded by other factors such as a perceived need to defend themselves or a usually misplaced desire to invade other nation states. In this context the concept that any developed nation state can survive as a predominantly service economy must be open to question.

By implication the UK needs a broadband technology economy rather than a broadband service economy. A broadband technology economy should be intrinsically efficient at realising a return on legacy R and D and manufacturing investment.

So back to the assumption that universal broadband connectivity will drive future economic growth. This is almost certainly true but the problem is that we have no obvious way of getting to where we need to be. The return on investment is not sufficiently attractive to bring in private investment and public cash is not presently a viable alternative.

The UK has the advantage of a one hundred and fifty year legacy of telecommunications R and D and manufacturing investment. It therefore makes obvious sense to couple broadband connectivity investment policy with telecommunications industry investment policy. Countries such as Germany, Canada, France, Italy and the US are in a similar position. An economic model that somehow aligns and combines these interests is probably the best and possibly only way forward.

12.7.5 The Role of Telecommunications in Emerging Economies

In emerging countries, telecommunications has a different though no less profound role to play. The provision of microtrading and microfinance over cellular networks discussed in Chapter 11 is one example of the potentially transformative power of low-cost and not necessarily broadband connectivity.

12.7.6 The Telecommunications Economy – A Broader View of Broadband

It is therefore useful not to get overfocused on a broadband debate but to take a broader view of the role that telecommunications and the telecommunications industry can play in developed and developing economies.

Developed economies often have legacy telecommunications R and D and manufacturing investment that should be nurtured and nourished as a national resource. Telecommunications R and D and manufacturing in these countries is not the problem but the solution to the problem.

Emerging economies do not have this advantage but have the potential to make rapid economic progress with telecommunications as the essential facilitator – a telecommunications economy rather than a broadband economy. They can and do insist on in-country R and D and manufacturing as a precondition for overseas companies wishing to do business in the country. The Digital Britain Report referenced earlier justified broadband investment as a precondition for economic and social progress. The thinking in the report reflected a global debate on the contribution that broadband connectivity is expected to make to developed and emerging economies.

Irrespective of whether broadband is delivered over fibre, cable, copper or wireless or any mix of these, it comes with a high price attached. From an industrial policy perspective, the provision of broadband connectivity is an obvious business opportunity but only if the technology and engineering economics make sense. If the technology and engineering economics do not make sense it becomes an expensive obligation. The issue of course is whether wireless is more or less economically efficient as a delivery platform when compared with other options including fibre to the curb and fibre to the home.

This is in turn dependent on the definition of broadband, both in terms of average and peak data rates and contention ratios, the demographics and topology of the country into which broadband is being deployed and the efficiency with which wireless is integrated technically and commercially with other delivery options.

For example, fibre arrived in our street in 2009 as an upgrade to the existing cable network and we were duly leafleted with the offer of 50 Mbit broadband, all you can eat TV and telephone connectivity at £50 per month. Problematically for the fibre supplier, we have acceptable ADSL2 connectivity and minimal appetite for all you can eat TV already adequately supplied without a subscription fee via a £100 Freesat satellite receiver that we fully own and are still happy with two years later.

This highlights that even in urban areas in developed economies (the UK can just about still claim to be a developed economy) the time taken to achieve a return on broadband connectivity investment may not be consistent with short- or long-term shareholder expectation. Local zoning restrictions preclude cellular as a broadband access option even if we needed it, which we don’t.

12.7.7 Australia as an Example

Simultaneously, the Australian government decided the Australian nation should have its own national broadband network delivering 100 Mbit fibre to 90% of the population over the next eight years. The cost is an estimated 43 billion dollars and is billed as Australia's largest ever infrastructure project

The proposed funding is made up of an initial government investment of 4.7 billion dollars with the balance as a mix of 51% government bonds and 49% private funding. It is claimed that the scheme will create 37 000 jobs.

The cost equated to a public funding investment per household of 1650 Euros. New Zealand has a similar programme that equates to 600 Euros per household. The equivalent UK national commitment is three Euros.

Whether the Australian and New Zealand initiatives prove to be economically efficient or politically popular is always going to be dependent on how much of the promised funding materialised within the political election cycle, how well it is spent and the terms on which private sector funding is secured. The latter is the detail that has been problematic. As one example, in Brisbane a private sector group called the i3 group had a plan to start work on a new fibre optic network in early 2011 using the underground sewer system. Unfortunately, the company became embroiled in an investigation by the Serious Fraud Office following a failure to deliver on similar projects in the UK. So far public funds have been spent, private capital has been raised and a fibre optic network has failed to appear, not the most promising of starts for a new funding model designed to transform broadband connectivity. The promised jobs have appeared elusive as well.

These projects are perhaps better attempted in command and control economies with a more deterministic political purpose. Singapore for example has a government plan to spend 500 Euros per household and it can be imagined that failure there will not be an option.

The justification for these investments is based on an expectation of improved industrial, social and political efficiency with additional benefits such as lower carbon emission based on an assumption that the need to travel will reduce. The underlying argument is that emerging economies need broadband in order to become developed economies and developed economies need broadband in order to remain internationally competitive.

The combined investment, however, is likely to be of the order of hundreds of billions of dollars and it is fair to question whether a better overall return could be realised by providing connectivity to bottom-end rather than middle- (emerging) or top-end (developed) economies.

12.7.8 Universal Narrowband Access in Survival Economies

A bottom-end economy is a survival economy in which the average annual wage is less than ten dollars per day ($3260 dollars per year). Emerging economies have incomes between $3260 and $20 000. Developed countries have median incomes greater than $20 000.6

Day to day life in survival economies is dominated by the need to find food, water and shelter (though TV is popular too). Many of the four billion people trapped in survival economies do not have bank accounts and have no way to exchange goods and services other than through basic bartering or cash, which can be dangerous to handle. When a natural disaster happens, survival economies are dependent on food aid brought in by oversees agencies by truck or air. The cost per calorie of this food aid once administration, transport and security costs are included is ridiculously high.

The Safari Com M-PESA scheme described in Chapter 11 has been an effective antidote to these problems. administered through a network of 10 000 local agents. Transaction volume is presently doubling every four months. The host operator, Safaricom, enjoys low churn rates and income from six million SMS messages a day. The publicity from the scheme helped with a stock exchange flotation which included a significant percentage of small shareholders to whom Safaricom has just sent a five-dollar dividend by SMS transfer – a transaction that would have been uneconomic using traditional banking.

From a user perspective menu prompts are in English or Swahili. An unexpected side benefit of the project has been to increase basic literacy, numeracy and memory skills (remembering a PIN number).

In the postelection riots subscribers were encouraged by Safaricom to share prepay credit in order to keep communication going. The riots meant that substantial food aid had to be brought in. The distribution of aid as cash instead of food parcels was piloted with SIMS issued to families who were then sent small amounts of money to allow them to buy food and supplies.

This reduced the cost per calorie and targeting the aid where most needed. The process also helped to establish local supply chains and sustain the local economy. A similar initiative was then started in Afghanistan, where recent events suggest there were other applications that could be usefully developed. For example, the cost of providing international observers in the Afghan election was estimated as being of the order of $300 million dollars. This does not include the loss of life that occurred and the political cost of evidential fraud. Spending this money on mobile phones and mobile infrastructure to enable voters to vote by phone is a viable alternative for elections in countries with limited democratic experience and would deliver the added benefit of a sustainable microeconomy on the Kenyan model.

Healthcare is an additional potential application based on the integration of M Health and M Wealth (poverty eradication using access technology).

12.7.9 Universal Narrowband Access in Survival Economies – The Impact on Developed and Emerging Economies

So if we diverted universal broadband infrastructure spending in developed and emerging economies and invested the equivalent amount, several hundred billion dollars, into universal narrowband access in survival economies what would happen?

We would suggest that globally there would be a net social, political and economic gain. Survival economies would move more rapidly to become emerging economies, in turn creating new markets for the developed world to serve. Attempts have been made to quantify this. The World Bank estimates that a 10% growth in broadband connectivity delivers 1.3% of GDP growth.7

In the wireless telecoms economy it is evident that profits in saturated developed economies are flat at best. Being forced into meeting universal broadband access obligations by wireless or other means will introduce additional cost and minimal profit. It is plausible that the taxpayer will absorb some of this investment pain but this hardly seems credible in today's international political climate. Given the growing interdependency of countries around the world it seems odd that the potential for survival economies to be the new engine of global social, political and economic progress has been largely ignored.

Partly, this may be due to fear of the unknown. For example, there is an argument that suggests that growth in emerging and survival economies will trigger large increases in carbon emissions. There are equally strong counterarguments to suggest this would not happen and that other positive metrics would outweigh any related carbon risk. Economic growth and a low carbon economy do not need to be mutually exclusive.

Note also that survival economies are not geographically specific and can be found within developed and emerging economies. Over 70 million Americans do not have a bank account.

12.8 The New Wireless Economy in a New Political Age?

The present political enthusiasm for broadband connectivity investment may be misplaced and is certainly too rooted in nationalistic competitive ambition. In a closely coupled global economy, future profits and business opportunity will be dependent on how well emerging countries perform and how fast survival economies can become emerging economies. Fortuitously, it would appear that mobile phones and mobile networks are important and perhaps essential to this transformation process.

We may be witnessing the birth of a new age of truly international growth and prosperity with the telecoms industry as a facilitator of global social, political and economic progress – the new wireless telecom economy. The fact that telecommunications networks have reinvented themselves as information networks does not alter this fundamental truth but should reassure us that the telecommunications industry continues to have a socially and economically profitable purpose in both emerging and fully developed economies.

The argument therefore is that narrowband access to survival economies should take precedence over broadband investment in developed economies. This would not be naïve altruism but a pragmatic way to maximise the social and economic dividend from global connectivity investment. Present broadband investment proposals in developed economies are justified on the basis of delivering nationalistic competitive advantage. This creates additional distance between the rich and poor on the planet, an unsustainable model for future global growth.

Access technologies, fibre, cable, copper and wireless, can, however, potentially deliver social and economic progress based on social and economic equality, the precondition for achieving at least a measure of political stability. Connectivity could be more broadly defined and delivered as a superutility in order to realise a step function reduction in connectivity cost and energy efficiency.

12.9 Connected Economies – A Definition

Connected economies can be interpreted on several levels.

There is a close interconnectivity between developed, emerging and survival economies and a growing recognition that South America, Africa and Greater Asia will be or will continue to be the engines of future global prosperity providing fast growth markets for the developed world to serve. The 2016 Olympics in Brazil will be a reflection of this global transition.

Additionally, these emerging economies are becoming increasingly able to undertake local technology and engineering development. Over six million young people graduated from Chinese universities this year of which about half a million are engineering graduates – an engine of innovation. India similarly is producing 400 000 new engineers per year.

The availability and cost and effectiveness of engineering resource is closely coupled with the world's ability to deliver connectivity. Additionally, telecoms connectivity is dependent on the availability of civil engineering skills and an ability to source and use energy as efficiently as possible.

The telecom companies that trace their heritage back to the Victorian era, Cable and Wireless, British Telecoms and Siemens being three examples, have built their business on the back of multidiscipline skills that include mechanical, civil and energy engineering.

Most telecoms companies today regard a capability to source and use energy efficiently as a core part of the connectivity proposition for both rural and urban deployment. Connectivity is, however, not a term that is exclusive to telecoms but includes transport, water and power.

To put this in an historical context, steam and coal, steel and oil, the motor car, the aeroplane, the valve and the transistor have created and shaped the world that we live in today – the discoveries and developments, inventions and innovations that have provided the foundation for today's modern connected economy.

However, technologies always come with a social and environmental price tag attached. As a general rule the costs and the benefits are not equally distributed geographically of demographically – technology does not have a good track record for delivering social or economic equality.

So, should technology and the engineering needed to exploit technology have an explicit social purpose and if so how should that purpose be measured and managed? An answer rather than the answer to the question can be found by analysing technology history in the developed world. The Science Museum8 helps with the analysis process by dividing ‘recent’ (1750 to 2000) technology history into eight overlapping periods

12.9.1 Enlightenment and Measurement 1750 to 18209

Henry the Eighth, a very modern politician, though with a marital record that would have not stood up well to modern media attention, laid the foundation for the age of Enlightenment by spending a disproportionate amount of money on the British navy. This allowed Britain to build its colonial empire generating the wealth that went partly into an expanded University system.

Two hundred years later, that same University system delivered Isaac Newton to the world and a generation of thinkers with the time and money to consider how the world could be understood and therefore managed through experiment, measurement and reason.

Most things that moved and many things that didn't were measured including rainfall, deaths and electric charge, an age of scientific curiosity and instrumentation. However, this was enlightenment for the few not the many – an important shift in intellectual thinking but with limited impact on the poor and disadvantaged.

12.9.2 Manufacture by Machine 1800 to 186010

The era in which ‘machines to make machines’ produced the looms and spinning machines that powered the industrial revolution destroying the Indian cotton spinning industry in the process.

The industrial revolution resulted in a process of intense urbanisation and rural social deprivation. The net wealth of the nation increased but so did social and economic inequality. In parallel, the traditional education system failed to respond to the need for engineers. This was to sow the seeds of Britain's industrial decline in the twentieth century.

12.9.3 Age of the Engineer 1820 to 188011

This underlying problem was hidden by a small group of Victorian engineering superstars, James Watt, Robert Stephenson and Brunel.

12.9.4 Industrial Town12

Between 1800 and 1900 Britain's population more than quadrupled and became predominantly urban. Production machinery made it possible for factories to employ hundreds of people. Towns with high-density working populations created new problems of housing, health and transport.

12.9.5 Second Industrial Revolution13

The first Industrial Revolution had been based on steam power, factories and railways. The second was based on new kinds of power – electricity, new chemicals, new plastics and new drugs – particularly from recently industrialised nations like Germany and the USA. The USA introduced mass-production techniques into the weapons industry – a change which was to have a profound influence on the world's-twentieth century political order.

12.9.6 The Age of Mass Production14

By 1914 Henry Ford's car lines in Detroit showed the world that production and machine repetition could transform the economics of personal transportation. Mass production generated mass employment, mass consumption and a World War in which technologically advanced mass-produced weapons enabled killing and wounding on an unprecedented scale, the age of mass destruction.

Wireless telecommunications started to have a substantial social, political and economic impact.

12.9.7 Defiant Modernism15

The atomic age had resulted in the age of mass destruction being taken to a new level of deadly efficiency, but as a small compensation, the Second World War provided the basis for inventions and innovations that could be potentially repurposed to realise social and economic progress.

The V2 rocket provided the basis for the space programmes of the 1950s and 1960s, the atom bomb was supposed to result in nuclear power stations that would produce electricity ‘too cheap to meter’ and emissions ‘too small to measure’, aspirations that significantly failed to materialise.

12.9.8 Design Diversity 1950 to 196516

The availability of new materials and the dawn of the age of plastic provided opportunities to differentiate the design of everyday products in order to meet consumer expectations of choice, cost, quality and functionality.

12.9.9 Age of Ambivalence17

Organised environmental movements and environmental campaigns reflected a growing unease and failing confidence in the ability of technology to deliver social and economic progress at an acceptable environmental and social cost. This included concerns about genetic modification techniques.

In parallel, there was a growing recognition that the gap between rich and poor countries and the gap between rich and poor within those countries was getting wider over time both in terms of wealth and health – a recipe for economic, political and social instability.

12.10 Inferences and Implications

A number of general inferences can be drawn from this particular slice of technology history.

The social and economic benefits of technology and engineering innovation are generally not evenly distributed geographically or demographically nor are the direct and indirect costs. The direct costs are those associated with initial deployment and maintenance costs thereafter. The indirect costs are the social and economic costs that can be both short term, the loss of cotton spinning jobs in India being one example, or long term. Only now, 150 years later are we becoming aware of the full environmental cost of the industrial revolution.

By the end of the nineteenth century, Britain had become the world's most advanced connected economy based on a postal system, the telegraph, the railway, including the London Underground, a canal system, a steam-driven water system, the world's largest network of sewers, an underground hydraulic system, gas, electricity and a large merchant navy, a form of mobile connectivity.

This was infrastructure investment on a scale never seen before or since, built on the profits of the British Empire. Today, we still post letters in Victorian post boxes, travel on Victorian railways and the underground network, navigate and ship goods on Victorian canals, drink water delivered through Victorian pipes and flush effluent away through now leaky Victorian sewers. The water-based hydraulic system was retired in 1977 and the curtains of the London Palladium are now opened and closed using electric rather than water power.

A number of specific inferences can be drawn from this.

The infrastructure of a connected economy is expensive and has to be financed from somewhere else by someone else. The return on investment has a long tail extending to hundreds of years.

Telecommunications is only one of several bidirectional connectivity systems including the postal service with which it was successfully integrated for over fifty years. Water and sewage is of course bidirectional but modern electricity grids are also now being designed as two-way systems. There has been some debate as to the practicality of homeowners earning money from allowing passersby to access their home WiFi systems, an arrangement that could potentially result in an extra ten million access points becoming available. The snag here might be the occasional home owner who decided to increase power and uplink gain beyond conformance limits in order to attract more revenue.18

These caveats aside there may be merit in reconsidering the concept of the repurposed super utility. Superutilities were invented by the Victorians as a mechanism for delivering multiple services more efficiently. The Gas Light and Coke Company is one example; the General Post Office (Post and Telecommunications) is another.

Utilities are delivered in the developed world with bizarre inefficiency, the only possible justification being the employment provided to the trench-digging, hole-digging and meter-reading community.

It is inherently inefficient to have multiple organisations separately supplying electricity, gas, water, sewage and fibre, cable, copper and wireless connectivity.

If Britain was a dictatorship there would be a strong argument in favour of setting up two organisations, the British Broadband Corporation and the National Connectivity Corporation.

The British Broadband Corporation would take over responsibility for public service broadcasting from the BBC but would also provide broadband connectivity using fibre, copper, cable and wireless and would be responsible for developing new broadband and narrowband delivery technologies that could be applied into emerging and survival economies. The logo could stay the same as well.

It would have a social remit – ‘entertain, inform and educate’ would do nicely, a term apparently coined not by the BBCs Lord Reith but by the American broadcasting pioneer David Sarnoff in 1922.

The National Connectivity Corporation would provide all other forms of connectivity with a similar remit to develop connectivity technologies, particularly water and electricity, that could be applied cost efficiently and energy efficiently into emerging and survival economies.

Both corporations would be publicly owned but profit making with dividends only payable to pensioners. The two organisations would be mandated to develop mutually supporting collaborative delivery models.

This of course will not happen but does suggest that there may be merit in considering new more closely integrated ways of delivering water, electricity and telecommunications to remote communities in bottom-end economies.

12.11 The Newly Connected Economy

As in the Victorian age, connectivity efficiency is very much determined by how the energy needed is created and how efficiently that energy is used. In the nineteenth century water and sewage systems were powered by coal and steam. In the twentieth century telecommunications were powered and are still predominantly powered today from the electricity grid.

In the twentyfirst century, connectivity is much more likely to be powered from locally generated energy including wind and solar-powered base stations, and wind and solar-powered water pump and recycling systems all of which could potentially be community owned and managed. But let's consider the particular contribution that cellular phones can make to connectivity.

12.11.1 Connecting the Unconnected

By the time you read this the number of cellular phones in the world will equal the number of people (not counting the number of machine-to-machine devices that will be interconnected by then). This does not mean that everyone will have a cellular phone but is a reflection of the growth of multiple device ownership in developed economies. Cellular phones also need cellular networks or more specifically access to information networks in order to be useful.

However, achieving 100% penetration demonstrates that we have the industrial ability to provide a cellular phone to every human being on the planet and the ICT ability to deliver socially and economically useful personal and individual services to those phones.

This industrial and ICT capability is not concentrated in a single country or continent but spread across Europe, Greater Asia, the US and South America, a globally distributed deployment resource. China for example has very successfully leveraged its local market both to build home-grown vendors and attract inward intellectual investment in joint R and D programmes. Huawei in 2011, a company few people outside China had heard of ten years before, had a turnover of $28 billion, employed 110 000 people, half of them in R and D and had 5000 staff in Europe.

The cost and energy efficiency of deployment will be dependent not only on the capability and resources of these newer vendors and their older competitors but also on how well these services are integrated with copper, cable, fibre and other wireless delivery systems and the best social and economic returns may be from narrowband rather than broadband investment but without a doubt the modern world has sufficient technology and engineering and financial resource to connect the unconnected provided we have the political will to do it.

12.11.2 The Dawn of a New Age of Enlightenment?

Developed nations have visibility to a plausible model for providing affordable and sustainable informational, educational and economic connectivity to the survival economies of the emerging world, delivered physically side by side with water and power connectivity – a light at the end of a dark tunnel of technology-driven social and economic inequality, the potential dawn of a new age of engineering-based personalised mass enlightenment, enlightenment for the many not the few.

12.11.3 Enabling Wireless Technologies – Femtocells

And then I noticed that what I should really have been talking about in this chapter was mobile phone network topologies, but then this prompted me to look at what we had covered in Chapter 12 of 3 G Handset and Network Design only to discover that not only had this topic already been fully addressed but also very little had changed in the intervening eight-year period.

Base stations had become smaller, smart antennas had become a bit smarter, integration with WiFi networks had progressed a bit but not a lot.

For at least the last five of those eight years femtocells have been promoted as the new panacea. These devices never made much sense from a user-experience perspective. A femtocell in your home promised access speeds rather lower than the WiFi already installed that you now could not use as it was providing the backhaul for your new femtocell. The femtocell was available with only one operator providing the service so if other members of the family had service contacts with other operators, additional femtocells were needed that would then produce adjacent channel interference with other femtocells. Meanwhile, WiFi access points became cheaper and faster.

And this is probably quite a good though slightly negative point to end this chapter. Innovations only become widely adopted if they deliver added value to all players in the value chain. In the case of femtocells, they solve an operator problem (poor indoor coverage from outdoor cells) but as far as the user is concerned, why should that involve fixing a problem that is clearly the responsibility of the operator who already takes a large monthly fee but now appears to be expecting the consumer to provide free backhaul as well.

This brings us to our next chapter.

1 http://www.makingthemodernworld.org.uk/icons_of_invention/technology/1939-1968/IC.059/.

2 http://www.makingthemodernworld.org.uk/icons_of_invention/technology/1968-2000/IC.110/.

3 http://www.ericssonhistory.com/templates/Ericsson/Article.aspx?id=2095&ArticleID=1378&CatID=362&epslanguage=EN.

4 http://webarchive.nationalarchives.gov.uk/+/http://www.culture.gov.uk/what_we_do/broadcasting/5631.aspx/.

5 Mr Kanwar Chadha speaking at the Future Wide Area Wireless Conference 27 January 2011.

6 http://www.wri.org/.

7 As reported by Bong Goon Kwah, Korea Telecom Cambridge Wireless Conference, 27 June 2011.

8 These extracts reproduced with permission of the Science Museum http://www.sciencemuseum.org.uk/.

9 http://www.makingthemodernworld.org.uk/stories/enlightenment_and_measurement/05.ST.05/.

10 http://www.makingthemodernworld.org.uk/stories/manufacture_by_machine/02.ST.01/.

11 http://www.makingthemodernworld.org.uk/stories/the_age_of_the_engineer/03.ST.03/.

12 http://www.makingthemodernworld.org.uk/stories/the_industrial_town/06.ST.02/.

13 http://www.makingthemodernworld.org.uk/stories/the_second_industrial_revolution/05.ST.01/.

14 http://www.makingthemodernworld.org.uk/stories/the_age_of_the_mass/.

15 http://www.makingthemodernworld.org.uk/stories/defiant_modernism/04.ST.02/.

16 http://www.makingthemodernworld.org.uk/stories/defiant_modernism/04.ST.02/.

17 http://www.makingthemodernworld.org.uk/stories/the_age_of_ambivalence/02.ST.06/.

18 Open forum discussion, William Webb, Future of Wireless Conference, 27 June 2011.