Mobile infrastructure sharing is on the rise, whether encouraged by regulators or adopted willingly by operators to reduce costs. Probably the largest tower sharing deal in the world is now on the cards, between the three Chinese mobile providers, and it could even be extended to base stations. This could significantly improve the economics of 4G rollout for the operators, especially the smaller ones, but squeeze equipment providers if the number of cell sites is reduced by as much as 25% (as some calculate).
Governments becoming positive about sharing:
Progress will be closely watched in large and small economies, especially those where data charges are stagnating and mobile penetration is approaching saturation. In such climates, it is increasingly hard to justify the cost of investing in next generation mobile networks, and that has persuaded many regulators round the world to shift their attitudes towards infrastructure sharing. While most have allowed, or actively encouraged, the sharing of passive equipment (towers, power supplies and so on), the sharing of the active base station network has often been regarded as a threat to competition and differentiated services.
However, now it may be the only way that an operator can be persuaded to invest in new networks and so, in order to stimulate 4G roll-out while making the business case more attractive for carriers, governments and regulators are easing their restrictions. Many, like the Chinese authorities, are now actively encouraging sharing. Some go further and support sharing of spectrum as well as infrastructure. In Kenya, Russia and other countries, there have been schemes, of varying levels of success, to create a single pool of spectrum and a wholesale network co-owned by all the mobile providers.
Other governments see network sharing as a preferable alternative to outright consolidation. Operators, especially in European markets, are increasingly looking to merge with one another as competition intensifies, but antitrust agencies remain wary of reducing the number of players in a market. Network sharing may be a useful compromise in some markets, such as the UK (EE/3UK and Vodafone/O2), Spain (Vodafone/Orange and Telefonica/Yoigo), France (SFR and Bouygues are negotiating) and Czech Republic (Telefonica and T-Mobile).
The rise of wholesaling:
The typical approach is to place the shared infrastructure, whether towers or access networks (and occasionally spectrum or core network), into a joint venture which is coowned by the operators but run autonomously, often by a managed services provider such as Ericsson. The logical extension of this pattern would be for mobile operators to stop owning infrastructure at all, merely leasing capacity from utility wholesalers. To some extent, this already happens with towers, in markets where the dominant owners are dedicated firms such as Crown Castle.
As operators adopt small cells – tiny, low power base stations often mounted on a house or lamp-post – they will often avoid the complexities of agreeing so many site deals by turning to a wholesale 'small cell as a service' provider, which will build a single network to be shared by all the mobile carriers. That trend could easily spread to all forms of base stations, following the classic pattern of utility businesses such as gas and electricity, which eventually move towards a single wholesale network supporting many retailers. This process could be accelerated in the mobile world by the increasingly dangerous chasm between the amount of new data capacity required, and the amount of incremental revenue that can be generated from that capacity.
Cost reduction is the main motive:
For now, though, these more extreme approaches to network and resource sharing are seen as futuristic in most markets, and real world deals are focused on the potential economies of conventional passive and RAN infrastructure sharing. In the case of China, tower-related spending – according to Morgan Stanley calculations - accounts for 15% to 25% of mobile capex, and may tend to the high end of that range, given the rapid buildout and density of 3G/4G in China, in high spectrum bands requiring many cells (2.1GHz, 2.3GHz, 2.6GHz). The largest carrier, China Mobile, has about 380,000 towers supporting almost a million cells, though its smaller rivals, China Unicom and China Telecom, have fewer, especially in rural areas. They could cut as much as 20% off their 4G deployment bill with tower sharing, and even more with base stations added to the mixture, while Mobile could unlock $30bn in asset value if the venture were to go public.
The Chinese deal is extreme, but the savings from infrastructure sharing can be considerable even in smaller markets. SFR, France's second mobile operator, expects to save €200m from its proposed deal with Bouygues, though this is being challenged by incumbent Orange.
Cost reduction is the primary initial objective of RAN sharing. When existing networks are combined, operators typically look for savings of about 25% or more on opex. When the partnership is primarily geared to new or significantly expanded networks, the initial measure of success is spending prevention – lower capex outlay than anticipated on the new infrastructure. This is commonly targeted at about 30% (per operator), as is the reduced operating cost of the shared RAN. About half the opex savings come from site costs but moving beyond passive sharing promises far higher savings, as well as greater complexity and regulatory scrutiny.
Fig 1. The primary objective of a RAN sharing program, and the top three objectives. Cost factors dominate among operators engaging in RAN sharing or considering it, but only when those savings are harnessed to improve service quality and competitiveness. Source Rethink/Amdocs survey.
Other benefits:
However, operators are increasingly looking for additional benefits, on top of simple cost reduction, to justify the risk and upheaval of entering a RAN sharing pact. They know that they have to adopt new business models to make a profit from LTE investment, and they want sharing deals to help enable changes such as more complex charging structures - including premium rates for guaranteed QoS (quality of service) - and brand new offerings such as machine-to-machine services.
These moves will depend on a network which is planned and managed flexibly to deliver QoS to the right users at the right times, and to maximize the capacity in any piece of spectrum. While coverage is becoming noticeably less important as a differentiator than in the early 3G days, competitive edge now revolves around services – quality, speed and range.
That means capacity is important as a reason to combine networks and is a top three factor for one-third of operators considering RAN sharing, and the main motivator for onefifth. Reducing time to market for new networks and services, by pooling resources, is also important. Other drivers for sharing RANs include personnel efficiency; reduced pressure to spend high on spectrum acquisition; environmental considerations; and the ability to divert resources to more competitively critical investments such as those supporting customer experience.
Network sharing – just at the bottom of the curve:
These incentives are likely to outweigh the risks, especially as the regulatory climate eases. However, network sharing will not be a static concept. The nature of the mobile network is changing and increasingly, the intelligence will be located far from the cell sites. Many of the functions of the network, such as user authentication, charging, security and even the processing of the signals, will shift over the coming years to be run in the cloud – as applications on standard servers. These will often be supported by cloud services providers from their huge data centers, which will be able to serve many mobile operators.
If these carriers retain their own equipment at all, there will only be the stripped-down equipment necessary to transmit and receive signals (antennas and radios) left at the cell site itself. That will mean most of the cost of the network will lie in fees to the cloud provider, while the cost of deploying a cell site could be slashed as low as $150 by the end of the decade.
That 'virtualized' approach completely cuts the ties between the network and the service, a process which begins with simple RAN sharing. It also opens the way for even more radical definitions of infrastructure sharing, such as the 'on-demand' network envisaged by Google and others. In this model, a vast pool of spectrum and network capacity would be managed by an independent provider (such as Google itself) from the cloud. Providers of mobile services would buy capacity in a dynamic way as they required it – perhaps only at certain times of day for some kinds of services – and would pay fees set by a marketplace engine.
In other words, the mobile industry is only at the beginning of a process of dramatically changing the economics of its networks and service delivery, in a way which will eventually see every aspect of the operator's power base deconstructed and shared.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
