Risks and Challenges for 5G

In an earlier blog post, I listed reasons for why 5G is needed. This blog post takes a look at what kinds of risks and challenges exist for 5G in the areas of spectrum, small cells, technology fragmentation, revenue, cost of rollout, IoT, NFV, SDN & cloud, alternative access technologies, network sharing, use cases, backhaul & fronthaul, regulation, and devices.

Spectrum

• In GSMA’s 2017 operator survey, 34% of the respondents ranked spectrum availability and spectrum fragmentation as the most important risk in delivering 5G, making it the second biggest risk identified in the survey [GSMA-2]
• The accelerated schedule for 5G that is, 5G NR (New Radio) non-standalone, means that there will be less time to agree on some of the open issues around 5G, such as the fact that no global spectrum bands have been identified for 5G [BMI]. Several ongoing trials are using frequencies that are only available in a specific market or region. As an example, the spectrum that the US wants to use is not available in most other countries [RW]. Global bands that are harmonized from one country to another need to be identified to ensure the necessary economies of scale for 5G device manufacturers and the ability for devices to function across many regions. The World Radiocommunication Conference (WRC) in 2019 plans to discuss spectrum from 6GHz up to 100GHz to harmonize its usage across the world. However, 2019 is fairly late, considering that pre-commercial 5G trials have already started now in 2017. Rethink Wireless is concerned that this might lead to gambles, guesswork and political maneuvering when it comes to which spectrum bands to use.
• The expectation is that by 2020, the number of LTE operators has increased a third compared to 2016 and LTE coverage has reached 75% of the world’s population [GSMA]. However, 150 countries will still have LTE adoption at less than 30% of all cellular connections, the majority of the connections being either 2G or 3G. In the long term, limited LTE adoption could mean reduced opportunity to shut down 2G and 3G networks to refarm spectrum for LTE Advanced and 5G. This could prevent or at least slow down 5G deployment. 
• The costs of using higher spectrum bands to build a 5G network could be enormous [LR]. Deutsche Telekom’s (DT) CEO estimates that the cost of covering Europe with 5G could be anywhere between $318 billion and $529 billion. DT’s CTO expects that radio will account for between 50% and 70% of the final 5G bill largely due to network densification (i.e., small cells). According to a report from Barclays, the Return On Capital Employed (ROCE) needs to be at least 8% for an operator so that the operator is able to finance a 5G rollout. Incumbents like DT probably have a ROCE that is at least at that level. However, tier-3 operators often have a lower ROCE and the overall industry barely reaches the 8% ROCE level.
• Rethink Wireless is concerned that 5G is in danger of not happening in the manner and timescale that vendors and operators are hoping for [RW]. Some of the reasons include the considerable complexity 5G aims to address and its cost. According to Rethink Wireless, the use of spectrum above 6GHz is likely to be slow and perhaps less than one percent of only very advanced regions will be covered by 2020, which may not generate enough of revenue to sustain the business of all the major players.

Small Cells

• To satisfy demand and enable the introduction of 5G, some estimates indicate that at least 10 small cells per current macro base station would be needed in dense urban environments, where cells are already today rather tightly deployed so that they can cope with traffic density [ARC]. 5G antennas are likely to be larger than current 2-4G antennas because of massive MIMO (Massive Input Massive Output) processing that will require the use of a very large number of radiating antenna elements. Additional antennas compatible with the new 5G spectrum bands will need to be deployed also. This means that the reuse of existing antenna masts could be problematic and new transmission sites will be needed. New sites are also needed to install small cells in semi-elevated locations. Due to the required high density of the new installations, operators will need to deploy the equipment on infrastructure like bus shelters, lamp posts, public buildings, billboards, street furniture, etc.
• A survey by TIA (Telecommunications Industry Association) indicates that lack of partner diversity is a likely challenge for 5G network densification efforts [GNW]. Since small cells are a key requirement for the maintstream rollout of 5G, operators need to ensure they have the right partnerships (e.g., with building owners, electrical utilities and public lightning companies) in place to address site acquisition challenges.
• Analysys Mason estimates that 5G Fixed Wireless Access (FWA) will most likely be cheaper than FTTP (Fiber-to-the-Premises) (except where this is aerial) but will cost more than G.fast [AM]. However, this may not matter in the long run because 5G fixed wireless access cells can be reused for future 5G mobile wireless access. 
• 5G may not appear in its full form at first, at least not until 5G coverage is built out using small cells across all the spectra [RW]. Rethink Wireless expects this not to happen until 2030 or 2035. By that time, today’s MNOs (Mobile Network Operators) and equipment vendors could be facing tough times. 5G elements might be able to sit inside a sub-6GHz LTE and LTE-Advanced network as easily as in a full 5G network. That would mean that no full-scale upgrade will be needed this time. Rethink Wireless believes that 5G will not involve another big bang upgrade since this is economically and logistically unfeasible for most operators for another 10 years or so. Instead, 5G will be phased in gradually alongside LTE on the MNO’s own schedule. Rethink Wireless expects that MNOs will take the quickest upgrades relatively quickly and then be very slow to deploy small cells. Since 5G can co-exist with LTE, operators can afford to wait perhaps even 15 or 25 years before fully deploying small cells. In the meantime, much more investment will go into macrocells.
• Vodafone’s CTO has stated that he is worried about the hype surrounding mmWave (millimeter wave) technology [LR-2]. There are practical problems with mmWave, including even leaves on a tree becoming barriers to signals in the mmWave range. Vodafone believes that there might be hotspot areas where one can use mmWave, but sees sub-6GHz as the sweet spot for 5G. Besides leaves on a tree, mmWave spectrum will not go through walls and not even all windows [RW]. Thus, the propagation qualities of the mmWave bands make it difficult to achieve widespread coverage especially in more sparsely populated areas [ARC]. However, it should be noted that very high frequencies are mostly an option for line-of-sight and non-line-of-sight outdoor-to-outdoor coverage and indoor-to-indoor coverage, rather than for outdoor-to-indoor coverage [IEEE]. The mmWave bands are also lacking in technological maturity when it comes to delivering consumer market communication services. For widespread coverage, bands below 6GHz are needed.
• DT is estimating that a cell site based on 28GHz spectrum, which is the focus of pre-commercial 5G efforts by Verizon in the US, would have a diameter of only 80 meters [LR-2].

Technology Fragmentation

• In GSMA’s 2017 operator survey, 6% of the respondents ranked technology fragmentation as the most important risk in delivering 5G, making in the fifth biggest risk identified in the survey [GSMA-2]
• There is a concern that the different timing of operator agendas for 5G and pre-commercial 5G deployment together with the race to be the first to introduce 5G could lead to the deployment of non-standardized networks, interfaces and end-user equipment, leading to cost, battery life and interoperability issues, similar to what has happened with some of the earlier mobile generations [ALT].

5G Revenue

• Gartner believes that while Communication Service Providers (CSPs) are the ones who will need to invest in 5G network infrastructure, they may not be the main beneficiaries of the economic value 5G will create [Gartner].
• Strand Consult foresees that 5G will find it difficult to create economic growth, contrary to what is being commonly claimed [Strand]. According to them, 5G will not be a total loss, but will probably be a financial disappointment for most players. Strand consult goes as far as to consider 5G as “just an orgy of hype”.
• Analysys Mason sees a risk that the investment required by MNOs to deliver widespread IoT services might exceed their revenue unless additional spectrum is made available [AM-2].
• 5G is expected to require more vendor financing since MNOs will want the revenues first and then pay the costs [RW]. This means that the network vendors will need to find ways to provide more that type of funding.
• It will take some time for the tier 1 opreators to replenish their cash supply after 5G spectrum auction activities have finished [RW]. If MNOs continue to go after existing markets (e.g., smartphone users and Ultra-HD video delivery), they may never replenish their cash reserves, given that ARPU (Average Revenue Per User) numbers are headed down and less cash per customer is generated every year. In addition, according to Analysys Mason, the introduction of ever higher speeds and capacity in 4.5G networks has diminishing returns [AM]. Thus, 5G will need to generate revenue from other use cases beyond eMBB (enhanced Mobile Broadband) services. A solution to these issues is for MNOs to seek out new markets, that is, the vertical industries (see the IoT section below).
• According to DT, there is a strong likelihood that cost challenges will slow down the rate of 5G adoption, especially given the improvements being made in LTE [LR], which may make the rapid introduction of 5G less critical. DT’s CTO says that the industry will survive beyond 2020 because 4G will evolve further (i.e., LTE-Advanced and LTE-Advanced-Pro) and therefore “It is not going to be the big bang with 5G”.
• According to Analysys Mason, there is no clear indication that 5G will reverse the current declining mobile ARPU trends in Europe [AM-2]. The same could also be said about the North American market. Thus, a key issue in these markets, especially where there is strong infrastructure competition, is that operators have to consider major new technology investments in the context of declining revenues. If mobile operators maintain the current levels of investment, the resulting population coverage for 5G may not be sufficient to meet market needs in all countries (e.g., Germany, France and the UK). 
• Also according to Park Associates, 5G data service ARPU has limited room for expansion from current levels [PA]. Thus, mobile operators are seeking revenue streams in the broader IoT market. 
• Gartner points out that also a weak global economy could naturally delay 5G deployments [Gartner].

Cost of Rollout

• According to GSMA’s Mobile World Live 2017 industry survey, the biggest barrier for 5G deployment will be the cost of roll-out [MWL]. For mobile operators, 5G cost is a significant pressure coming rather soon after the industry’s multi-billion dollar investments in LTE. 
• Also, 5G may cost more than 4G simply based on historical trends, which show that each new “G” has required greater numbers of cell sites to account for signal propagation of higher-frequency signals and greater need for frequency reuse [IPC]. And 5G will certainly result in an increase in the number of cell sites due to mmWave spectrum and small cells.

IoT

• Analysys Mason claims that almost all the envisioned use cases for IoT can use a variant of LTE, or even something simpler, without needing a major network investment. This means that any business case for 5G investment cannot rely on IoT (and eMBB, as discussed above) alone [IBN]. The most commonly cited IoT application for 5G is the use of 5G for V2X (Vehicle-to-Everything) communications since it benefits from low latency. However, for example Nokia has stated that it has not yet found any vehicle platooning use case that would need 5G. Analysys Mason is also pointing out that IoT players are already developing solutions and are not waiting for the launch of 5G [TL].
• In 2016, revenue from IoT accounted for less than 1% of most operators’ total revenue. If assuming a CAGR (Compound Annual Growth Rate) of 20%, which is in line with the current growth rates of operators’ IoT businesses, it would take until 2025 for IoT to exceed 5% of operators’ total revenue [TL].
• The integration of connectivity within new industries poses multiple challenges, such as the need for expressing industry-specific KPIs (Key Performance Indicators) as network requirements [ALT]. This requires cross-industry collaboration and development.
• Gartner believes that a lower than expected demand for IoT could delay 5G deployments [Gartner].
• According to Rethink Wireless, the industry is attempting to sell a vision of 5G being the future of the networked enterprise and IoT [RW]. To achieve this, 5G has become an “unachievable science project” that tells everyone it can do everything with low cost. Rethink expects 5G to fail in its key aim, which is to keep cellular valued higher than other wireless technologies.
• Strand Consult finds it difficult to see big financial value in IoT for telecommunications companies that build and operate infrastructure [Strand]. This analysis, however, misses the development where some operators, like Verizon, are making acquisitions to build an IoT service business.

NFV, SDN and Cloud

• Parks Associates sees that it will not be easy for operators to transform their networks from hardware-centric to software-defined and virtualized, which is a transformation that is needed so that networks will have the flexibility to support the diverse use cases and unique requirements of industrial IoT verticals [PA].
• Some challenges that the transition to virtualized networks (SDN and NFV) brings include NFV infrastructure performance, workload configuration and multi-vendor integration, interoperability, and standardization [ALT]. The mobile operators are far behind the web players (e.g., Google, Facebook, Amazon, etc.) when it comes to the adoption maturity of SDN and virtualization technologies. Operators will need a mentality and cultural change to adopt DevOps, continuous-X, automation-oriented workflows and agile technologies together with the integration of network and IT operational tasks. This could turn out to be more challenging and time-consuming than expected.

Alternative Access Technologies

• In GSMA’s 2017 operator survey, 23% of the respondents ranked competition from alternative networks as the most important risk in delivering 5G, making it the third biggest risk identified in the survey [GSMA-2]
• Also Gartner sees that challenges from alternative access technologies, such as Wi-Fi could lead to the delay of 5G deployments [Gartner]. 
• According to Analysys Mason, 5G IoT use cases face challenges from LPWA (Low-Power Wide-Area) networks [AM]. The only use case that exploits the unique 5G capabilities might be critical IoT, but that is also the use case that is the farthest away in time.
• Parks Associates sees that IoT verticals’ unique requirements and diverse use cases open up market opportunities for alternative network providers like SIGFOX and Ingenu [PA].
• While Park Associates expects the global mobile data revenue to increase from $386 billion in 2015 to $630 billion in 2020, North America and Western Europe will see only minimal growth. Especially in the US, there is consumer interest towards alternatives like Wi-Fi to reduce mobile data bills. Cable operators are promoting Wi-Fi hotspots and roughly 70% of US mobile consumers are interested in Wi-Fi only data plans [PA]. Rethink Wireless cites a report that suggests that US cable operators could grab 20% of current MNO customers in the US by 2021 using Wi-Fi first models [RW]. This could make major US operators like Verizon and AT&T less profitable in the near term. The resulting debt rating misalignment would likely make the rapid buildout of 5G difficult to pay in practice.
• According to DT, the industry has to collaborate to drive the cost of 5G down. DT mentions Facebook’s Terragraph initiative (which uses spectrum in the 60 GHz range to provide high-speed connectivity in densely populated areas) as an example of the kind of innovation and partnership that has the potential to improve 5G economics [LR].
• Rethink Wireless forecasts about 13 million non-domestic LTE-Advanced small cells to be installed by 2020 [RW]. In contrast, Wi-Fi first opponents will be relying on 500 million homespots and perhaps 10 million public hotspots. Rethink Wireless believes that the advantage of Wi-Fi may not last forever, but it will last long enough to create financial issues for some operators.

Network Sharing

• Modeling done by Analysys Mason suggests that extending the coverage of a 5G MBB (Mobile Broadband) network to provide 100% road coverage along highways and motorways for ITS (Intelligent Transport System) use cases is not attractive for operators under assumptions of current market structure (i.e., multiple competing operator networks) [AM-2]. However, if competing operators collaborate to deploy a single shared infrastructure along major road routes, the business case can become viable.
• Deutsche Telekom has stated that it is looking at network sharing as a way to reduce the capex bill for 5G [LR].
• Another scenario where network sharing is likely to become necessary is indoor deployments of 5G small cells as it would be expensive and impractical for every operator to deploy their own indoor small cells in every building. This will require new partnerships and business models such as neutral host arrangements.
• Analysys Mason believes that a potential solution to the problem posed by diverse vertical industry 5G use cases and their different regulatory requirements could be greater use of common platforms between verticals and increased network sharing, including accessing infrastructure owned by public authorities to roll out 5G networks [AM-2].

5G Use Cases

• In GSMA’s 2017 operator survey, 51% of the respondents ranked lack of a clear business case as the most important risk in delivering 5G, making it the biggest risk identified in the survey [GSMA-2]
• BMI Research believes that there is yet no real business case for 5G services [BMI]
• According to Cisco, operators must solve the challenge of effectively monetizing video traffic (by 2021, 78% of the world’s mobile data traffic is expected to be video) while developing profitable business cases that support the investments needed for 5G infrastructure [Cisco].
• The functional requirements for some 5G vertical use cases are diverse (e.g., healthcare, transport, energy, etc.) and each industry vertical operates under different regulatory regimes. This creates complexities for MNOs [AM-2]. 
• Rethink Wireless believes that the industry needs to find new revenues for 5G [RW]. Ultra-HD video will not pay for 5G investments since people that use it have not shown intentions to pay and are happy consuming it over Wi-Fi. According to Rethink Wireless, never before in the history of the telecommunications industry has a technology been put forward to solve so many problems at once, without anyone being clear on who will pay for what and in what timeframe.
• New business models need to be developed for 5G [ALT]. With the emergence of NaaS (Network-as-a-Service) and network slicing, third parties can sell services relying on the MNO-owned networks. This will reshuffle the value chain and create new platform business models (e.g., a car company paying to an operator for a 5G network slice to be able to offer guaranteed connectivity services to their driver customers). The 5G network will become an integral part of business models in diverse vertical industries, which enables new actors to enter the resulting ecosystems. Operators need to assess the relationships with their current and future partners to optimally position themselves in the market and define their role in IoT.

Backhaul and Fronthaul

• One of the biggest issues for 5G might be the cost of backhaul and fronthaul [RW]. Rethink Wireless estimates that between 30-50 small cells will be needed per each current LTE macrocell (a more conservative estimate referred to earlier in this blog post is 10 [ARC], but the figure depends among other things on the frequency band used). This could imply a rise of more than 7 times compared to the current operational costs. Thus, technological innovations are needed for backhaul and backhaul also needs to be made much cheaper. However, there are no standards for backhaul.
• Cisco believes that backhaul capacity and efficiency must increase so that mobile broadband, data access, and video services can effectively support consumer usage trends and keep mobile infrastructure costs under control [Cisco].
• Ensuring regional connectivity might be one of the challenges for 5G since mobile networks have not earlier employed so high frequencies which require a high number of relay antennas to be installed [ARC]. This raises the question of the cost of deploying optical fiber to connect 5G radio or baseband units. Thus, the industry needs to design technologies that will make it possible to minimize the cost of 5G rollouts especially in rural areas.
• It is expected that backhaul will be a challenge in India for the migration from existing networks to 5G since less than 20% of Indian networks are running via fiber optic cables [ET]. Around 80% of cell sites in India are connected via microwave backhaul. Today's microwave backhaul does not provide enough bandwidth for 5G, whereas fiber-based backhaul can offer nearly unlimited capacity and low latency. According to Reliance Communications, strong backhaul is one of the fundamental requirements for 5G. Building backhaul is time-consuming and extremely expensive in the present conditions in India. Besides fiber-based backhaul, another option is installation of higher-capacity microwave links.

Regulation

• In GSMA’s 2017 operator survey, 20% of the respondents ranked regulation as the most important risk in delivering 5G, making it the fourth biggest risk identified in the survey [GSMA-2]
• According to several enterprises and electronic communication sector stakeholders, overly strong enforcement of network neutrality e.g., by BEREC (Body of European Regulators for Electronic Communications) could have a negative effect on the 5G rollout schedule [ARC]
• To overcome issues with the cost of 5G deployment, Analysys Mason recommends the regulatory environment for 5G to promote network and infrastructure sharing where it is commercially and technologically advantageous [AM-2].

5G Devices

• Rethink Wireless estimates that a standards-compliant 5G handset would today cost at least $1800 to build, since the handsets are more complex and offer more and more onboard radios and antennas so that the devices can speak throughout the 5G spectrum [RW]. Chips with MU-MIMO (Multi-User Massive Input Massive Output) logic built into them may be exceptionally expensive until they are produced in volumes of hundreds of millions. 5G devices may also have a dramatically lower battery life than LTE devices, at least initially. If the US moves early and thereby cements the 5G spectrum, it risks compromising the economies of scale that 5G devices will need.
• Rethink Wireless believes that the smartphone is ready to become a commodity and the sudden halt of smartphone growth starts to look more and more probable [RW]. This would mean that there is less money available for the mobile phone industry to invest in the R&D required by 5G.

References

[ALT] 5G as the enabler of cross-industry convergence, https://www.altran.com/fileadmin/medias/1.altran.com/files/PDF/Mobile_World_Congress_2017/Position_Papers/5G_as_the_enabler_of_cross-industry_convergence_White_paper.pdf

[AM] The investment case for 5G mobile is more distant without fixed wireless, http://www.analysysmason.com/Research/Content/Comments/5G-fixed-case-Aug2016-RDTW0-RDCS0/

[AM-2] Regulatory options to promote investment in 5G and IoT – a study for Qualcomm, https://www.techuk.org/component/techuksecurity/security/download/10241?file=Janette_Stewart_Analysys_Mason_-_Presentation_from_Analysys_Mason_Qualcomm_study_for_SPF_140217.pdf&Itemid=182

[ARC] 5G: Issues & Challenges, March 2017, http://www.arcep.fr/uploads/tx_gspublication/Report-5G-issues-challenges-march2017.pdf

[BMI] Early 5G Still Needs A Business Case, http://www.bmiresearch.com/articles/early-5g-still-needs-a-business-case

[Cisco] Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2016–2021 White Paper, http://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/mobile-white-paper-c11-520862.html

[ET] Migration to 5G will be challenging for India: Experts, http://economictimes.indiatimes.com/tech/migration-to-5g-will-be-challenging-for-india-experts/articleshow/58886148.cms

[Gartner] 5G — Who’s in the Driver’s Seat? http://www.gartner.com/smarterwithgartner/5g-whos-in-the-drivers-seat/

[GNW] TIA Survey Reveals Deployment Hurdles for Operators Before Commercial 5G in 2020, https://globenewswire.com/news-release/2017/03/16/939987/0/en/TIA-Survey-Reveals-Deployment-Hurdles-for-Operators-Before-Commercial-5G-in-2020.html

[GSMA] Uneven 4G adoption has implications for 5G outlook, https://www.gsmaintelligence.com/research/2017/04/uneven-4g-adoption-has-implications-for-5g-outlook/618/

[GSMA-2] The 5G era: Age of boundless connectivity and intelligent automation, https://www.gsmaintelligence.com/research/2017/02/the-5g-era-age-of-boundless-connectivity-and-intelligent-automation/614/

[IBN] MWC 2017: IoT has no compelling need for 5G – focus was on LTE variants for IoT, https://iotbusinessnews.com/2017/03/15/80790-mwc-2017-iot-no-compelling-need-5g-focus-lte-variants-iot/

[IEEE] Outdoor-to-indoor coverage in high frequency bands, http://ieeexplore.ieee.org/document/7063463/?arnumber=7063463

[IPC] 5G Will Cost More than 4G, But How Much More?, https://ipcarrier.blogspot.fi/2017/05/5g-will-cost-more-than-4g-but-how-much.html

[LR] DT CTO: Costs Must Fall or 5G 'Won't Work, http://www.lightreading.com/mobile/5g/dt-cto-costs-must-fall-or-5g-wont-work/d/d-id/730664

[LR-2] Vodafone CTO 'Worried' About 5G mmWave Hype, http://www.lightreading.com/mobile/spectrum/vodafone-cto-worried-about-5g-mmwave-hype/d/d-id/730679

[MWL] Mobile World Live, Annual Industry Suvey 2017, https://www.mobileworldlive.com/wp-content/uploads/2017/02/19308%20MWL%20Annual%20Survey.pdf

[PA] Global mobile data revenues to reach $630 billion by 2020, https://www.parksassociates.com/blog/article/pr-07112016 

[RW] The reasons why 5G can never happen, http://rethink-wireless.com/2016/07/19/the-reasons-why-5g-can-never-happen/

[Strand] 2017 Predictions and review of 2016, http://www.strandconsult.dk/sw7283.asp

[TL] Telecom operators strategies for IoT revealed by Analysys Mason, http://www.telecomlead.com/5g/telecoms-operators-strategies-for-iot-revealed-by-analysys-mason-73315