In several regions of the world, network operators seem to be so set against sharing infrastructure that they’ll actively install duplicate/parallel systems to serve their own customers. Elsewhere, a focus on cost encourages wider co-operation. What is the right answer to best improve the end customer experience?
In-building cellular service remains an unsolved problem
Despite huge technical advances in the engineering of mobile networks and smartphones, there are still many places where performance falls far short of what customers want or expect. It can be particularly bad indoors, in underground car parks and industrial areas.
New building materials with higher RF isolation, the use of higher frequencies that have lower in-building penetration and the ever-growing demand for more data are all generating momentum to solve the issue.
Network planners and equipment vendors have achieved remarkable results with an “outdoor-in” policy, but it has its limits. Poor performance measurements for some buildings may be masked by those who either switch to Wi-Fi or to another operator for better service. The operator may be under a false impression that service is adequate when it continues to fall far short.
The big question is: "who pays?"
In the past, network operators have been expected to provide “always-on coverage, everywhere, at all times”. That’s not just difficult to achieve but expensive, involving interactions with many different building owners in many different jurisdictions. Investment has typically been focussed on the larger and higher profile locations such as sports stadiums, shopping malls and the larger office complexes.
DAS has been a popular technology for these projects because it can interwork with existing basestation products from all vendors. The interface is the same simple RF analogue connection used for an antenna. Costs have been high, partly because of the requirement to install separate basestation equipment for each network, partly because each operator often insists on support for multiple frequency bands and partly due to the specialist nature of DAS equipment.
Commonly, the DAS system would be designed and installed by a third party – there are several specialist DAS system integrators with the capability. Often this is led or supervised by one of the network operators, with others participating.
In some countries, the DAS system and installation would be funded entirely by the building owner. In others, the network operator would pay for everything. In Europe, its common for each network operator to expect the building owner to pay not only for the in-building DAS system but also for any basestation equipment required to drive it.
Sharing vs duplication outdoors
In the early days of cellular network deployment, there was virtually no co-operation or sharing. Each would have its own individual macrocell sites, often closely located next to each other. This allowed each network to have complete control over its cellsites, differentiating on quality and depth of network coverage.
I’d point to the Indian cellular companies as the pioneers of full cellsite sharing, where the huge pressure for lowest possible cost led to Reliance and others leasing towers, cellsites and backhaul from third parties. Other large operators elsewhere have sold off their tower assets on a lease-back basis. The use of shared macrocell sites, often owned by third parties such as American Tower or Crown Castle, is a commonly accepted business practice worldwide.
We can then look at the UK for the next step, where three UK operators created a joint company that owned and operated their RAN networks. Consolidation of cellsites radically reduced costs by dismantling several thousand of their most expensive sites. This took sharing a step further than simply sharing sites, but equipment, staff and transmission. It was only RF frequencies that remained independent.
Indoor spectrum sharing
It’s that insistence to retain full control over their spectrum that seems to continue to drive more costly solutions. ip.access has been promoting a shared spectrum approach for some years, highlighting that a single set of small cells could provide excellent service for all occupants from every network.
In the US, regulators are about to enable independent small cell systems using CBRS (now branded as OnGo). Building owners can initially setup their own Private LTE networks, and potentially interoperate with all the major networks to provide a truly seamless service. Analysts indicate this could achieve much lower price point, something like 40 cents per square foot rather than $1-2 using DAS.
Small cell solutions for larger buildings (such as SpiderCloud) and Distributed Radio Systems (such as Lampsite and Radio Dot) are one approach used in some countries.
Specifically in China and Japan, we see quite a lot of duplication where each operator installs their own independent system. It’s not uncommon to find three Huawei Lampsite systems or similar equipment in the same building. In the UK, OpenCell has pioneered a business model of installing separate sets of small cells for each operator, under a common management and leasing agreement. ClearSky takes a similar approach in the US. Although the physical equipment is duplicated, both companies claim that this is still lower cost than traditional DAS with independent basestations, especially for small to medium sized buildings.
It is perhaps surprising to see network operators funding their own independent in-building systems in some countries. This allows the most differentiation and quality control of the service provided, but you would think that this could be achieved at lower cost with well defined design criteria and using approved contractors.
Perhaps less surprising is the continued insistence that spectrum cannot be shared in-building. This is a factor that adds considerable cost and drives the need for more complex DAS solutions. As we heard at last year’s DAS Europe Congress, some operators continue to insist on as many as five different frequency bands for all their in-building systems.
Commercial CBRS service may well be the trigger for much more cost effective inbuilding spectrum sharing. Whether or not the same frequency or solution is adopted elsewhere, it may start to encourage operators in other countries that these lower cost solutions can achieve the quality they require – particularly for buildings with relatively lower capacity demands that are also equipped with good backhaul links.