Special Report: New operators for industrial 5G

We have often written about the opportunity for various industries to make far more transformative use of mobile connectivity, if they are given greater control over how the networks are optimized.

This is not just about 5G, although it is critical to 5G that there is dramatically more industrial use of the new networks. This is true for the traditional operators, which need their investments in infrastructure to support a far wider range of revenue streams than more video and virtual reality traffic generated by budget-sensitive consumers. It is certainly true for many governments, which have placed 5G at the heart of industrial and digital plans and are expecting the new networks to improve economic competitiveness and social inclusion. And it is true, above all, for many industries which are on the brink of digital transformation processes and believe 5G could enable new processes and behaviors, which in turn would improve efficiency and productivity, and enable new services for customers.

Before 5G can deliver all these benefits, a deployment and operations model needs to be found that supports a good business case for the MNOs and the industries. In many cases this means private networks, and these are starting to be rolled out in 4G. After all, very few applications actually require 5G ultra-low latency or multi-gigabit rates, though they may be enhanced by those in future. What most enterprises really need is predictable performance, to the extent of enabling mission-critical and SLA-backed reliability. And they need to be able to manage all the data traffic they generate – to secure it, to comply with privacy and legal intercept regulations, to apply data analytics to support fast decision making at the edge of the network (on their own premises).

The best way to achieve this is to have a network which can be optimized, for an enterprise’s users, customers or locations, to support its particular requirements, such as fast response, high reliability or perhaps low cost/low data rate for some machine-to-machine services. And this network must be able to be secured and the data handled by the enterprise or its service provider, not entrusted to an MNO. Finally, it will increasingly be a localized affair, integrated with the company’s cloud platform at the edge.

As we have often argued, MNOs find it very hard to build a network which can be tailored for individual sectors. Their model relies on building an increasingly fast, high capacity pipe which can be used for any number of purposes by customers and (whether the operators like it or not) over-the-top applications. Optimization is carried out to ensure the best efficiency and performance on a blanket basis and to chase the elusive goal of predictable data rates and QoS (in the age of peak data speeds which are excessive for 90% of subscribers, predictability has become an increasingly important driver of satisfaction, rather than speed).

This increasingly powerful wide area network starts to creak when users are deep indoors or want a particular type of performance. Operators have found the ROI challenging when they look to push coverage deep indoors, and only in rare cases have they managed to strike a mutually acceptable balance of investment with enterprises. But in Rethink Technology Research’s surveys, it emerged that many companies would be happy to share the cost burden of building cellular networks on their premises, or even along their railway lines and waterways, if they could ensure those networks were under their control and fully tuned to support their connectivity and data requirements.

Hence the rising interest in private networks. These may be deployed directly by enterprises or their integrators, like WiFi; they may be built out by neutral host providers, which optimize a network for a particular sector and allow multiple enterprises, and the MNOs themselves, to ride on it; they may be supported by a large network equipment or cloud vendors with on an as-a-service basis; they can be deployed and managed by MNOs themselves (though multi-operator support can be a challenge in public-facing environments).

Eventually, 5G network slicing should address many of the requirements for connectivity and services optimized specifically for individual industries or user groups, but even in the most advanced 5G economies, full end-to-end slicing – with dynamic reconfiguration of network resources to suit specific needs – is years away. It requires a full implementation of the cloud-native 5G core, for a start. For large national or international industrial players slices will need to be called up on a cross-border basis, so issues of interoperability and roaming will come into play. New charging platforms will be required to commercialize slices; the slices will need to tap into physical and virtual resources, in fixed and mobile networks and cloud infrastructure, which are owned by many companies.

All this will take a long time to work out, especially on an international basis. In the meantime, some operators will create static slices for specific vertical sectors or use cases – a virtual reality slice or an automotive slice, for instance. But these will mainly serve the operators’ own services, such as VR gaming, or those of very large enterprises. And they will not, in this initial phase, support enterprises’ desire for their own security and control of their networks – that will need integration with cloud and edge resources and the implementation of localized cores which can work with existing corporate resources such as WLANs and IP PBXs.

So an LTE-based private network is a short term fix and one that will have a life of a decade or more in many industries and countries. That opportunity is attracting vendors, especially Nokia, which want to extend their reach into direct-to-enterprise networks and services; companies which specialize in enterprise equipment such as indoor small cells, edge compute nodes and virtualized EPCs (evolved packet cores); and vertical service providers, some of them moving from the world of traditional private networks, deployed for large sectors such as utilities or public safety with proprietary spectrum and technology.

The opening of the market to this diversity of players should create a richer ecosystem around industrial cellular services which will boost innovation and offer alternative profit models to those of the MNOs. This will help enterprises move from proprietary connectivity to the better economics of LTE, and future 5G; or have optimized, controlled quality of service for the first time, indoors. If the MNOs take a more positive view of sharing the enterprise opportunity in this manner, and see it as a way to penetrate the in-building and industrial environments without having the cost of building everything themselves, there is a far better chance that they can monetize 5G across a wide range of vertical sectors, not just a few chosen ones like auto; and that 5G will be able to deliver its promises to enterprises, governments and the MNOs themselves.

Small sliver of spectrum could open floodgates for sharing and neutral host

A proposal by the US regulator, the FCC, to open up a small sliver of midband spectrum for sharing between government incumbents and commercial operators, has caused a stir far greater than its 5 MHz size seems to warrant.

This is for two reasons. One, it indicates a willingness to push the sharing model, pioneered in the CBRS band and the TV white spaces, into more of the midband spectrum which is so prized for 5G. The USA is at a serious spectral disadvantage compared to China and many other markets, because so much of its midband airwaves are occupied by federal, local government or satellite users, and moving these incumbents to alternative bands would take many years. Sharing, then, is an attractive solution.

The second reason concerns Ligado, the US operator which emerged from the ashes of bankrupt terrestrial/satellite venture LightSquared. It is still fighting to be able to use its mid-band (L-band) mobile satellite spectrum, in 1525-1559 MHz, for 4G or 5G networks. It has an ongoing satellite communications business, but – like its former incarnation – wants to expand its reach with a terrestrial cellular network, which would support a wholesale-only model mainly focused on the IoT. The 5 MHz slice at stake is close to Ligado’s holdings and would enable it to aggregate a 40 MHz band, adding to the capacity to support a wide range of 5G services and service providers.

This is important because the US market would certainly benefit from a wholesale platform – like the 5G networks envisaged by the large cablecos, this could accelerate adoption of 5G in vertical industries by supporting specialized MVNOs and private operators; and that could, in turn, spur the large MNOs to pay more attention to industries. Ligado could also, potentially, be an interesting partner for the cloud giants – Amazon AWS, Microsoft Azure and Google all have a strong interest in closer ties to a network which would support enterprise cloud and edge services, and reduce their reliance on telco connectivity. An alliance with Ligado, similar to the one often rumoured between AWS and Dish, would be an interesting counterpoint to the established telcos.

First, Ligado needs to be able to build out its network. Its attempts have been repeatedly scuppered by objections from the occupiers of nearby bands, including GPS operators and government users. They claim the higher signal strength of a terrestrial system, compared to satellite, would create interference. It was the GPS community’s opposition which prevented LightSquared building out its network, even after the FCC had, perhaps prematurely, granted it leave to use satellite spectrum for LTE.

Now Ligado has shifted its attention to 5G, and it is hailing the latest FCC move as a step forward. The FCC’s chairman, Ajit Pai, has announced that the regulator will vote this month on a proposal to reallocate the small piece of spectrum – between 1675 MHz and 1680  MHz – for shared use between federal and commercial users. Ligado took this as a sign that the FCC is warming to the idea of allowing sharing in many of the midrange bands – particularly valuable for 5G – where government users are incumbent.

More broadly, this suggests that the FCC is confident about the success of its biggest spectrum sharing venture so far, the CBRS band in 3.5 GHz, even before those airwaves are in commercial use (see separate item). It is certainly looking to apply the technology which prevents interference in CBRS – the database-driven spectrum access system (SAS) – to other bands to increase the capacity available for 4G and 5G by sharing underused public sector airwaves.

Of course, it remains to be seen whether the US government continues to allocate portions of these bands for unlicensed use, as it has in CBRS, or succumbs to the lure of new revenues, as the value of midband spectrum rises. A government analysis estimates that the 5 MHz chunk of spectrum between 1675 MHz and 1680 MHz is worth about $600m.

The lure of a boost for the Treasury should be resisted, since sharing will help to open up the service provider ecosystem as well as providing more beachfront spectrum for 5G. This is much needed in the midband ranges. Even the CBRS scheme has been 4G-focused on the unlicensed side, a decision heavily criticized by T-Mobile and others for putting the USA behind in the ‘5G race’.

The licences which will be awarded later this year can be used for 5G, but are more fragmented than the large chunks of 3.5 GHz airwaves which some regulators, including China, are allocating. The company in the best position is Sprint, whose 2.5 GHz band has been accepted as an official 3GPP 5G allocation – it has over 100 MHz in most markets, with much of that not yet fully used for LTE.

Ligado said in a statement: “This NPRM [Notice of Proposed Rule Making] has significant potential to free up critical midband spectrum that will help ensure the United States commercializes the wireless spectrum necessary to accelerate, strengthen, and secure 5G networks across the country.”

Ligado’s chairman of the board, Ivan Seidenberg (a former Verizon CEO), added: “As a company, we are ready to invest in this spectrum and help deploy the 5G technologies that will not only improve all our lives but also give our economy a competitive edge against the rest of the world.”

The company has pledged to spend $800m to build a national 5G network based on both terrestrial and satellite and said this will create 8,000 jobs as well as giving US businesses and service providers an additional platform. Like Dish, it will kickstart its build-out with 4G-based IoT systems and then move to 5G.

Ligado insists that its midband spectrum will be optimal for emerging IoT applications. It said in a recent interview: “In order to accelerate delivery of innovative services in such verticals as healthcare, automotive, industrial, shipping, home and municipal sectors and to more broadly fulfill the promise of next generation mobile networks, more mid-band spectrum is needed. Midband spectrum is part of the solution to deliver ultra-reliable, highly secure and capacity-rich connectivity.”

One of the decisions the FCC must make is whether to auction the 5 MHz or allocate it to a particular company such as Ligado. If it goes for an auction, that will be at least a year away.  And Ligado is still waiting for a final verdict on whether it can use all its 35 MHz of L-Band spectrum for 5G –  nine years after LightSquared came up with its original plan.

Last November, Ligado announced one step towards commercial reality, signing deals with Ericsson, and with device chipmaker Sequans, to supply equipment and chips for its planned network. The plan is reminiscent of Dish’s – take the quick and low cost deployment option of 4G IoT networks first, to establish a beach-head for 5G. Ligado said it would roll out 4G LTE-M and NB-IoT technologies first to “take advantage of the broad ecosystem that is developing around LTE-M and NB-IoT specifications”.

Ligado battles on in quest to build a cellular network:

Ligado is still fighting to be allowed to use its spectrum for terrestrial usage. In May 2018, Ligado filed an amendment to its pending licence modification to reduce its operating power levels considerably,  and that brought some powerful influencers onto its side, including American Tower (which may be hoping for a new customer); the Competitive Carriers Association, which represents smaller operators, some of which could benefit from a new wholesale IoT network; the Wireless Infrastructure Association (WIA), which is campaigning for better smart building and smart city networks; and satellite operator Inmarsat.

However, one major player still remains to be convinced. A group led by satellite provider Iridium, and also including weather data players, still claims Ligado’s network will interfere with their services.

Last year, Iridium claimed, in an FCC filing, that Ligado’s IoT services would “pose a far greater interference concern” to Iridium’s airwaves than satellite operations. “Far from being concerned about any competitive threat as suggested by Ligado, Iridium objects to Ligado’s proposal because of the potential for significant harmful interference to Iridium’s operations in violation of the FCC’s rules if Ligado is permitted to deploy a

terrestrial service in the adjacent frequency band as currently proposed,” it wrote.

Ligado (which means ‘closedly tied together’ in Spanish) emerged from Chapter XI last March under the ownership of several investment groups, having won approval for its restructuring plan. Its owners are now Fortress Investment and Centerbridge Partners (34.3% between them), JPMorgan, and LightSquared’s original backer, Philip Falcone’s Harbinger Capital (44.45%).

It went on to sign agreements and settle litigation with the three GPS manufacturers (Deere, Garmin and Trimble) which had made its original launch impossible with claims that high power LTE signals would interfere with their transmissions in neighboring spectrum. The deals see Ligado committing to reducing out-of-band emissions from its planned LTE services, and avoiding spectrum adjacent to the GPS airwaves.

Virtualization will help Dish afford its wholesale, IoT-focused 5G vision

Dish has been very clear, ever since it mooted a plan to build its own cellular network in 2012, that it will not be able to adopt the traditional MNO cost model. As a new entrant with few infrastructure assets of its own, it would face a huge capex and opex bill to create a US-wide network from scratch.

At first, it aimed to work with an infrastructure partner, which would allow it to deploy its RAN on existing sites and backhaul, and it had an agreement with Sprint to support this approach, though that later collapsed. Then it aimed to acquire an MNO, but was gazumped by Softbank when it went after Sprint and Clearwire.

It may still pursue an acquisition – if the Sprint/T-Mobile merger plan falls apart, for instance, or with a cableco. Or it may just sell on its valuable midband spectrum to another 5G player. Some analysts think Sprint’s airwaves have now been devalued because there is such plentiful millimeter wave spectrum coming to market in the USA for 5G, but there are still challenges of range, battery life and device availability for the 28 GHz and 39 GHz bands, so we believe AT&T and Verizon (and TMO if it is denied the Sprint prize) would still be interested in adding to their midband arsenals.

But chairman Charlie Ergen insists that having to sell the spectrum, to please shareholders or for regulatory reasons, would be a “personal failure”, as he put it on Dish’s most recent earnings call. But when Dish builds out in earnest, it will do with a fully virtualized architecture that will slash operating costs and support new business models in wholesale and the IoT.

Last year, Dish was forced – by the FCC-mandated coverage obligations attached to some of its spectrum – to start building in the conventional way, but it is so far only rolling out an NB-IoT network, whose architecture allows for small numbers of low cost cells – they have wide range, and only need to support IoT data rates, not the huge capacity that Dish will eventually need if it wants to offer 5G video to its pay-TV customer base.

So, NB-IoT is very much a stopgap for now, and Dish is looking for a cost-effective way to do 5G and broaden its services offering. This is not to say it will stop deploying NB-IoT – though the technology, and the higher-powered 4G IoT standard, LTE-M, have a roadmap to 5G, this is not yet standardized. Indeed, it is likely to be some years before true 5G is available for the IoT – since the latest versions of NB-IoT and LTE-M are brand new, there will be little motivation for operators to move to 5G for M2M applications until they have sweated their new 4G assets a little, and until future 5G standards support the very low latencies that are promised.

But Dish will only accelerate its NB-IoT deployment when this can be done in parallel with 5G, harnessing the same virtualized architecture which is at the core of its envisioned business model to support many private and industrial operators. This platform would enable it to mix and match 4G and 5G technologies, and call any combination up to support a particular application or network slice.

Ergen says virtualization will enable his firm to slash the cost of entering the market. On the company’s Q1 quarter earnings call last week, he said: “When I go to one of our towers, I want to throw up. We’re building something that we know is… technically obsolete. The architecture has to change for a wireless network.”

Dish says it plans to develop and deploy a virtualized network which will place the baseband processing units in the cloud, and this will be rolled out along with a 5G core, with a view to supporting network slicing. It is logical that Dish would wait until the 5G New Radio Standalone specifications are included in mature commercial equipment – 5G SA requires a 5G core, unlike Non-Standalone (NSA), which works with the 4G core. Most operators are deploying NSA as a relatively simple first step, but of course, Dish does not have a 4G core. Now that SA specs are finalized, stable, mainstream equipment should be available around the turn of the year, though some operators believe they will wait longer for two key elements – a fully cloud-native 5G core, and a readily deployable virtualized 5G RAN.

If these two elements are deployed in a greenfield network, the running costs will be dramatically lower than a conventional network plus they will offer the flexibility to support many different services with different network behaviors, and to allocate resources cost-effectively where they are most needed. All that will lead to slicing, and the kind of platform that would enable Dish to do what it has often hinted at – not just provide 5G and quad play services for its consumer base, but support MVNOs, cloud service providers and industrial network players on a wholesale basis, allocating virtual slices of the network to each one, increasingly on a dynamic, on-demand basis.

Ergen was more explicit about this than he has been in the past, telling the analysts on the Q1 call that the plan is not just to offer mobile broadband services but to “tap into the power of 5G” and support robotics, smart cities, smart agriculture and other vertical markets.

Dish is also jumping on the USA’s current fixed wireless access (FWA) bandwagon, following its investment in Tarana Wireless, which makes advanced equipment for FWA and also has funding from Ergen’s other company, EchoStar. Of course, FWA could fit well with Dish’s pay-TV business to help bring multiplay bundles to rural areas, or to bypass the cable and fiber providers to offer cut-price packages in towns (though Ergen said his initial focus is rural, since other providers like Starry are targeting urban dwellers).

Ergen also highlighted Japan’s new mobile entrant, ecommerce and content company Rakuten, as his technology role model. Rakuten is deploying a fully virtualized, cloud-native network, initially for 4G and later 5G. Ergen said of that effort: “That architecture… is starting to gain credibility.  We’re in position to build a virtualized network from the ground up… In our case, we don’t have any [mobile] customers, so we can build it right the first time.”

One of the competitive advantages of a virtualized, flexible network is to support on-demand allocation of capacity – for consumers, and to replace the static MVNO arrangement for B2B and B2C providers. This goes hand-in-hand with dynamic pricing, and Ergen said this would be central to Dish’s model, not just to attract more customers, but to ensure the network assets are fully utilized. Just as Uber makes use of cars which would otherwise sit in the garage, so he wants dynamic, marketplace pricing to drive full usage of the 5G network at all times of day.

He added that Dish does not need to raise more capital for its NB-IoT network, but will consider various ways to raise the estimated $10bn needed to deploy full 5G over the coming years.

Dish will still need to find a cost-effective way to deploy and manage the elements of the network that have to stay physical – the antennas and remote radio heads. It may still manage to find an operator infrastructure partner, or could turn to a group of tier two tower and fiber owners to avoid heavy duty contracts with the major towercos. It could also look to work with several smaller MNOs.

Ergen told the most recent conference of the Competitive Carriers’ Assocation (CCA), which represents smaller US operators, that he aims to work with some of its members on tower access, and on cross-selling services on the new network. He also sees this as potential users of his sliced, wholesale platform, saying: “We can be a path for you for 5G. With network slicing, you should be able to lease a slice of our spectrum and it should look like yours.”

All this is a visionary approach to 5G enterprise and IoT networks, adopting a neutral host model which the USA could benefit from significantly. However, Ergen is well known for his big ideas, some of which – like the 4G network itself – never come to reality. And the core business at Dish has plenty of financial challenges, which may limit its ability to make the partnerships, investments and acquisitions it requires to build the platform it envisages. The state of its main business is certainly not comparable to that of Rakuten’s cloud/ecommerce/content operations in Japan, for instance, which are financing its planned mobile networks as well as leveraging them.

With the pay-TV unit in sharp decline, some analysts believe the longer game is for Dish to build up alternative businesses around 5G and the IoT, and eventually sell off its original business entirely. In that case, it will have a very different argument to make to potential investors in its 5G plan, and that platform will need to develop very quickly to become a standalone business, capable of outmanoeuvring the big telcos.


Satellite will have a key role in the 5G enterprise, as Amazon jumps in

Gone are the days when satellite was scarcely mentioned with regard to the mobile market, except to provide backhaul or access in very remote areas or to emergency services – or because of disputes over spectrum allocations.

Now, some companies are looking to use satellite spectrum for terrestrial 5G (in the USA, Dish and Ligado – see previous article), while others want to expand the use cases for satellite networks to support a wide range of enterprise and consumer applications, and give the technology a central role in 5G.

This week, satellite operator Telesat reported the results of a project with Vodafone and the University of Surrey in the UK, to carry out the world’s first demonstration of 5G backhaul using a low earth orbit (LEO) communications satellite. During the live test last month, Telesat’s Phase 1 LEO satellite connected to the university’s 5G testbed, and delivered a round trip latency of between 18ms and 40ms, among the lowest latency recorded for a satellite connection.

The backhaul was used to support video chat, web browsing and simultaneous streaming of up to 8K video, as well as 4K video transferred to the edge of the 5G network.

Other companies are planning to launch new satellite networks to support high speed Internet access, which in many cases will complement 5G, especially in industrial use cases, rather than merely filling gaps in ultra-remote areas. OneWeb, which is partly funded by Softbank, has already launched its first six satellites, and the much-discussed SpaceX is working on trials for its Starlink broadband deployment. OneWeb recently closed a new round of funding, raising $1.25bn, to bring its total to date to $3.4bn. The latest round was led by SoftBank, Grupo Salinas, Qualcomm Technologies and the Government of Rwanda, while existing stakeholders Virgin Group, Coca-Cola and Airbus are still involved too.

Now, Amazon plans to launch 3,236 satellites to support Internet access, according to filings with the International Telecommunications Union. The web giant has become increasingly active in supporting and trialling alternative connectivity options to those offered by the telcos – presumably to assert greater control over quality of service, optimization and cost, as it increasingly aims to manage and monetize the connections to its cloud, not just the cloud itself. This will become particularly important as the cloud is distributed to the network edge, and as the market moves towards end-to-end, multi-network slicing. In the latter scenario, cloud players like AWS will want to manage the orchestration of the slices, rather than letting telcos take that critical role in the value chain.

Amazon’s Project Kuiper aims “to launch a constellation of low earth orbit satellites that will provide low latency, high speed broadband connectivity to unserved and underserved communities around the world,” a spokesperson said. The initiative clearly echoes the work Facebook and Google are doing to bring connectivity to the ‘next billion’, and in so doing, push their services as the new users’ first introduction to the world of the web.

Amazon, based on insider comments and its track record, will have a strong initial focus on the enterprise, in addition to consumers.

At the Amazon AWS Re:invent conference in November, AWS’s CEO Andy Jassy announced AWS Ground Station, a service that allows customers to download data from satellites into AWS’ cloud infrastructure. This initially leverages a network of 12 ground station antennas located on the same grounds as AWS data centers.

Mobile satellite provider Iridium is already working with AWS, to connect its satellite network to the cloud giant’s IoT services. The company has joined the AWS Partner Network to develop Iridium CloudConnect, which it says will be the first satellite cloud-based solution with global coverage for the IoT.

The service will enable corporate users of AWS IoT services to reach more than 80% of the world’s locations which lack cellular coverage. And when it comes to applications that require 5G, the cellular coverage will be even lower for many years – 5G will mainly be targeted at urban areas in the first phase and, as it will mainly be deployed in relatively high frequencies around 3.5 GHz in the early years, it will not achieve very broad reach.

Its CEO, Matt Desch, said recently: “Iridium CloudConnect will completely change the speed at which a satellite IoT solution can be deployed and will allow existing AWS customers to keep everything the same on the back end, while opening up the opportunity to quickly expand their coverage. This is a major disruption for satellite IoT. Costs will drop, time to market will speed up, risk will be reduced.”

Iridium has a constellation of 66 cross-linked satellites to deliver global coverage and is pursuing a $3bn plan to replace its entire original constellation with new satellites, known as Iridium NEXT.

Farms and ports – among the industries that must look to non-MNOs for wireless

Nokia has been the most vocal of the major OEMs in targeting the opportunity for private and industrial cellular networks. It has multiple platforms which can support IoT and enterprise self-contained ‘sub-nets’ from the cloud, with or without the involvement of a mobile operator (it is careful to claim that MNOs will be part of most of these deployments, but in shared or industrial spectrum, there is no need for that).

The Finnish vendor also has an architecture, part of the overarching FutureX framework from its Bell Labs R&D arm, for future enterprise networks. ‘FutureX for Industries’ is designed to lay the groundwork for an Industrie 4.0 implementation, following the highly automated and intelligent approach pioneered in the German manufacturing sector in particular.

The Bell Labs blueprint is not just designed to provide an architectural foundation for Nokia’s expansion of its enterprise business; it also aims to inject 5G into Industrie 4.0. While the 5G community cites the transformation of manufacturing and industrial processes as a key use case for the next generation wireless networks, the same rarely happens in reverse, and most Industrie 4.0 demonstrations and conferences manage to go by with only passing references to 5G.

Frameworks like FutureX, and the trials and deployments that Nokia, Ericsson and Huawei are all conducting with industrial players like Volvo, Bosch and others, will all be aiming to change that perspective and convince enterprises that 5G is not just another connectivity option, but will add significantly to the applications they can deploy and the results they can obtain from their digital transformations.

Marcus Weldon, president of Nokia Bell Labs, said at the launch of the latest FutureX extension: “Industry is driven by a constant quest for leading profitability, which in a dynamic marketplace requires maximum productivity. The Future X for industries architecture is designed to describe the essential foundation of this Industrie 4.0 revolution, much as the Future X network vision we published in 2015 is serving as a blueprint for service provider network evolution.”

But Nokia, more than any other supplier, is seeking to push its industrial networking vision ahead well before 5G is universal. Two new projects highlight this, and the fact that, even if some large enterprises decide 5G doesn’t actually add a significant amount on top of 4G, WiFi and fiber, Nokia will still be there with a digital platform offering that may become increasingly divorced from actual connectivity.

One of the projects is in Brazil, a country that is unlikely to have widespread 5G for some years, especially in the agricultural sector which this initiative targets. Nokia has joined ConectarAGRO, an industry group that is looking to bring IoT technologies to the country’s farmers and their supply chain. The group consists of local mobile operator TIM, as well as major agribusiness vendors AGCO, Bayer, CNH Industrial, Jacto, Solinftec and Trimble.

ConectarAGRO says that it will bring connectivity to the 93% of Brazilian farmers that currently have no access to wireless broadband – around 500,000 farms. TIM will be working with Nokia to expand 4G coverage to support IoT and precision agricultural technologies. These include connecting farming equipment fleets, robotics, temperature and moisture sensors, and drone aerial images.

For Brazil, farming is a major strategic market. ConectarAGRO says that the segment accounted for 23.5% of GDP in 2017, and 80% of GDP growth. Consequently, if ConectarAGRO can provide efficiency and yield improvements here, it would be to the benefit of the national economy.

Luiz Tonisi, head of market unit Brazil at Nokia, said: “The ConectarAGRO initiative provides an industry-coordinated approach to facilitate the development of these precision agricultural technologies for Brazil’s farmers. This open approach will help to spur innovation by creating an ecosystem of companies that can support the digital transformation of the Brazilian agricultural sector. The Nokia team is very pleased to help realize Agriculture 4.0 for Brazil’s farmers.”

In this case, Nokia is leveraging its relationship with an existing customer, TIM, to create a channel to a new vertical market that needs IoT platforms and services. In many cases, this will be the logical entry point to new industries, providing the easiest way for Nokia to gain new business for offerings such as its WING IoT platform, while not creating conflicts of interest with its core MNO customers.

But it is very clear that WING, and other options like private sub-nets enabled by Nokia’s Cloud Packet Core, do not have to involve an MNO, provided there is another source of spectrum. With the growth of cellular technologies for unlicensed frequencies, like MulteFire; and the willingness of some regulators to allocate spectrum directly to enterprise sectors, as in Germany; Nokia will increasingly have back-up options if the MNOs prove slow or reluctant to roll out the systems enterprises need.

And it is worth remembering that WING supports non-cellular wireless technologies too, including WiFi and unlicensed low power WAN (U-LPWAN) systems like LoRa. These can be deployed by a wide range of service providers, and this will be a more likely scenario in some of the markets, including agriculture, which MNOs find it hardest to address with their current models of network deployment and charging.

The agriculture market is a big one, and there are major efficiency savings to be reaped if IoT technology is sown properly. But return on investment is tough to achieve if LTE or 5G networks have to deployed over huge sparsely populated areas – based on current pricing models, IoT services to farmers will not justify that investment. And the U-LPWAN community has high hopes for sectors like agriculture because the MNOs so often have insufficient coverage in place – while their networks, in free, sub-GHz spectrum, can deliver wide coverage at very low cost, with just a handful of base stations and gateways.

Broadly, the MNOs are approaching IoT applications in the same manner as they have done in consumer markets – that there is a per-use cost for sending data over the network, and a finite bucket of data available per tier, sold on a monthly allowance.

For many IoT applications, this is a very expensive way of doing things. For instance, an alarm application which sends an alert if a sensor threshold is met, such as water leakage, might only need a single weekly or monthly message to be sent, to check that the sensor is still active. Paying for a higher allowance of unused messages is extremely wasteful, and battery-constrained devices can’t simply make use of the extra unused messages in the bucket, as every message sent depletes the battery reserve.

Most battery-constrained applications favor delta-based messaging, where communication is only made if a sensor reading exceeds predefined parameters. In an application such as water level monitoring, there’s no point sending updates that say the tank is still full.

This is especially true in agricultural applications, where a lot of the IoT demand for sensing is for monitoring tasks. The farmers don’t want to know that all their thousands of cows are still doing normal bovine things; they want to know when a single cow is showing signs that intervention is needed.

So, if the MNOs collectively fail to find a way to price their offerings to suit these sorts of messaging demands, they risk being dislodged by the U-LPWAN rivals – which can place a low cost LoRa gateway on a barn, add a few sensors, and only charge a small monthly fee.

Nokia is clearly determined still to be in the mix, even in that scenario, and especially for enterprises which might want the wireless connectivity managed and secured, on an as-a-service basis, along with management of other resources such as edge nodes and data analytics.

Its second private networks launch of the past week shows how the vendor can start by supporting connectivity and then add more and more services so that it becomes a full digital platform enabler for a large industry – a role that its MNO customers covet too, but have so far (with a few exceptions) been poor at grasping it.

Nokia is working with a large port, Kokkola and Ukkoverkot, which has set up its own private LTE ‘micro-operator’ offering connectivity services to any companies using the port. The third largest general port in Finland says it handles massive amounts of data each day, which need to be transferred between its various operations, and it can now extend the services to partners and customers using its facilities. Among its connected data requirements – many of which use wireless networks because of the mobile nature of the port’s operations – are security, device management, automation of port functions and support for rescue and logistics vehicles.

Jyrki Roukala, development manager from the port, said: “It’s challenging to create data transfer channels in the vast port area. At first we considered using Wi-Fi, but an LTE network is superior thanks to its security, reliability and pervasive coverage. At the same time, we can commercialize our network by offering slices to other actors in the area.”

The FutureX for Industries architecture:

The new architecture has four technology layers held together by a common approach to cybersecurity (“defense in depth”, as Nokia puts it), and a range of new user interfaces.

The layers are:

  • Business Applications – specific applications for each segment, including key areas such as predictive maintenance, workforce efficiency and safety, and asset optimization.
  • Digital Value Platforms – this supports industrial automation, cognitive analytics and digital operations, and enabling network and device platforms.
  • Multi-cloud – this delivers proximity and elasticity of compute resources to ensure meet the performance requirements of business-critical applications using the most appropriate cloud models (from local clouds to edge and hybrid clouds).
  • High-performance Networking – dedicated ultra-reliable connectivity (wired and wireless) to link everything together.

Private cellular will be the business that makes or breaks the CBRS model

Initial commercial launches in the shared access element of the USA’s CBRS band could kick off within weeks, as the government completes years of testing and a decade of study and deliberation.

The FCC reached a significant milestone at the end of last month when it approved to the Environmental Sensing Capabilities (ESCs) of Google, Federated Wireless and CommScope. ESCs may only be used in conjunction with a fully certified Spectrum Access System (SAS) and within the geographic areas covered by the approved ESC sensor registrations.

Auctions for licensed usage of the band should follow later this year, and this three-tiered system has the potential to enable many new types of operator, including those offering fixed wireless; and private or neutral host industrial networks. Of course, the spectrum will also provide a low cost way for MNOs and cablecos to increase their cellular capacity rather than relying on WiFi alone.

The three-tiered system for the CBRS spectrum in 3.5 GHz sets up three levels of priority access. At the top are the federal incumbents, which will always have priority in the spectrum, but whose usage tends to be focused around the coasts (for naval radar) and on a few other applications, leaving most of the spectrum underused. ESCs may be used to detect the presence of federal radar transmissions.

The second tier is the holders of PAL (priority access licence) rights, to be auctioned later this year; and the third is the GAA (general authorized access) layer, which is open to anyone which registers with a spectrum database and follows the rules.

The PAL operators’ signals take priority over GAA traffic in their licensed airwaves, although there is also 80 MHz set aside entirely for GAA. GAA traffic can use up to 150 MHz in the band but is guaranteed the 80 MHz. The SAS allocates channels according to the priority of those seeking to use the band, and the database, combined with mechanisms such as listen before talk (LBT) prevent multiple users in GAA interfering with one another.

The future winners of PAL licences will be able to use the spectrum for 4G or 5G, while GAA is currently targeted at 4G, though this will change when standards for 5G in shared spectrum are finalized early next year as part of Release 16. These are being spearheaded in the 3GPP by Qualcomm, which is also the backer of MulteFire, a key technology that allows cellular technology to run in unlicensed bands without a licensed-spectrum anchor network (and therefore puts 4G and future 5G in the hands of non-MNOs).

One of the first applications expected to be supported by early GAA deployments is rural fixed wireless access (FWA). There were already some FWA incumbents in the 3.65 GHz band, which are being offered PAL or GAA status in return for their old licences, and these are set to be early deployers, along with some WISPs, and rural cablecos like Midco (see separate item on cable operators).

The maximum impact of CBRS will be felt if it is heavily adopted by industrial, private and neutral host operators, providing a low cost way for such providers to expand their reach into new locations, use cases and the IoT. Potential deployers include large enterprises, which might also look to acquire PAL rights, something GE, for example, has said it would be interested in doing. Others would be pure play neutral hosts, vertical market integrators, enterprise-focused cablecos, and cloud giants like AWS, which demonstrated several applications for CBRS at its Re:invent developer conference in November.

LightReading gives the private networks example of Landmark Dividend, which plans to use 3. CBRS spectrum to support a variety of private LTE networks for its customers. The company owns leases on the ground beneath cell towers, billboards and other infrastructure, which it will leverage to add industrial wireless services to its portfolio. Its first CBRS contract is with the Dallas public transit authority, which will use a private LTE network to connect up to 500 information kiosks across 90 bus and railway platforms. This will be a far cheaper option than fiber, and will use Ericsson equipment and Cradlepoint receivers.

Private LTE is important for the success of the CBRS experiment because it offers a different proposition from that of WiFi, to a far greater extent than in consumer, home or fixed wireless use cases. There are enterprise and IoT applications which are better served by cellular technology than WiFi, or can use a combination of the two, whereas for consumer applications – as Joel Lindholm, VP of LTE business at Ruckus Networks, puts it – “WiFi is cheaper. It’s always going to be cheaper,” from the chips to the end product, and it’s pervasive.

Ruckus made its name in carrier WiFi but now offers CBRS equipment too, and sees the two as complementary. Lindholm told FierceWireless that he thinks private LTE will be the main driver of CBRS uptake, and provided a list of use cases which require cellular coverage, mobility, security or low latency, and which Ruckus has seen sparking interest in its base of trials and prospects:

  • Secure, wide area, HD video surveillance
  • Critical communications for security and operations teams
  • Remote vehicle and equipment control
  • Automated guided vehicle (AGV) connectivity
  • Mobile point-of-sale and kiosk connectivity
  • Mobile high volume data transfer for imagery, video, diagnostics
  • Mobile connectivity for utility, public safety and passenger transfer vehicles
  • Low latency connectivity for industrial automation (Industry 4.0)

CBRS router maker Cradlepoint is more assertive about the benefits of cellular over WiFi. “It’s really about replacing WiFi,” said Ken Hosac, VP of IoT strategy and business development. “A lot of enterprises are trying to use WiFi in applications where it’s just not working.” He used the example of managing data from large numbers of HD cameras in a secure environment such as a large warehouse. Cellular technology is desirable, as is a private network – “There’s no reason to boomerang that to a public network and back, and, it would be way too expensive anyway,” he said. The main justification for CBRS, he believes, is to make it affordable to deploy LTE for the IoT, especially in industrial settings “where WiFi has never worked, and they like the capabilities of LTE”.

Hosac said in an interview: “The whole world is watching. If it works, it’s a model for others around the world.”