Read this article to learn:
- How siting issues could be the biggest challenge for MNOs in densifying cellular networks
- What needs to be overcome in order to make densified network economics viable
- Why MNOs are migrating capex from spectrum acquisition to small cells
- How MNOs see the role of traditional infrastructure in the development of distributed networks
Caroline Gabriel is the Co-founder and Research Director of Rethink Technology Research; Senior Contributor, Analysys Mason, Next Generation Wireless; and a member of the Small Cell Forum.
Densification in 2017 is no longer just an ugly word, but an ugly business challenge. It is now a real and urgent part of many cellular operators’ strategies to turn the inexorable rise of mobile traffic into profits. A few MNOs in Japan and Korea, blessed with plentiful fibre and access to the latest technologies, have blazed the trail. But for most operators, there are far bigger hurdles in the way of building massive capacity, in targeted areas, into their networks, and so supporting new business cases and laying foundations for 5G. The essential technologies – various small cell form factors, effective self-optimizing network (SON) tools and so on – are falling into place rapidly. The bigger barriers lie in logistics – reducing the red tape around site acquisition and leasing; streamlining the deployment process to be less labor-intensive and more replicable; automating as many operational activities as possible.
The operators have grumbled about these issues for years, but densification was not yet necessary outside a few select environments such as transport hubs, where DAS (distributed antenna system) would often do the job. Now MNOs, especially in the US, are acknowledging that a layer of heterogeneous small cells, spanning licensed and unlicensed spectrum, is critical to their revenue model. They are nearing the capacity ceiling in some areas, affecting the consistent quality of
experience which is the main way that high value customers choose their carrier. Traditional ways to increase that capacity are non-economic. The US operators have spent vast sums on acquiring new spectrum in the past decade, via auctions and M&A, and they have one more big sale to come, in 600 MHz. But that is not a capacity band, and with ARPUs under pressure even in the US, splashing out on airwaves is no longer attractive.
This is why all four national operators are now committed to small cell densification. Verizon’s CFO Fran Shammo said last year that it would be cheaper to achieve the capacity required for a major city like New York by using small cells and Wi-Fi, rather than acquiring additional spectrum. Meanwhile, AT&T said its macro network is currently about 90% of the size it will be in five years’ time, but the carrier has only deployed between 5% and 10% of the small cells it expects to need by then. Sprint CFO Taek Robbiati told a JP Morgan conference last year: “There are not enough towers in the country to do what we intend to do to densify the Sprint network, period. So, you’ve got to think differently. And really, there is great benefit around this, because the cost of deployment of cell sites is a fraction of the cost of the deployment of a tower site.”
He says a traditional tower site costs “probably a couple of hundred thousand dollars”, including labour and equipment, while small cell sites are “60%, 70% lower”. Sprint’s small cell deployment partner Mobilitie told the Wall Street Journal that building and operating the small cells “costs about $190,000 over 10 years, whereas a traditional tower costs $732,000 because of real estate rents, power and other costs”. However, given the large numbers involved, those costs clearly need to go down a lot further to make Sprint’s target of 70,000 small cells by late 2018 commercially viable. The notion of industrialization – a fully automated, rapid process which can be replicated anywhere – is central to the business case. But this shift will require a radical change in the regulations around sites – currently it can take up to two years to deploy outdoor small cells, said Shammo, “once you get a location, you negotiate with the landlord, you get the fibre to that location, because every single one of our small cells has fibre backhaul to a macro cell”.
The US is just one example. The issues are seen almost everywhere, and need to be addressed urgently. Brendan O’Reilly, CTO of Telefonica’s O2 UK arm, says MNOs in Japan, where some flagship dense HetNets are seen, can access four times more spectrum than their UK counterparts, and far more dark fibre at an affordable cost. He said Japan has also devised a more flexible regulatory environment that allows a new site to be acquired and a cell deployed in under three months. Nor is this just an issue for operators in developed mobile broadband markets whose customers’ usage is outrunning their macro layer build-out. Some MNOs in developing markets are taking a small cell-first approach to urban and enterprise deployments, but there are barriers there too. India is one country where, in some ways, site challenges are getting worse. The Municipal Corporation of Delhi (MCD) recently stopped granting permission for new towers in residential areas, and this may affect some small cell sites too.
The site issues could have a serious negative impact on MNOs’ business plans in most regions, preventing them from keeping up with capacity and QoE expectations, or supporting MVNOs such as cablecos effectively. In the cable case, especially in the US and parts of Europe, that may be an opportunity for Wi-Fi-first models to be accelerated; but for MNOs, while Wi-Fi offload is a well-established way to boost capacity affordably, in the dense outdoor environment this comes with a host of quality issues. Wi-Fi-first is fine for a mass consumer model, and for disrupting established cellcos with a low cost proposition, but for operators which need to chase higher value business such as mobile enterprise, it has as many challenges as small cell siting does (and if new, dense, carrier-class Wi-Fi build-out is required, the site and backhaul issues are the same).