TV switch puts radio spectrum on the radar
The following contributed article - written by InternetNZ policy staffer Reg Hammond - was published on Stuff.co.nz on Tuesday 20 November 2012.
Radio spectrum is like air - it is largely invisible and we only really pay attention to it when it starts to become scarce.
We notice it most when we cannot listen to the radio, watch TV or make mobile phone calls. As most of us are able to do all those things (I will come back to the few who miss out shortly) radio spectrum issues generally remain off our radar.
We are relatively blessed in New Zealand that the useful radio spectrum within our geographic boundaries is shared between just 4.5 million of us. The same amount of useful spectrum in the UK, a country approximately the same size as New Zealand, is shared by 60 million Brits. They have lots of close neighbours who can potentially share their spectrum as well.
Successive New Zealand governments over the past 20-plus years have talked of the efficient use of spectrum and the need for competitive auctions of spectrum management rights. They have seen this as the best means of allocating a scarce spectrum resource because it incentives those with the greatest perceived need to purchase the spectrum rights at auction - for up to 20 years.
It also allows the purchaser flexibility to "change horses" mid-stream in that they can on-sell the management right or use the spectrum for a different purpose should technology change.
One aspect of radio spectrum allocation has in recent months been the centre of attention - a strong blip on the public's radar screen. Several other aspects have been but peripheral shadows seen only by those paying close attention. This article attempts to "connect some of the dots".
The strong blip on the radar is the switch to digital television. Each evening, we are all reminded of the impending switchover and that if we want to continue watching Shortland St, Fair Go and more adverts then we will need to dump the old TV and go digital. For many of us, it means having to purchase a new TV and possibly a new antenna. In return we get better quality pictures and more channels.
Thanks to the wonders of digital technology, digital switchover means those additional channels along with all the existing ones use far less spectrum than before and so makes available a considerable amount of very useful spectrum (112 MHz of spectrum in the 700MHz band) available for new uses. So does this mean that useful spectrum is no longer scarce? Yes, and no.
Cellular companies (Telecom, Vodafone and 2degrees) will tell us that the growth in demand for mobile data is exponential and that the only way they will be able to satisfy that future demand is to let them have the spectrum in the 700MHz band.
It's true that globally the growth in demand for mobile data is exponential and America's 320 million citizens or China's 1.4 billion or India's 1.2 billion may struggle in future to get a coverage bar on their cell phones.
What New Zealand's mobile companies don't tell us is that they have considerable amounts of useful radio spectrum already - much of it unused or underused. It is not the same for us - New Zealand is blessed. There are only 4.5 million of us - better still we are spread out, even in Auckland - so the same frequencies can be used at the same time in different locations.
So why would the Government, knowing that there was no real scarcity, start a process that first freed up the 700MHz spectrum band to then make it available to those who already have surplus spectrum anyway? And, why would those with the surplus spectrum want more (we can probably leave 2degrees out of the argument here as they have rights to significantly less spectrum than others) particularly if they might have to bid millions or tens of millions of dollars to get it?
There are a number of reasons which deserve detailed explanation but for now can be summarised as: in the case of the Government, a mistake from the past that has been perpetuated and the lack of an agreed strategic plan.
In the case of the cellular operators it is rational, commercial behaviour. First, by creating and maintaining an artificial shortage of spectrum the cellular operators are able to keep lower cost competitors out of the market - 2degrees will certainly attest to this. Secondly, and particularly in the case of the 700MHz band, it has attributes that make it far better for mobile cellular use than their current spectrum holdings i.e. they will be able to roll out more profitable services at lower cost to them - not wholly a bad thing as consumers will get some improved service benefits as well.
Let's now go back to my first paragraph and deal with those who may not be able to watch terrestrial TV, make cell phone calls or get mobile data - remote rural communities. This small, but important, part of the New Zealand population has oodles of unused spectrum buzzing round its ears - but what it doesn't have is a cellular company prepared to provide a service over that spectrum. Worse, because the mobile companies own the management rights they are within their rights to refuse to allow any other company to use the unused spectrum to provide a service.
The Government, with some foresight, recognised at least part of this problem when developing its plans for the Rural Broadband Initiative (RBI).
It recognised that rural communities would require access to good broadband but that the massive cost of supplying fibre to remote and sparse rural communities was impractical. It also recognised that future 4G wireless technologies, with limited numbers of people using them, as would be the case in remote areas, were capable of deploying good broadband services at a fraction of the cost of fibre.
The contract with Vodafone and Chorus to roll out the RBI will see a significant percentage of rural New Zealand given access to mobile broadband, particularly if Vodafone is able to acquire a sufficient block of the 700MHz spectrum which has especially good attributes for rural use.
There is little doubt that Vodafone will have the incentives to acquire a sufficient amount of the 700MHz band and many rural users will undoubtedly benefit. However, there are two blips on the radar screen - problems, both of which are able to be solved.
The first problem is the remaining remote communities that fall outside the RBI coverage area. The second is the perpetuation of the artificial scarcity of spectrum which reduces competition overall and leads to higher prices. The first of these problems is relatively easy to solve; the second is more difficult.
Taking the easy one first, there is plenty of unused spectrum available for the use of remote communities, including, in the future, the 700MHz spectrum band. It is just that the costs of deploying the network infrastructure are disproportionate to the amount of business that will be generated. In these more remote locations the network costs rise exponentially and the returns diminish - there is no business case.
There are two or three possible solutions and much depends upon who pays for them. Who pays will largely depend upon the outcome of the 2013 review of the TSO. For the purposes of this article however we can assume that the status quo will apply which means the end user receiving the service will have to pay. This leads to a choice between the existing satellite broadband service which can be considered the baseline in terms of cost and possible new services that could be leveraged off the RBI roll out.
The first leveraged solution would be for Vodafone to allow remote communities (or small regional operators) to install their own local network infrastructure and use unused spectrum or white spaceen with Vodafone providing the backhaul and national coverage beyond the local network. There would potentially be benefits to Vodafone (or other spectrum/network owner) that might persuade them to consider doing this voluntarily.
If they were not prepared to allow their spectrum to be used they could be required to do so as part of the tender process for the 700MHz spectrumen or separately via regulation.
The second leveraged solution would be for local communities to establish a fixed wireless or Wi-Fi network using a rural school that had been supplied with fibre under the RBI as a base station and source of backhaul.
Again, there are some potential difficulties that would have to be overcome and InternetNZ has joined the 20/20 Communications Trust to look into these as the possibility of a remote school sharing its own broadband costs and services with its community is well worth exploring.
Let's go back to the more difficult problem of the artificial scarcity of useful spectrum generally reducing competition.
The origins of the problem lie back 20-plus years with the introduction of the spectrum managements rights regime. At the time it seemed a sensible solution, allowed a market price to be put on a resource, helped the government to avoid having to pick winning technologies and gave some certainty to cellular operators that the significant investment they would have to make to deploy cellular networks would be returned over time.
Early problems arose - for example, Maori considered that if property rights were being allocated they were entitled to their share - a problem that has persisted and which is now holding up the allocation of the 700 MHz band.
The next problem was the belief that by allocating the spectrum rights via competitive auction it would see the spectrum awarded to those that valued it the most and who would use it most efficiently. The combination of those two assumptions was a mistake. Those that valued the spectrum most (at least those who paid the most for it at auction) did not necessarily have any thoughts about efficient use they simply wanted exclusive use.
There are solutions. Technology now exists to make exclusive management rights unnecessary. Three examples:
Mobile cellular technology has historically been based upon frequency distribution duplex (FDD). It was developed along the same lines as analogue telephony with fixed, equal amounts of bandwidth on defined frequencies being allocated to the send (talk) and receive (listen) components of a phone call.
The increasing trend away from phone calls and towards data and the differing requirements of data (an individual will usually download much more data than they upload) means that allocating constant equal amounts of bandwidth to the send and receive functions is very inefficient. It's far better to allocate bandwidth dynamically - allocating the bandwidth at the time it is required to either the send or receive functions.
This time division duplex (TDD) i.e. allocating sufficient bandwidth at the time it is needed is far more efficient for data and doesn't require operators to have exclusive use of frequencies - they can share.
Cognitive Radio is a technology that allows the handsets of users to monitor the network environments they are in and identify the best parameters, such as frequencies, to use. Again, allocating frequencies for the exclusive use of rights holders is unnecessary to the establishment of communications - the equipment simply chooses the best frequencies to use at a point in time in order to avoid interference with other users.
Wi-Fi - although not new technology, this is a great example of a technology that operates very efficiently outside the management rights regime.
It is also very popular and low cost. So popular that iPads, tablets, smart-phones and many other devices increasingly have Wi-Fi or dual Wi-Fi and cellular capability. The costs of deploying Wi-Fi are low enough that coffee shops, hotels and even cities such as Wellington allow you to use their Wi-Fi freely or in exchange for viewing an advert or buying a coffee. This low cost, availability and easy access has its downsides and busy hot spots become congested. A solution to this congestion again could be simply the allocation of more spectrum bands for Wi-Fi.
InternetNZ has proposed relatively straightforward solutions to these problems. First, allow Vodafone, Telecom and 2Degrees, if they wish, to purchase the 700 MHz spectrum band management rights. As part of this arrangement Telecom and Vodafone would have to release other underused spectrum which would be removed from the management rights regime. They would also have to allow open access to their tower infrastructure much like Vodafone has already agreed under the RBI contracts.
The released spectrum would be made available for use of alternative technologies such as Wi-Fi, TDD, or cognitive radio that could share bands of spectrum without the need of exclusive management rights. Maori, the police, Civil Defence, small fixed wireless operators could all operate in these bands the only major requirement of them being that they do not cause interference with other users.
These technologies would bring much needed innovation and competition into the New Zealand market - competition from different technologies and also different operators. New Zealand has long promoted itself as a communications laboratory and with our relative abundance of spectrum New Zealand was, and still could be, an ideal environment to test new communications technologies.
This is only going to happen if there is an agreed radio spectrum allocation strategy as the Minister, Amy Adams, identified in her foreword to the spectrum allocation five-year outlook. These technologies and the companies that will deploy them are, for the most part, going to take five years to become established.
They need long term certainty that the spectrum they require will be available and not sold for exclusive use to the highest bidder. They need certainty that the equipment they and their customers will purchase will be available and will operate in the allocated bands. They need certainty that they can use established network infrastructure at reasonable cost. This all requires long term planning and an agreed long-term strategy between government, operators and users that is cognisant of the international trends.
This is currently the faintest of the blips on the edge of the public radar screen - it urgently needs to be brought front and centre. The current delay in allocating the 700MHz spectrum provides us with an ideal opportunity to develop and agree that strategy.
Reg Hammond is policy consultant for InternetNZ, private consultant and former manager of ICT regulation at the Economic Development Ministry.