In our view, one of the most important factors which will determine whether 5G will be a commercial and socio-economic success is whether it will trigger the blowing apart of the tight circle of the mobile industry to date.
This has been much discussed in terms of the supplier ecosystem. The combination of a virtualized, automated platform based on white box hardware, with the efficient and agile 5G radio, should allow mobile operators finally to achieve their long-held dream of introducing a wider variety of vendors into their domains, in a multivendor architecture with significantly reduced upfront and operating costs.
But in opening up the platform and the supply chain, MNOs may also expose themselves to new competition. 5G is hard to justify if its only business case is the initial one – driven by the MNO themselves – of enhancing fixed and mobile broadband. 5G must fulfil its other promises, of enabling new connected services with demanding performance requirements in terms of agility, latency and availability.
But MNOs may not be best placed to build those specialized networks to support cities, enterprises and the Internet of Things. They are struggling to work out the business case, and in turn, some industries express doubts that the traditional operators are the right partners, and would rather work with integrators and service providers focused on their own sectors.
Some of the key elements of 5G, such as a greater focus on very dense, localized, edge-oriented networks, potentially deployed in shared spectrum, offer the wherewithal to support specialist vertical service providers, and new entrants such as neutral hosts or city integrators. And increasingly, governments are interested in these developments, as pressure grows on them to ensure that 5G delivers broad socio-economic benefits such as smart transport, smart cities and reduced digital divide – this is especially true, of course, is those governments are making any direct or indirect investment in 5G infrastructure.
So will anything actually change? Some of the hopes of an open platform, enabling new suppliers and service providers, were aired in the early days of 4G, but despite some potential enablers, like neutral host small cells and dynamic spectrum systems, the entrenched habits of the large MNOs, and of most regulators, meant that these hopes were mainly dashed.
Networks remained in the hands of a few end-to-end suppliers; large operators continued to win expensive, long term exclusive licences which precluded new entrants; disruptors and new providers either had to secure massive financial backing, as Reliance Jio did in India, or rely on unlicensed spectrum technologies like WiFi and LoRa, which do not support all of the high reliability, high security, high mobility aspects of 4G and 5G.
The first prerequisite of 5G being an enabler of new entrants is that regulators support a more creative approach to spectrum licensing, and that MNOs recognize that this could present them with opportunities as well as competitive risks. At Mobile World Congress this year, it was noticeable that most of the large MNOs, in their keynote and panel addresses, were making the same demands for exclusive long term licences that they did a decade ago, with few concessions to the discussions about 5G being based on a far wider range of spectrum bands and sharing schemes.
In the higher bands, from 3.5 GHz to millimeter wave, there will enough spectrum to go around more players, and these airwaves are well suited to supporting local or vertically specific networks – dense, high capacity zones for a city or business park perhaps; or specifically optimized systems for a particular enterprise, vertical or transport route.
The large operators may battle against this valuable spectrum being carved up among more companies, but a few are starting to consider the upsides too. In a recent survey about 5G intentions conducted among 60 large MNOs, by Rethink Technology Research, about 15% said they would favor a model in which they would take responsibility for the nationwide mobile broadband network and specialist companies would shoulder the burden of building networks for specific industries.
After all, most MNOs have failed to find a convincing case for delivering critical enterprise services directly in the past. Instead, they may look to support an expanded wholesale model, in which they take revenue from wide area roaming (via MVNO deals with the specialists) and from providing higher layer services from cloud-based device management and security assurance, to full network-as-a-service or network slicing platforms.
To be sure, 15% is a small percentage and most MNOs believe that the expanded, dynamic wholesale model – enabled by shared spectrum, cloud platforms and slicing – is too complex or futuristic to model. But the indication that some progressive MNOs are accepting their own limitations in terms of vertical and local services is encouraging.
The next step will be to convince spectrum regulators. Only a few ‘5G’ auctions have been held yet, mostly in the C-band (3.4-4.2 GHz or subsets), plus the USA’s 600 MHz incentive auction – though some operators will deploy early 5G in repurposed 2G, 3G or even 4G bands. But so far, most of those auctions which have taken place look very similar to those of the 4G era, with a few, expensive national licences being snapped up by the established MNOs.
The UK’s recently concluded C-band auction is a case in point. The country’s four MNOs were the only bidders apart from one new entrant, Airspan, which is bidding for spectrum in various countries to support a neutral host, small cell platform play – exactly the kind of approach which would help open the market to the kind of new, vertically oriented services and providers outlined above. Although Airspan did secure 3.5 GHz spectrum for this purpose in Ireland, it dropped out quickly in the UK race, in an auction which pushed prices higher than expected, inevitably favouring the established MNOs.
The auction, which included some 2.3 GHz airwaves – mainly used for supplemental downlink in LTE – alongside 3.4 GHz, saw Telefonica’s O2 UK subsidiary emerging ahead of the pack, while Hutchison’s Three gained the least.
Regulator Ofcom was offering 40 MHz in 2.3 GHz, and 150 MHz in 3.4 GHz. The total raised was £1.356bn, more than double the pre-auction estimates of about £630m – good for the Treasury, but less so for a would-be new player like Airspan.
In the event, O2 got all the 2.3 GHz spectrum on offer, plus 40 MHz in the more prized 3.4 GHz band. Market leader, BT’s EE unit, and Vodafone, both also got 40 MHz in the 5G band, but Three got only 20 MHz. O2 paid £5.147m per MHz for the 2.3 GHz assets and £7.943m per MHz for its 3.4. The others paid £7.565 per MHz in 3.4 GHz.
Telefónica UK’s CEO, Mark Evans, said in a statement: “The airwaves we’ve secured allow us to further enhance our network, both now and in the future. We’ve thrown down a major marker for our future commitment to the UK. Our investment in 3.4 GHz enables us to move forward to further improve connectivity whilst boosting the economy and laying the foundations for 5G in Britain.”
Three, which had engaged in legal action to try to get more stringent caps on BT’s holdings, has not succeeded in redressing its 4G spectrum disadvantage, though it does have existing C-band assets, via the 40 MHz of 3.4 GHz, and 84 MHz of 3.6-3.8 GHz spectrum, which it acquired with UK Broadband. That highlights another source of hope for new entrants. In many countries, some 3.4-3.6 GHz spectrum was auctioned for broadband wireless access, often at low prices, and was often secured by local ISPs to deploy WiMAX or TD-LTE. No doubt these owners, like UK Broadband, will be aware of the rising value of their holdings, just like the companies which own once-unloved millimeter wave spectrum in bands like 28 GHz. But they may still be a lower cost source of airwaves for new players than the auction process, especially if that does not change far more radically than has been seen in the UK and elsewhere.
“The outcome for Three will do little to improve its precarious market position,” said Kester Mann of analyst firm CCS Insight. “Having campaigned tirelessly for more favourable conditions, it was surprising not to see it spend more. Three remains sub-scale and without fixed-line assets in a market gradually moving towards multiplay services and today’s outcome will do little to dampen doubts over its long term future.”
All true, but it demonstrates how often government’s desire for a quick income boost from auctions outruns the objective to expand mobile services for industries, transport and consumers – spending on a massive scale remains the only way to gain advantage, outside of countries, like Finland, which often conduct beauty contests as well as conventional auctions.
“That operators paid higher than expected prices is good for HM Treasury but not so good for consumers, as it leaves the operators with less money to invest in their 5G networks and services. In today’s market, operators are capex constrained,” Bengt Nordstom of telecoms consultancy Northstream told Mobile Europe. “It would have made more sense for Ofcom to agree a set 5G licence fee with each operator, which would then leave them with more money to invest in their networks. The results of the auction confirm our view that the mobile industry has passed the stage where we should expect new entrants in national markets to build and launch new mobile networks that compete with the incumbents. This in turn makes the idea of a competitive spectrum auction even more strange. It was a given that the UK’s four incumbent operators would win the 5G licences – therefore why would Ofcom drive up the prices?”