Major US carriers are rethinking neutral host networks. While MOCN promises multi-operator coverage on shared spectrum like CBRS, Verizon avoids it entirely and T-Mobile has withdrawn. But what are the alternatives? Here, we take a look at MOCN, MORAN, and an MNO-led neutral0host alternative.
MORAN is often used when the host provides shared Radio Access Network (RAN) infrastructure – towers, base stations, and so on – but lets each mobile network operator (MNO) use its own frequency licence. MOCN is preferred for indoor neutral host deployments – in airports, business districts, campuses – because it supports spectrum pooling and makes it easier for a single shared RAN to serve multiple operators’ subscribers seamlessly. Let’s discuss.
1 | MOCN (multi-operator core network)
MOCN allows multiple MNOs to share the same RAN – meaning they share base stations (gNodeBs/eNodeBs) – but keep their own core networks. The RAN equipment (antennas, radios, baseband units) is shared. Spectrum can also be shared, or MNOs may contribute different frequency bands. Note this is still different from the alternative MORAN model, outlined below – on the grounds the different frequencies are pooled, and not separated.
Even where MNOs contribute different licensed bands, the RAN setup in a MOCN system dynamically manages and presents them as a single, shared resource. The difference is in how the spectrum and RAN resources are logically integrated and controlled – not simply in who owns which frequencies. Like with MORAN, the core network (EPC/5GC) is separate for each operator, so each controls its own subscribers, services, and policies.
A neutral host provider can deploy and manage the shared RAN infrastructure in a building, campus, or stadium and connect it to the core networks of several MNOs. This means all MNO subscribers get services via the same shared RAN system – in turn, enabling efficient spectrum and site utilisation, while preserving operator-specific control of services and subscribers. It is also regulatory-friendly because the core networks are not shared.
2 | MORAN (multi-operator radio access network)
MORAN also allows multiple MNOs to share the same RAN equipment, but each uses its own spectrum, which remains separate per operator – unlike MOCN, where spectrum can be pooled and jointly managed by a single shared RAN. In MORAN, each frequency block belongs to one MNO; scheduling is per-operator and independent. Like with MOCN, operators share the physical RAN infrastructure, but each runs its own logical RAN instance on top.
In each case, this is tied to its own licensed spectrum. Each MNO broadcasts a single Public Land Mobile Network Identifier (PLMN ID), the code a device uses to recognise the network (and distinguish from others when multiple networks are available), and connects only to that operator’s core network. There is no need for multi-PLMN broadcasting or inter-operator signalling within the RAN. MORAN is simpler to deploy and operate.
It does not require shared spectrum, inter-operator scheduling logic, or requirement to coordinate fairness and quality-of-service (QoS) between PLMNs. Its simplicity makes it attractive where regulations limit spectrum sharing, or where operators prefer clear separation between their air-interface assets. It is often used for rural coverage partnerships, shared small-cell roll-outs, or early-stage neutral host programmes to reduce site and capital costs.
3 | MOCN vs MORAN – complexity and cost
While MOCN delivers efficiency by consolidating spectrum and equipment, it also introduces complexity on the grounds that a shared RAN must handle simultaneous connections to multiple core networks and manage traffic from users of multiple PLMNs within the same cell. As above, this requires inter-PLMN fairness policies so no operator dominates the RAN resources, scheduling algorithms to coordinate users on the same carrier and spectrum pool, and QoS enforcement to map each operator’s KPIs and SLAs into a shared RAN scheduler and buffer management.
These mechanisms make the RAN multi-tenant by design, demanding advanced software features and sophisticated orchestration. Integration and testing are more involved, since faults or upgrades in the shared RAN can affect several operators simultaneously. Performance monitoring, fault isolation, and SLA assurance must all be handled per operator, even though the infrastructure is common. This drives higher engineering and management overheads. MOCN also requires carefully defined agreements around spectrum use, service levels, and governance.
But the payoff is a more compact and efficient network that can deliver consistent multi-operator coverage in dense indoor environments – where duplicating full RAN stacks per operator would be impractical, arguably. MORAN, by contrast, keeps things simple. Each operator manages its own air interface and QoS, and coordination between cores and schedulers is minimal. It reduces capital and site costs, but not spectrum duplication. So take your pick: one minimises complexity and the other maximises efficiency. The choice depends on this trade-off.
4 | MOCN in shared spectrum (CBRS)
Where MOCN is deployed in a shared spectrum environment such as in the CBRS band in the US, the economics and engineering requirements shift significantly. Because all participating operators can use the same shared spectrum, the RAN hardware does not need to host multiple separate radios or antennas for each MNO. This reduces equipment costs, simplifies antenna design, and lowers site power and backhaul requirements.
This is in stark contrast with MORAN, where each operator’s licensed bands require separate RF chains and multi-band support. MOCN in a shared band like CBRS allows the neutral host to maintain the multi-tenant software features that handle inter-PLMN fairness, shared scheduling, and QoS enforcement, but without the cost and complexity of supporting multiple parallel frequency chains.
The same shared RAN can broadcast multiple PLMN IDs, route traffic to separate cores, and dynamically allocate resources across all users while relying on one pool of spectrum. This model makes indoor deployments highly efficient and easier to manage than duplicating separate MORAN stacks for each operator. But there is a wrinkle, of course, because most operators don’t like it – at least in the case of CBRS.
5 | MNO-led neutral host networks
T-Mobile has just pulled back from supporting MOCN in CBRS, citing better performance in its licensed mid-band spectrum. Licensed spectrum is the foundation of its neutral-host strategy, it has said, with CBRS playing a role only in selective complementary scenarios. Verizon has never (publicly) supported MOCN deployments in CBRS spectrum, also focusing on its own licensed bands for neutral hosts systems.
Which leaves AT&T as the CBRS outlier, currently. This trend makes clear their (obvious) preference for the control and performance advantages they get from their own licensed spectrum, versus a shared access model – even despite their various support, until recently, for a well-oiled supplier ecosystem to connect indoor venues (and load their networks, and placate their subscribers) on well-worked MOCN systems.
An MNO-led neutral-host model is different from a true MOCN setup as the neutral-host RAN works primarily on the carrier’s own bands. The carrier with the host contract – and host spectrum – controls the RAN, scheduler, and core connectivity, so subscribers automatically receive service. The site is “neutral” insofar as it is hosted by a third party and accessible to other tenants, but the indoor spectrum and radio resources are under one MNO roof.
Other carriers can join on a roaming deal or a spectrum lease. Roaming allows rivals’ subscribers to connect via another’s network; a spectrum lease lets a second operator deploy its own carrier over the same hardware, similar to a MORAN configuration. Both approaches maintain separation between operators’ spectrum and services, but they require additional coordination, operator-specific configuration, and potentially separate RF chains.
