Roaming, where cellular service is provided by partner networks, has been a tremendous benefit for foreign travel – extending coverage to all corners of the globe. National roaming is used in relatively few countries to fill in isolated pockets that may be otherwise uncommercial. The success of neutral host enabled small cells will rely on successful national roaming agreements. What’s involved and how could it come to pass?
Aggregator vs Neutral Host
Up to now, small cells have been focussed on direct deployment by an existing cellular network using their own spectrum. In some cases, the cost of the small cells might be paid for or subsidised by the building owner. These form an integral part of the existing network and are connected directly into the operator’s core. The potential for large numbers of independent enterprise small cell deployments (perhaps thousands or tens of thousands per country) make it likely that several aggregators will appear who can scale to support large numbers of independent installations and interwork with the major cellular operators in each country. I’d term these organisations as Small Cell Aggregators.
It is also proposed to allow spectrum from one network operator to be shared by all cellular operators within a location, simplifying the deployment of small cell solution. Technical standards such as MOCN already support this feature. I’m not aware of this in commercial use anywhere as yet, but if it was we could expect the aggregator to have more core network functionality and become a so-called Neutral Host. This term has often been used in the past to describe organisations that install and operate DAS (Distributed Antenna Systems) connected to multiple network operators. Unlike a MOCN or shared spectrum solution, these DAS systems use different spectrum for each operator and don't take an active role in call routing or core network functionality.
With the introduction of LTE small cells using shared and unlicensed spectrum (specifically CBRS at 3.5GHz and MulteFire at 5GHz), these can be deployed and operate completely independently as standalone private LTE networks. Their full potential will be achieved if they can seamlessly integrate with existing cellular networks, which would require roaming to be established in the same way as between the existing cellular networks worldwide today. Any organisation facilitating this could be appropriately named a Roaming Neutral Host.
Recap of international roaming arrangements
Cellular networks allow each other’s customers to use their networks on a reciprocal basis. The common technical standards for smartphones and networks ensure compatibility for 2G, 3G and 4G although not all frequency bands or modes are supported by every device. 2G GSM is always a common denominator embedded in every phone and likely to remain so for a long time.
Networks have two separate types of connections with their roaming partners – a signalling/telecom one and a billing/charging one. They use signalling which initially validates end-users, authenticating (ie. verifying identity) and authorising (i.e. permitting use). Calls and data may be routed via the home network or breakout locally. All calls are logged with standard format billing records generated as the basis for wholesale charging.
With over 800 cellular networks worldwide, the complexity of managing individual partner relationships is considerable, especially for the smaller players. Wholesale roaming businesses are established which handle the signalling and/or the billing interworking, ensuring that the sometimes large imbalances between inbound/outbound roaming fees are paid on time. These roaming hubs don’t have core network functionality and don’t directly process or control the calls, instead validating/routing signalling messages and translating formats where required.
Some of the larger multinational networks optimise their roaming costs by handling directly the transactions between their subsidiaries and/or their largest partners.
Introducing a new partner
Any new network provider need only prove interworking by testing with a single roaming hub and signing commercial agreements. They would require their own unique mobile network identity code in order to route traffic and identify the source, and could issue their own SIM cards should they wish.
I’d expect the roaming hub would require a deposit to cover against risk of large imbalances or abuse, but they have systems in place to terminate service in that unlikely scenario.
This makes it unlikely that each individual small business could become a full roaming partner on the same basis as existing large scale cellular networks.
Instead, I would expect to see several neutral host organisations emerge within each country to provide this service, plus perhaps a few regional multi-national players. They would need to be designed to operate at scale, with large numbers of separate partners – perhaps as many as tens of thousands each – for which they would provide the core network and roaming features needed.
Fraud and abuse protection
Network operators are a very conservative bunch, and would be concerned about introducing large numbers of unknown service providers into their tightly controlled systems. Even today, you only have to look at marketing from fraud prevention vendors to see there are concerns about roaming fraud. For example there’s also a growing grey market in wholesale text messaging, where SMS can be sent at lower cost from third parties (typically there is no wholesale fee to send an SMS internationally).
However, roaming neutral hosts may choose not to operate or host such services. They may decide simply not to issue their own SIM cards or block them from use elsewhere and only accept inbound roaming, so would not generate traffic from roaming revenue elsewhere. All that would be seen by roaming partners is their own users accessing the system, sending SMS and making voice calls via their own core networks.
It seems more likely (at least in the short term) that relatively few trusted partners will qualify for roaming relationships in each country, rather than the system upscaling by several orders of magnitude. Roaming neutral hosts will have to ensure they have suitable checks and balances in place – it’s not just network performance itself that will be needed.
I see parallels with the credit card industry where the large international card networks (Visa and Mastercard) operate through a number of accredited banks and credit card aggregators in each country, with businesses signing up to a single aggregator/bank for all of their transactions.
Matching speed and performance
Many people have observed that mobile networks don't seem to work as well when abroad - despite fast measured LTE data rates being reported worldwide. Not everybody realises that their traffic is routed back through their home network and/or that traffic profiling may be applied. There has also been some underinvestment in roaming interconnect by operators in the past, throttling roaming throughput.
These constraints don't usually apply to national roaming traffic, but operators will need to ensure that their systems are suitably dimensioned and fit for purpose. It's an aspect that's easily overlooked because few complaints are directly received about it and these could be attributed to the roaming partner.
Establishing an ecosystem
Network operators are moving at a glacial rate to satisfy demand for third party funded in-building wireless. Up to now, Wi-Fi has succeeded partly because it can bypass approval or interworking the the cellular networks. CBRS and MulteFire could also succeed on a similar basis but in my view would be much more beneficial with some form of roaming in place.
This needs the industry to move quickly to establish similar guidelines and business processes to adopt, interwork and support roaming neutral hosts in each country.
I believe this is far more urgent that some of the technical research being made into other areas of cellular technology, and would provide greater long term financial benefit for the industry as a whole.