The world’s first country-wide multi-operator core network sharing project

The Future Network spoke to Graham Mitchell, CEO of Crown Infrastructure Partners, the company overseeing New Zealand’s broadband and cellular network expansion in remote rural areas. This innovative, government backed project evolved from a nationwide broadband access project to incorporate cellular networks, and uniquely managed to partner with a joint venture representing all incumbent MNOs who will be sharing spectrum and RAN network infrastructure to provide mobile coverage in remote rural areas and coverage of highways.

The Future Network: Please introduce yourself and your company.

We are a government owned corporation, composed of private sector-background executives and we report through to a board and government. We were set up initially to accelerate broadband access in NZ, supplying fibre to the home, so we ran a large tender to cover 75% of the population. Following the tender, we set up four large public private partnerships (PPPs), providing about NZ$1.3 billion of government funding alongside our private sector partners’ total co-investment of approximately NZ$4-5 billion. The original ‘ultra-fast broadband’ initiative (“UFB”) as it is known is now about 80% complete, and we have recently expanded the programme to provide fibre to about 87% of the population – about 390 towns in total. When we started, New Zealand was near the bottom of the OECD rankings for population covered by fibre. We are now about 11th, and when we finish we’ll be about 5th for fibre broadband coverage, just behind Korea and Japan.

The Future Network: Could you please explain a little about the business model that you have adopted?

The business model is that we have a main partner Chorus (similar to Openreach in the UK) which was structurally separated (through a share market transaction on the New Zealand and Australian stock exchanges) from Telecom New Zealand, the previous incumbent (now called Spark New Zealand). Chorus is responsible for approximately 73% of the fibre build. Then there are three other partners, two electricity companies and one Council owned metro fibre-network provider, all of whom are also wholesale only. We also have about 90 retailers on the fibre network, which represents approximately 38% uptake. It’s a voluntary migration and it’s growing at about 1% each month, and we’ve already had some capital returned to us from the partners as well, so we lent them what was effectively interest free money in order to get the private sector engaged, because as you know, these sorts of projects are often very hard to build, and the construction of an economic case and securing uptake is challenging.

The Future Network: And so how did you go from a country-wide broadband project to incorporating a multi-operator core network (MOCN) solution?

The MOCN, or shared RAN project came about because we were asked by the government to run a tender and invest again with the industry, but this time using the Telecommunications Development levy which is similar to a universal service obligation (USO) levy and utilising some of that and some of our own capital which we’ve had returned under the original UFB initiative. We are targeting improved broadband service for the last 4% of New Zealanders in rural areas and increasing mobile coverage on our state highways and into a number of key rural tourist spots. So, we ran that tender, and the challenge was to make the economics work, which historically has been difficult. In the previous model (the Rural Broadband Initiative phase one), when the government ran the process it was awarded to Vodafone and they ended up with the other operators co-located on the Vodafone owned towers. The challenge with co-location is that you needed to have quite large towers with a lot of power, which is expensive, even with the subsidies. You also tend not to get all the operators signing up under a model like this, although you do have emergency services and carrier roaming. And despite having varying market shares (Vodafone 39%, Spark 37% and 2degrees 23% source: IDC), the three operators got together and put in a bid through a joint JV entity they called the Rural Connectivity Group, with each having one third of the JV. They have a contract which will initially cover about 70,000 rural households, and 1,000km of highway across 32 different highways, and 100 tourist areas scattered throughout the country. The target initial build is approximately 400-450 towers and there is an expansion option which could expand the number of towers. They improved the economics by using a common tower and sharing a RAN through multi-operator core network (MOCN) technology. It will be a 4G LTE with VoLTE and then they will go back and integrate this into their core network, so all three of the operators will be able to offer service on day one. They will be offering a fixed wireless broadband service and a mobile service. There are also open access requirements whereby the RCG has to offer wholesale backhaul and co-location to at least one wireless ISP. These different services are all co-located on the same tower. All the operators will compete on the network at a retail level, and there’s no pricing regulation applied. The only thing the MNO retailers share are the towers and the radio equipment, which are purchased from the JV as wholesale blocks of capacity; the MNOs then compete in the market for subscribers. The unique thing is that the whole mobile industry here is using the shared RAN.

The Future Network: So, are the operators using their own spectrum or are they sharing that too?

They are pooling their spectrum. They will be using 700 MHz initially because that’s best suited for rural, then they will move up to 1800 MHZ, 2100 MHz and then 2600 MHz and they will deploy a quad band antenna and use carrier aggregation. An interesting fact is that the final 3-4% of the population in New Zealand will have more spectrum available to them than any other portion of population. Currently the operators can only use their own spectrum to offer a fixed wireless broadband service, and it is a good service, but once it gets popular it gets constrained because of the spectral nature of it, so the ability to add additional carriers on to it, within the same spectrum band, is quite attractive and quite economic.

The Future Network: Resistance from the operators when it comes to sharing spectrum is something that is familiar to many people around the world, can you talk us through how you coordinated them and got them to agree to pool their resources in this case?

What really drove them there was first that they wanted to participate with us in the high-profile project at government level, and the development of the programme was industry-led. Secondly, the MNOs advised us that the economics stacked up. They worked out that if they tried to do it individually the economics weren’t viable, so they decided to work together. It’s interesting that the result we have got and contracted for is that three competitive MNOs agreed to work together on infrastructure, but on more than just towers, on the spectrum layer too. There is a mobile black spots initiative in Australia but that is a single operator solution with co-location options which is fine from a public safety and emergency services point of view, but it doesn’t help with the tourism agenda, which has been a crucial driver in New Zealand.

The Future Network: So what is the timeline on the project?

We’re aiming to be completed by 2022, but we’re yet to build our first tower with the Rural Connectivity Group as the contracts have only just been awarded. The first year is mainly going to be a series of land acquisitions and equipment tenders, and then we anticipate the build happening quite rapidly after that. It’s a reasonable sized tower build and the terrain is large and varied. There will need to be some satellite backhaul to accommodate the more challenging environments which adds to the complexity. When it’s complete, the project will have covered an additional 15-20% of the land mass of the country, and will include one very remote island which currently doesn’t have any mobile coverage at all.

The Future Network: What is the technology that you are deploying? Is it all macro cells, or are there some small cells and other network densification technologies that you’re going to be employing?

It ranges from outdoor femtocells to use for track head ends at the commencement/end of tourist walking routes which can be used by the emergency services, to small cells, macro cells and transmission towers. The network will be 4G, with a smaller 3G overlay to provide for voice services until the penetration of handsets with 4G voice capability increases. There are also a range of backhaul technologies, as I mentioned earlier, some will be provided by satellite, but it’s predominantly provided by digital microwave radio (DMR) and possibly some fibre. When it comes to the power solutions, where it can be connected to the grid it will be, but about 10-15% of sites are off grid, for which we are using solar solutions. These work quite well, and we’ve been advised by the MNOs that the power requirement for the sites isn’t as great because the operators are not co-locating all of their separate equipment.