Top Trends for Private Networks in 2023

Improved employee safety and productivity: Improved connectivity can significantly improve safety through the use of remote control of equipment, proximity awareness, or the use of sensors to spot early signs of danger.

Cybersecurity: Security challenges are becoming multi-layered and highly complex. Private wireless networks may be used as backups in case of failure of fiber or other links, and may also reduce risk associated with older proprietary systems.

Data and analytics: Companies are now rapidly transforming legacy analogue infrastructure and manual processes. Better connected equipment, IoT sensors, and video input can improve asset management, enable fault detection, improve productivity and resource allocation, extend assets’ working lives, help diagnostics, and optimize maintenance. The vast amount of data will require much more higher performing and robust networks than in the past.

Geopolitics: Many sectors are having to deal with ongoing shifts in the international political arena. Some countries and sectors have trade restrictions and embargoes. Taxes, tariffs, and import regulations have constant shifts. Governments demand clarity on employment and environmental compliance across entire supply chains. Companies have to deal with more paperwork and maintain better records. This is inherently more data- and network-intensive.

Equipment mobility and reusability: The increased utilization of vehicles, individuals, and assets both between locations and with partners such as rental companies or expanded supply chains necessitates constant and dependable network connectivity for effective monitoring and connectivity.

Private networks moving from 4G to 5G

The role of mmWave in private 5G

Moving from proof-of-concept to large-scale deployment

Private networks for public venues like stadiums, hotels, and airports

Combining private 5G with Wi-Fi in enterprises

Some countries have allocated spectrum for local networks that is only suitable for 5G, especially in the upper midband 4GHz+ and mmWave ranges (see next section). There simply cannot be private 4G in those bands, as it is not supported in devices or network equipment.

Many 5G MNOs initially deployed their main networks using “non-standalone” mode. This requires a 4G network radio and core as an “anchor.” This meant that for isolated deployments such as mining, oil, and gas facilities, or even larger ports, they would have needed to install both 4G and 5G coverage.

Until 2022, there were few 5G standalone (SA)-capable small cells or core network products on the market. These have now become more diverse, mature, and tested, although they can still be too costly for bulk IoT applications.

The fragmentation of 5G bands has made it difficult for low-volume IoT vendors to justify creating and certifying devices, until prices of chips or modules fell from initial smartphone-oriented cost points. Some frequencies such as the 3.8-4.2GHz part of Band 77 used in the UK were especially rare until recently.

Certain products and applications have only been available in 4G variants. For sites using private networks for critical voice communications (push-to-talk), the main handhelds used LTE – often as hybrids with PMR systems such as P25 or TETRA.

In the U.S., CBRS product guidelines and certifications were heavily 4G-focused until 2022.

Many enterprises have viewed 5G as immature and a “moving target,” especially as they could see timelines for releases 15, 16, 17, and beyond – with many of the more useful features such as ultra-low latencies only becoming viable later. As a result, some have used 4G either as the basis for proofs-of-concept or for full but “de-risked” deployment in the short term.

Many regulatory authorities are now focusing their private network allocations in 5G-only bands, such as 3.8-4.2GHz, 4.9GHz, and higher mmWave ranges. Sometimes this is intended to avoid “set asides” of MNO-oriented pioneer bands, while another reason is policymakers view 5G adoption as a strategic national goal.

Multiple vendors now have 5G SA cores available for deployment by enterprises and systems integrators, both for on-premises and cloud-based delivery. These can also often work in hybrid mode with 4G and/or 5G non-standalone, for greater flexibility and backward compatibility.

More MNOs have 5G SA networks, which can potentially support private or semi-private networks, either through coverage extension or (eventually) some form of network slicing.

Some MNOs’ regulatory and spectrum-management teams are becoming more agile at localized allocations for enterprises. B2B solutions teams can see if there is suitable “inventory” of unused public-network frequencies in specific locations that can be used for dedicated systems.

The device ecosystem (and especially IoT modules, gateways, and cameras) is adopting 5G more broadly, and is increasingly supportive of both 5G SA mode and a range of different frequency options – although local bands in some countries are still mostly 4G-only.

5G RedCap (reduced capability) is starting to emerge, aimed at optimizing 5G for lower cost IoT modules. It can be thought of as “5G light” and should be ideal for some private network scenarios in the future.

Industry groups such as CBRS Alliance, also branded “OnGo,” are starting to certify 5G solutions, which opens the market to a broader set of integrators and also allows easier interaction with the SAS authorization database providers.

Few deployments by public MNOs of widespread mmWave 5G networks that could be extended for private/enterprise use.

Few countries licensing mmWave bands for local/indoor shared use directly by enterprises.

Very limited numbers of suitable mmWave-enabled devices, especially in categories most frequently used for private networks, such as cameras, vehicle gateways, IoT modules, and industrial handhelds.

Typical poor range and propagation characteristics, meaning that mmWave 5G often struggles with mobility, or environments with restricted line of sight.

Lack of suitable small cells and other infrastructure elements capable of dealing with mmWave bands.

Several different mmWave bands used around the world, fragmenting potential demand.

Only suitable for 5G use, so no existing base of 4G equipment/devices to stimulate the market.

More spectrum available than in mid-band ranges, with the potential for 100-800MHz allocations for enterprise usage. This allows for higher throughputs with multi-Gbps peak speeds.

Less politically contentious for regulators than “set-asides” of precious midband frequencies, especially in countries with 4+ public MNOs battling for allocations.

Easier deployment of true ultra-low latency networks, reaching 1ms response times or lower. In some midband ranges there are sometimes technical restrictions on the way spectrum is used, to enable coexistence with other wireless users in the band.

MNOs can often use existing licensed mmWave bands for private enterprise customers’ sites, without impacting nearby macro networks. Conversely, when using midband frequencies they may face more complex network planning scenarios to manage public and private domains’ coexistence.

Potential to easily add private network capabilities to indoor mmWave neutral-host infrastructure. A single network of small cells could support 3 or 4 public networks, plus a dedicated local network for the venue, each with separate tranches of spectrum.

Enthusiasm by some network and semiconductor vendors to find alternative markets for mmWave products, given limited uptake for public 5G services owing to short range.

Growing ability to engineer enhanced mmWave coverage, especially with the addition of new indoor systems/small cells, plus new repeaters and range-extenders.

Ability to have customised uplink/downlink ratios reflecting specific application needs, or to using uplink carrier-aggregation to combine mmWave and midband 5G frequencies.

A car manufacturer deploying two 4G/5G small cells in a prototype pre-production facility. This installation was part of an industrial demonstrator and testbed setup, intended as a “lab” for a future full-scale factory that will be built in 2024-25.

A large city-based hospital and healthcare authority announced deployment of a P5G network. The initial deployment was in fact for one specific building (and non-clinical applications) at a remote suburban campus site.

A metropolitan authority used a local 5G connection at a specific traffic junction, connecting cameras and sensors to a roadside unit for optimizing flows of public buses and bicycles. This was a government-funded testbed, which has not yet resulted in scale-out to the rest of that city or elsewhere.

Availability of 3GPP Release 16 networks and devices, with various maturing features for network quality, low latency, and positioning accuracy.

A much wider pool of case studies, robust ROI models, and documentation about other deployments.

More trained and skilled individuals capable of deploying – and running – private 5G networks.

More regulatory authorities making spectrum available for enterprises, with particular bands such as 3.8-4.2GHz becoming seen as (approximately) harmonized across multiple countries.

Greater maturity of “templated” models for private networks requiring less customisation for each site. Some of these will be cloud-delivered “network-as-a-service” propositions while others will be more “network-in-a-box.” Both should support scale better.

New financing and commercial models, such as pay-as-you-grow cloud models, external bank loans or leasing approaches, bundling into larger industrial systems’ CAPEX and depreciation schedules, or various CSP-based subscription or services frameworks.

Growing awareness and trust of private 4G/5G networks by the enterprise IT functions, finance, and procurement teams, OT and industrial systems units, and various others.

Initial business case: In some cases, there is an urgent operational need for a specific network where the costs are clearly small compared to the financial or operational benefit. The replacement of old critical-communications push-to-talk systems with modern 4G/5G upgrades in industrial or military environments is a good example. Conversely, deployments in universities may have less-clear applications, weaker financial business cases, or more alternative technology options.

Simple/complex stakeholder models: Some deployments can be made by an individual company or government agency, on a specific site that it controls, with its own cash reserves, with few other parties involved in the decision process. Mining is a good example of this. Conversely, some sectors may involve a complex array of site-owners, network operators, integrators, tenants, infrastructure providers, government agencies, financiers, and so on. Inevitably the latter multi-stakeholder models are more cumbersome to negotiate, contract, and implement.

Long-term projects: Some private network deployments are fairly “standalone” – for instance connections for perimeter security cameras at an industrial site. Other implementations are linked to much larger projects, such as creating an overall industrial automation system for a next-generation factory. In those cases, the private cellular deployment may be tethered to much longer timelines for decision-making and investment.

Use-case “stacking”: In some industries, an initial trial for a particular P4G/P5G application may not just be successful in its own right but also may highlight an array of additional follow-on applications that can make the business case easier after proving out the technology. For instance, in ports, an initial trial for cameras on cranes could stimulate interest in 5G for drones, trucks, personal handhelds, traffic control, and so on. That “stacking” makes the investment decision much easier.

Technology dependencies: Some industries or specific applications may be heavily dependent on technology evolution and maturity. Some industrial use cases of P5G are quite clear – but really need features in 5G Release 16 or 17, which may not be available yet. Others may work perfectly well with 4G LTE, with 5G as a future evolution option. This one is tricky to score, as a given site or company may have a mix of short and longer term technical needs.

Regulatory dependencies: Some sectors may require particular regulatory developments in order to exploit private 4G/5G. Spectrum availability and usage rules are particularly important here – if a band is only available at low power or with short-term (e.g., 3 year) licenses, that may be prohibitive. Other regulatory issues may relate to industry-specific certifications, especially where there are questions around safety or interference (e.g., medical equipment and wireless).

Onboarding of visitors’ devices, to provide localized services with branded “Venue Mobile” icons on phone screens, or reduced reliance on public Wi-Fi or cellular roaming (especially in airports or hotels).

Venue-owned “neutral host” networks on which public operators become (paying) tenants or roaming partners. This is especially relevant in remote locations with limited public network coverage, such as resorts or offshore islands.

Dedicated wireless networks for venue IoT or automation systems, such as payment terminals, ticket kiosks, or security cameras.

Monetization of the private network by providing it “as a service” for on-site tenants such as retailers, event management and broadcast, catering staff, and so forth.

Segregated wireless networks for staff communications, for instance, on ruggedized handhelds.

Property companies seeking to use private mobile networks as a value-add for business or residential occupiers, for building management/IoT purposes, or (as above) with the intention of becoming neutral hosts or enabling “offload” from public carrier networks. Many already provide fiber backbones, Wi-Fi access, or rent rooftops for mobile base-stations, so private 4G/5G seems a natural extension.

There is a strong overlap between public venues and those which were most affected during the pandemic. Hotels were empty, sports venues closed, and retail shifted largely to delivery models. Few had the confidence – or spare capital – to deploy a new and largely untested communications infrastructure. New models of operation reduced some of the use cases as well – for instance, hotel staff now clean rooms less frequently, so there is less need for a sophisticated new communications system.

Early private 4G and especially 5G systems were relatively expensive, especially for locations with limited direct monetization or demonstrable ROI. While a mine can justify a private network’s cost for improved worker productivity or enhanced automation, it is harder to argue similarly for visitors to a town hall or kiosks in a shopping mall.

Most of these venues already have Wi-Fi. And while that has sometimes suffered with congestion or clumsy onboarding, those issues are increasingly being fixed with Wi-Fi6/6E/7, or simpler log-on pages, or automation with OpenRoaming or PassPoint. This makes private networks less easy to justify.

Onboarding of consumer/visitor smartphones to private networks is hard. There are 100s of types, with various OS generations. While many now have eSIM or dual-SIM capability, many people are unfamiliar with using that function – or else it may be locked down by a carrier. There are questions over what happens with phone calls or messages on the user’s normal phone number – critical in the days of 2FA security or SMS-delivered tickets. No venue wants to run a support desk for these issues.

Neutral host business models have proven tough to realize in practice. There are regulatory hurdles for public/private hybrid networks in some markets, questions over trust and cybersecurity, complexities about integrating a venue-run network into an MNO’s wider infrastructure and management system, and so on. Commercial models are still being tested

Specific niches and use cases have been identified as having value. Revenue-generating or safety-critical systems such as card-payment handhelds or gate-line ticket barriers need to be ringfenced from potential congestion on the public Wi-Fi network. Security cameras may need wireless backup connections in cases of cable cuts. It is easier to create business cases for such important applications.

Private 4G/5G costs and complexity are reducing fast. The arrival of cloud-based solutions, and a general intention to make dedicated cellular “as easy as Wi-Fi” to install and operate are starting to bear fruit. Various vendors are creating packaged “in a box” solutions, as well as scalable pay-as-you-grow cost models.

End users are much more aware of private 4G/5G than they were in 2021. Whereas vendors and even regulators were talking about the need for “evangelism,” there is now mainstream coverage in technology publications, while Wi-Fi and enterprise solutions conferences will also now include sections on private mobile. Private wireless is discussed at industry-specific events for hotels, airports, healthcare, or “proptech” (property technology).

The ecosystem has expanded rapidly. There are now vertical-focused specialist integrators and solutions providers for public venues such as sports arenas or hospitality. Various enterprise managed service providers (MSPs) have added private cellular to their existing portfolios. Planning and design tools are more readily available and understood.

Some public venues have adopted P4G/P5G for their outdoor areas in particular. Retailers’ parking lots, university campuses, business parks, and transport hubs often cover large expanses, which are impractical for Wi-Fi. While they may have public mobile coverage, few such sites are willing to pay per-GB for video footage from security cameras – and may want to link IoT systems more tightly into their IT platforms.

Rather than offering neutral-host capabilities as an add-on function to private networks, the reverse may prove easier. Some DAS-type systems are starting to allow private networks to be offered as a secondary function on existing or new neutral-host platforms, depending on the frequencies supported.

There is likely to be a growing opportunity for “P5G-aaS,” for instance with temporary networks being set up for festivals or pop-up sites. This may require a bit more flexibility on the part of regulatory authorities, however – especially in markets where licensing of spectrum is done “manually” rather than via an automated system like a CBRS SAS.

Separate devices and applications using P5G and Wi-Fi in the same places, such as staff handhelds (P5G) vs. visitor smartphones at stadiums. This can isolate different traffic types, as well as ring-fencing each domain in terms of security, interference, and congestion – and budget/administration.

P4G/P5G gateway units with local Wi-Fi deployed in vehicles or at remote enterprise structures, such as maintenance sheds.

CBRS or P5G backhaul from outdoor public Wi-Fi access points, where fiber runs are difficult or costly over longer distances.

Conversely, Wi-Fi or 60GHz backhaul to 4G/5G small cells may prove be useful, where fiber runs are difficult over short (indoor) distances.

Dedicated cellular networks (e.g., using CBRS) for fixed-wireless access for “unconnected” homes, deployed by education, healthcare, or local government agencies. These typically use on-premise equipment using Wi-Fi for final device connectivity.

Handover of connectivity from one network to the other, for example an indoor Wi-Fi robot or AGV to switching to P5G for travel between buildings anchored in a 5G core network (For example, a multipath demo^[5] from Deutsche Telekom for AGVs on campus networks).

Production and sale of IoT devices (e.g., cameras) with pluggable or switchable modules for different customers and sites. They may accommodate 4G, 5G, Wi-Fi, or some combination.

Devices which can bond or load-balance between multiple radios. Various approaches exist here, which may involve some sort of intelligent “combiner” unit in the network, or just connection-management software on the device combining independent connections. Technologies such as multipath TCP are becoming more helpful here.

Combined planning, design, operations, and security tools, which cover multiple network types (likely P5G, Wi-Fi, and also the LAN/Ethernet infrastructure).

Concepts for “offloading” public 5G inside buildings to both Wi-Fi and private 5G, perhaps for different MNOs’ traffic or different classes of device.

Open RAN and private wireless: In theory, P5G should be a natural win for some new disaggregated models for building network infrastructure. However, most work on standards and equipment has focused on the macro RAN and carrier domain. A few vendors have focused on indoor and private network scale offers, but smaller budgets mean that customised “mix-and-match” integration and testing is hard to justify for individual private network projects. Instead, there will likely be more pre-integrated or single-vendor solutions which use OpenRAN-type interfaces internally, which gives futureproofing while still allowing the quick and simple deployment of small networks in the short term.

Government-run private networks: A growing trend – albeit somewhat patchy across geographies – is the rise of government-focused verticals for P4G/P5G. This ranges from municipal networks deployed for smart-city applications or healthcare through to security and defense sites and bases. 2023 is likely to see a significant rise in the adoption of private networks for the public sector.

Combining private networks and neutral host: This report has references neutral-host wholesale mobile networks in several places, although commercial development has been slow. This is starting to change more noticeably, and 2023-24 should see growing use of such hybrids as vendors, venues, and regulatory authorities start to understand all the various moving parts. That said, business models and technology gaps still need to be addressed, for example around cybersecurity/lawful-intercept and mechanisms for dealing with payments and monitoring QoS across boundaries.

Smaller private networks: Most P4G/P5G networks have been deployed at campus-scale sites or larger wide-area domains like utility grids. As products become more standardized, easier to deploy, and automated (“zero touch”), the baseline should shrink considerably, especially with the use of cloud-based NaaS approaches. Companies with multiple similar sites, such as retail chains or hotels, could benefit from “cookie cutter” designs and rollout models. There will likely be important roles for new channel partners specialised in such verticals.

Cloud: A growing number of P4G/P5G vendors are leaning on cloud-delivered or cloud-managed functionality, especially for core networks. We can expect this model to evolve further although mass-market deployment is questionable for 2023, but more probable for 2024-25, as companies such as Amazon AWS push new business models and “pay as you grow” approaches with little upfront cost and tighter integration with applications such as IoT and AI platforms. That said, there will still need to be customization to particular building designs, radio environments, and company-specific details that are hard to automate or deliver remotely. There will also always be a need for onsite surveys, installation, and maintenance.

Sustainability: There is a growing focus on the environmental footprint of all connectivity and compute infrastructure, from energy consumption to CO2 “embodied” in equipment and device manufacture. Often, this compares to energy savings that are enabled by new automated and efficient buildings or processes, although the maths/accounting for carbon is complex and often debatable. For private networks, there will be opportunities in some new types of sites such as wind, solar, and hydrogen plants, while others such as coal mines and oil production may be challenged in the medium term. Elsewhere, smart buildings and cities generally require more connectivity, which is positive for this sector. The amount of energy used directly in P5G deployments tends to be low compared to the industrial or transport systems they assist.

India: Ongoing consultations and (often emphatic) debate between telco industry, government, and other stakeholders on suitable spectrum and rules for P5G.

Europe: EU regulatory agencies and various individual nations are looking at new bands, especially in the 3.8-4.2GHz range, as well as updated strategy for its 5G “action plan” and the “digital decade” ahead. Norway has already opened this band, and France has some early test licenses.

China: Early signs of a pivot towards direct allocation of spectrum to certain enterprises, perhaps in the 6GHz band. Until recently it has been very MNO-dominated.

Canada: Canada does not have a direct equivalent of the U.S. CBRS band but has been examining the 3.0-4.0GHz range.

Middle East: A number of countries such as UAE and Saudi Arabia have been investigating options such as the 4.0-4.2GHz range. There may be more movement in 2023.

South Korea: Increasing emphasis on private 5G should pick up pace during 2023, including dual-connectivity solutions running in 4.7GHz and 28GHz bands. Samsung is aiming to be a major global participant and has support from the government, including building P5G for some agencies’ own internal needs.