Urban rail networks – both above-ground commuter rail and underground railway systems like the New York City Subway – increasingly need to have reliable cellular coverage, with adequate and growing capacity.

Passengers and staff need dependable access to all major public mobile network operators (MNOs), plus in some cases private cellular networks as well.

I was recently invited by BAI Communications to comment on its Connectivity outlook report Transport and connectivity: smarter infrastructure for a smarter city” which shows that effective connectivity improves passenger satisfaction with transport services and operators. It also improves business productivity by enabling remote work.

In a broader context, connectivity is important right across urban environments, beyond rapid transit. Smart city authorities see the benefits in coordinating multi-modal mobility (blending rail, bus, tram, cycling, walking and e-vehicles), as well as improving health, education and environmental outcomes. These goals often require ubiquitous network coverage for all MNOs, including inside many public buildings and structures. Some leading municipal authorities are even considering becoming wireless infrastructure operators – and even spectrum-owners – themselves, to facilitate this.

The diverse benefits of connectivity will become especially important in the post-pandemic era, which also involves the extra opportunities and complexities of 5G. Together, I believe that these challenges are likely to favour the establishment of new neutral-host shared mobile infrastructure and associated business models, such as private networks.


A Neutral-Host Network (NHN) is a third-party owned cellular network, providing wholesale mobile coverage solutions to MNOs or other communications service providers (CSPs). NHNs have the potential to improve 5G and 4G network reach and capacity in rural areas, inside large buildings, in city-centres and along major roads and railways.

There are various technical and business-model options for NHNs, including examples where dedicated locally-licensed spectrum is used. Together they represent forms of network-sharing which can enhance competition, improve utility for consumer and business mobile users, and enable objectives of government and municipal authorities to be realised.

NHNs have many potential operators and stakeholders – tower companies, wholesale divisions of existing telecom providers, well-funded startups, rail and utility companies, municipal authorities and property developers are all active in this space. There is also a strong overlap with private and ‘vertical’ mobile networks, which are being deployed for industrial sites and transport hubs.

Railway and multi-modal city transport systems have variable wireless coverage

Today, good mobile network coverage is available on most urban rail systems where they run above ground. However, capacity and performance can often be constrained, especially during rush-hours on crowded trains, when many commuters are using their smartphones, catching up on news and social media feeds. Such applications are often video-rich, generating significant traffic loads. Streamed music, gaming and some business/cloud applications can also be demanding.

In addition, navigation tools on smartphones – and apps to rent shared-access cycles or call a ride-hail vehicle – are also dependent on good multi-MNO coverage, including at stations, access walkways and other interchange points. Rail is increasingly just one component of an integrated city-wide mobility network, with authorities emphasising multi-modal transport, where rail, bus, cycling and even e-scooters are linked together to optimise overall journey times and reduce environmental impact.

The rail element of a smart city has specific challenges. A train can hold 400-1000 passengers, and some places on the system may have multiple trains passing or running in parallel. This creates huge capacity-density requirements, spread across all the public MNOs. While this can generally be serviced by each MNO’s macro towers, there is often limited physical space for infrastructure alongside the tracks, as well as planning complexities for antennas on nearby buildings.

Underground rail can also support cellular connectivity, especially the newer transport networks constructed over the last 20 years, which have been designed with passenger communications as a priority. However, some older systems (or those with narrower / deeper tunnels) have significantly greater challenges for wireless propagation – there is less physical space for radio equipment, poor lines-of-sight in tunnels and strict constraints on scheduling of engineering work to avoid disruption to train services.

Typically, some form of shared distributed antenna system is used, with cable along tunnels, connected to MNO radio equipment in “base station hotels” in certain locations. Stations are somewhat easier and may have Wi-Fi access points or cellular small cells installed directly on platforms, perhaps linked more broadly to smart city deployments for good connectivity in public spaces. In addition, there may also be a need for public-safety communications systems to work underground – in future, these may converge with the cellular networks, using LTE or 5G as a platform.

Operationally, many MNOs’ workforces are not optimised for working in railway and underground rail environments. They have standardised processes and may lack appropriate rail-related safety and security accreditations. They may need customised scheduling to fit with the rail network’s operational hours. Multiplying this by three or four separate MNOs is even more complex. Conversely, the rail operating companies generally do not have the ability, financial resources (or perhaps regulatory permissions) to act as telecom providers themselves.

A similar set of issues apply to mobile coverage in other areas of the urban footprint – underground bus stations, urban “canyons” between high-rise blocks, bridges and tunnels with bike-rental stands, newly-built property complexes and so on. Creating functional smart-city applications becomes exponentially harder, when connectivity is poor – yet MNOs are not always able or willing to prioritise investment or may simply find it takes months or years to get permits and suitable sites.

As a result, a third-party commercial NHN provider can perform an important role as an intermediary, both for underground rail and other city domains. Such an organisation can specialise in the network’s operational and logistical requirements, and act as an aggregation layer for the MNOs’ technical needs. It may also be easier to structure the financial aspects, where the NHN deploys its own capital for upfront costs and matches its operating costs with wholesale access revenues from the MNOs. Potentially, the rail operator, or other municipal asset-owner, can expect a revenue share agreement.

The challenges of deploying 5G infrastructure in dense environments

Around the world, mobile networks are being upgraded to support 5G, both for enhanced passenger broadband access, and a variety of additional smart-city applications for businesses and IoT applications in public spaces. Urban rail networks can be expected to be at the forefront of this shift, alongside other busy city environments. All have high densities of demanding visitors and passengers, as well as internal operational use-cases such as staff connectivity, security cameras, digital signage, environmental monitoring and many others.

However, there are significant extra challenges for deploying 5G in cities’ transport systems and public areas, especially where it is intended to support the most demanding new use-cases. These extra constraints may apply above ground and indoors, as well as underground.

  • 5G brings new spectrum bands which will likely require new equipment, typically at closer distances to the track (or road) for above-ground coverage, particularly for mmWave frequencies which have much shorter propagation range.
  • The new frequencies, and a greater dependency on TDD spectrum may mean that existing DAS systems struggle to cope with 5G requirements, instead necessitating greater use of small cells and remote radio units. “Dynamic spectrum sharing”, blending 4G and 5G in the same frequency bands, may also create new problems.
  • Many MNOs are hoping to use 5G networks to enable new features such as network-slicing and the delivery of ultra-reliable low-latency communications (URLLC) for novel, super-demanding applications and IoT devices. These capabilities may prove impossible to deliver without much more granular and optimised radio systems, rather than today’s basic shared DAS.
  • Rail operators, public safety agencies, municipal authorities and other smart-city organisations may want to operate their own private cellular networks, separate from the MNOs’ infrastructure and operations.
  • Continued evolution of the 5G standard, from Release 15 to 16, 17 and 18 in coming years, will necessitate continued upgrades of infrastructure, addition of new bands and support of new features.
  • 5G maturity will coincide with the rise of various forms of open and disaggregated RAN technology. Neutral host providers may be at the forefront of adoption, as they are “greenfield” deployments, with less need for legacy equipment integration.

Taken together, 5G could well increase the viability and sophistication of the NHN model for underground rail transport and more broadly across cities, especially where obtaining new sites is challenging. The growing availability of local-licensed spectrum – including mmWave in some countries – could also support high-capacity ground-to-train connectivity to feed onboard Wi-Fi access points or cellular 4G/5G small cells.

Pandemic impacts on wireless infrastructure and use-cases

The COVID-19 pandemic has also increased the arguments for ubiquitous mobile coverage across transport systems and throughout urban areas, even though some passenger loads have fallen because of a shift to working from home (WFH).

The potential for transmission of infection is relatively high in transportation environments, as it is not always possible to maintain adequate social distancing, particularly during busy hours for travel. This is likely to mean additional mechanisms to manage the risk – some of which will likely use mobile connectivity.

For example:

  • Authorities may want the ability to notify passengers of delays, crowded stations, or to enable journey re-routing. We may see the emergence of smart “multimodal transport” apps that could help authorities load-balance the various transport options in realtime to minimise transmission risks.
  • It may be desirable to support future cellular-connected personal IoT devices, such as proximity-tracking wearables.
  • National contact-tracing operations may use a variety of connected sources of information, such as security cameras, mobile phone data, Wi-Fi connections and so on.
  • Smartphones or other devices will be unable to use GPS positioning underground, and will need other means to establish location – which will in any case be moving rather than static. On-platform or in-train connectivity may help establish contacts, proximity and timing more accurately.
  • Post-pandemic smart-city management will be reliant on ubiquitous coverage for all major MNOs. Citizens will want to use their smartphones to access bike-share and ride-hailing apps as part of their travel experience, as well as use trains and buses. They will also be ever-more reliant on delivery drivers and ecommerce services.

The precise mechanisms for all of this are unknown – and there may be other use-cases for mobile connections such as disinfecting robots, or mobile thermal-camera units and so on. Taken together however, the importance of cellular (and likely Wi-Fi) connectivity will rise, including the support of multiple MNOs’ networks and frequencies.

Conclusions: The strengthening case for neutral host

As a telecommunications analyst, I believe that urban rail systems, above and below ground, will continue to require greater investment in high-capacity mobile coverage. More broadly, the trajectory towards smart cities implies good wireless connectivity throughout the urban environment and public spaces.

Given the challenges involved in the upgrade to 5G in coming years, as well as the possible new use-cases (and government requirements) for managing the current pandemic or other future possibilities, the impetus will evolve with greater complexity and urgency.

There is a significant potential for new Neutral Host models to go beyond today’s basic coverage models, with additional sources of added value around high-performance shared infrastructure. A variety of new business models are likely to emerge, especially as the timeline coincides with the rise of private 5G networks and more-open wireless infrastructure. Citizen and IoT access to (all) MNOs’ networks will need to improve towards ubiquity, for cities to function effectively.

As well as NHNs providing improved coverage for passengers in trains, on platforms, or in other public spaces, urban rail systems and metropolitan authorities may also benefit from adjacent new revenue streams as well. They may also be able to use their other fixed assets such as station buildings, bridges, elevated sections of track and other infrastructure as suitable sites for masts or radio units, aimed at streets or trackside buildings. The installation of fibre to support new NHNs could also have wholesale potential in some locations.

Taken together, I think that city-oriented NHNs should help deliver improved connectivity, both for rail and transport applications, but more broadly for citizen connectivity and other applications such as security and environmental monitoring.

About the author

Dean Bubley is a technology industry analyst, futurist, speaker and consultant. He specialises in wireless, telecommunications and the internet of things. Dean is recognised as one of the world’s leading analysts covering 5G, telco business models and regulation, the future of voice and video and the emergence of technologies such as blockchain and AI. Dean was invited by BAI Communications to examine and comment on the findings of the Connectivity outlook report 2020.

About the Connectivity outlook report 2020

BAI Communications surveyed more than 2,400 rail users in five global cities: Hong Kong, London, New York City, Sydney and Toronto. Respondents were asked how they saw the state of transport, connectivity, and its role in the future of their city.

Our data shows that the benefits communications infrastructure enables, such as 4G and 5G wireless, personalised services and connected public spaces, are critical to innovation, opportunity and wellbeing.

You can access the findings as captured in the Connectivity outlook report 2020 here.