5G: A Reality check
|
by Detlef Eckert |
23 April 2021 |
Europeans, by now, should be able to enjoy 5G services. The industry should be deploying 5G services as part of digital transformation. The reality is different. Europe is behind its ambitions in deploying and using 5G.
Sure, the pandemic apparently prevented spectrum auctions in some countries. But, why were they so late in the first place? On the contrary, the pandemic should have motivated EU Member States to accelerate deployment and not slowing down.
Instead, the EU and member countries were busy to address 5G as a matter of national security concerns. We would talk about 5G security as an important but technical problem if it were not for the push of the US government against Huawei. Europe has become entangled into the US—China power struggle, even after Trump left the White House.
The German politician Norbert Röttgen referred to 5G as Europe’s digital nervous system that cannot be trusted to a Chinese company. This statement demonstrates both a limited understanding of 5G and the tense geo-political climate, associated with the rise of China as a global hi-tech power. Thus, let’s unpack 5G to assess where Europe stands today.
A quick look at 5G
What is 5G? What makes it faster than 4G? The short answer is more spectrum coupled with clever antenna technology (in particular massive MIMO, 3D beamforming). True, the 5G standard also contains a number of other improvements over 4G and adds useful features but spectrum and antennas generate the big leap. 5G shows its optimal performance when it uses the 3.5 GHz band, also called C-Band, as this spectrum offers an attractive combination of reach and speed/capacity. Although 5G can use a wide range of spectrum, its optimal use requires the C-Band.
Advanced antenna technology exploit the characteristic of the C band and are smaller, lighter, and individually more power efficient than the ones for 3 or 4G.
When you deploy 5G on 700 MHz users will not experience a significant increase in download speed. It will be better but by far not 10x the speed of 4G as you often read. When operators deploy 5G on 26 GHz, the mm wave, users will get high-speed connections but only at a short distance. Therefore, 5G at that spectrum will mainly support ‘Fixed Wireless Access’ (as an alternative to fibre to the home) or indoor scenarios, for instance, replacing Wi-Fi, and some city hotspots. This spectrum could also be interesting for industrial campus solutions, which has not really materialised yet.
For instance, so far the US operators have used 700 MHz and the mm Band. The US only recently allocated parts of the C-Band (auction raised more than 8 bn USD). Therefore, the US is not – as some reports suggest – a leader, but lagging behind. The leaders are to be found in Asia. More later.
Industry and policy makers do not get tired to stress that 5G is more than a faster 4G. The features of 5G are usually displayed as a triangle, with faster speed, larger device density, and low latency at the corners. Attached to these features are user scenarios, as depicted in the image below.
Sure, the pandemic apparently prevented spectrum auctions in some countries. But, why were they so late in the first place? On the contrary, the pandemic should have motivated EU Member States to accelerate deployment and not slowing down.
Instead, the EU and member countries were busy to address 5G as a matter of national security concerns. We would talk about 5G security as an important but technical problem if it were not for the push of the US government against Huawei. Europe has become entangled into the US—China power struggle, even after Trump left the White House.
The German politician Norbert Röttgen referred to 5G as Europe’s digital nervous system that cannot be trusted to a Chinese company. This statement demonstrates both a limited understanding of 5G and the tense geo-political climate, associated with the rise of China as a global hi-tech power. Thus, let’s unpack 5G to assess where Europe stands today.
A quick look at 5G
What is 5G? What makes it faster than 4G? The short answer is more spectrum coupled with clever antenna technology (in particular massive MIMO, 3D beamforming). True, the 5G standard also contains a number of other improvements over 4G and adds useful features but spectrum and antennas generate the big leap. 5G shows its optimal performance when it uses the 3.5 GHz band, also called C-Band, as this spectrum offers an attractive combination of reach and speed/capacity. Although 5G can use a wide range of spectrum, its optimal use requires the C-Band.
Advanced antenna technology exploit the characteristic of the C band and are smaller, lighter, and individually more power efficient than the ones for 3 or 4G.
When you deploy 5G on 700 MHz users will not experience a significant increase in download speed. It will be better but by far not 10x the speed of 4G as you often read. When operators deploy 5G on 26 GHz, the mm wave, users will get high-speed connections but only at a short distance. Therefore, 5G at that spectrum will mainly support ‘Fixed Wireless Access’ (as an alternative to fibre to the home) or indoor scenarios, for instance, replacing Wi-Fi, and some city hotspots. This spectrum could also be interesting for industrial campus solutions, which has not really materialised yet.
For instance, so far the US operators have used 700 MHz and the mm Band. The US only recently allocated parts of the C-Band (auction raised more than 8 bn USD). Therefore, the US is not – as some reports suggest – a leader, but lagging behind. The leaders are to be found in Asia. More later.
Industry and policy makers do not get tired to stress that 5G is more than a faster 4G. The features of 5G are usually displayed as a triangle, with faster speed, larger device density, and low latency at the corners. Attached to these features are user scenarios, as depicted in the image below.
The 3GPP – a cooperation of standardisation organisations and industry – has been releasing sets of specifications underpinning these scenarios. ‘Release 15’, finalised in 2019, laid the first building block, mainly specifying the broadband capabilities. In Summer 2020 3GPP published the final version of 'Release 16' with a focus on massive machine and low latency communications as well as improvements of mobile broadband. The organisation also determined the work programme for 'Release 17'.
These timelines already signal that a revolution is not around the corner but that deployment of 5G is a gradual process. So, where is Europe in 5G deployment?
Europe and Asia
The EU adopted an ambitious Action Plan in 2016. So far, however, this plan has not lived up to expectations. Many countries such as Poland, Portugal or Belgium, have not even assigned the C-Band, others were late. The mm band is almost nowhere available. Operators roll out of 5G moves in slow motion.
China has turned out the world-leader in absolute number of upgraded and newly installed 5G base stations. Korea remains the world-leader in terms of coverage ratio, with more than 40% and 60% of the geography and population, respectively, covered. Ten million+ Koreans have already subscribed to 5G services.
China’s three mobile operators have rolled out more than 700,000 5G sites by end of 2020 and they will add at least another 600,000 in 2021. They will also deploy Standalone 5G core networks which are essential for industrial applications. China Mobile, for instance, invested around 15 billion USD dollars in 2020 for 330,000 sites, serving almost 150 million 5G customers, about 15.5% of its customer base of about 945 million. The city of Shenzhen with 47,000 5G base stations (more than the entire EU) is fully covered.
Compare this to Germany. Beginning 2021, Deutsche Telekom has 50,000 antennas (approximately 16,000 sites as usually there are three antennas per site) operating 5G, of which only 1,400 use the C-Band. Vodafone Germany enabled 8,500 antennas, of which 1,000 C-Band antennas (4,000 are planned by end of 2021). Telefonica in Germany deployed 1000 antennas on about 330 base stations, all activated on the C band. So far the core networks of all three operators still run on 4G, although according to announcements this is supposed to change during the course of 2021.
It should become good reporting practice to distinguish between software upgrades and antenna deployments. 5G roll-out communication should contain data about the number of massive MIMO antennas using the C-Band. In addition, meaningful comparisons require information how many networks are 5G stand-alone and how many still operate with a 4G core.
5G Observatory – Tracking 5G developments
5G: A driver for Industry 4.0?
The short answer is not yet. First, as mentioned above, the 3GPP has not released all relevant standards yet. In addition, advanced industrial applications, for instance, enabling extremely low latency or network slicing, require a stand-alone 5G core.
Second, more importantly, you need additional investments, in particular mobile edge computing which manages the devices on an industrial campus, road infrastructure or in smart cities. Without having computing close to the ‘action’ (e.g. robots moving around) the user will not get the required latency.
Finally, you need ‘intelligent’ software that controls the action, which typically runs on servers owned by the company or an organisation. All of this costs money which demands a sound business case. When you listen to the industry stories, there is no shortage of fancy scenarios, such as 5G controlled drones, automatic driving, robotics, remote surgery, etc.
https://etno.live.ft.com/agenda/session/316111
There is no shortage of studies that predict huge economic benefits either. In November 2020 Analysys Mason, commissioned by Ericsson and Quantum Technologies, found that manufacturing would be the greatest potential for 5G in the UK. A Huawei study, carried out by STL Partners, on manufacturing estimated a gain of 740 billion USD globally by 2030. DLA Piper’s European Tech Index report saw great potential in smart cities.
We should take all these studies with a grain of salt. Have you ever heard about a new digital technology that is not supposed to generate a revolution? Policy makers love to take up these visions, but ironically without always taking action. There is huge gap between ideas, proof of concepts and scaled-up business implementations. How many of 5G scenarios are already feasible with existing technologies such as 4G, Wi-Fi, Zigbee or NB-IoT? These are not a trivial questions as companies would have to justify investments to use 5G.
At the moment, we are still in an exploratory phase. For instance, Siemens has set up a demonstration lab to show case how to connect a smart factory to the public 5G. Ericsson and Telefonica developed with Mercedes in Sindelfingen (Factory 56) a private 5G campus solution.
https://www.daimler.com/innovation/production/5g-network-production.html
https://new.siemens.com/global/en/products/automation/industrial-communication/5g.html
The company Bosch launched a 5G trial in its Worchester/UK factory, for instance, testing sensors for preventative maintenance and real-time feedback.
https://www.bosch.co.uk/internet-of-things/leading-in-5g-technology/
Whilst autonomous driving does not require 5G, it will be helpful to enhance safety, to manage traffic, and offer services to drivers. 4G already supports V2X applications and 5G will add useful features.
Going forward, the industry will improve 5G technology. Thus, some scenarios might become more competitive against alternatives. Recently, Huawei announced a 5.5G concept with faster uplinks and downlinks, high-precision indoor positioning capabilities, and wide-area high-resolution sensing.
Security
Policy makers talk a lot about security, and rightly so, only a bit more understanding of the issues would help. A 5G network is not a monolithic block. Broadly speaking, a mobile network comprises a radio access (antennas and some boxes) -short RAN -, a transport (fibre, micro-wave) and a core (servers) network. Operators prefer a multi-vendor strategy, mainly not to become too dependent on a dominant supplier, which is also encouraged by policy makers. However, buying from many vendors also implies more complexity when managing the network. This will become a big issue if Open RAN becomes more popular. Do not expect Open RAN to be more secure just because it is called ‘open’.
The core network is probably the most sensitive part as it stores and processes data, user identification, gateways, billing or cryptographic keys. Therefore, a proper risk assessment should disentangle the network, identify areas that need more protection than others and apply mitigation tools. For instance, generation, storage and protection of crypto-graphic keys should be a focus point, which is hardly mentioned in the political debate, and does not show up in the security tool box recommended by the European Commission. Otherwise, this tool box contains a comprehensive risk assessment but at a price of complexity. In reading it I was reminded of Einstein's quote: “Everything should be made as simple as possible, but no simpler.”
The EU toolbox for 5G security | Shaping Europe’s digital future (europa.eu)
The risk profile of a campus solution also depends how the network is built. We can distinguish two scenarios. An organisation, for instance, a company, can build a private 5G network or use the public one. The company can acquire both edge computing and software from different vendors and need not rely on an operator or one vendor.
Occasionally you hear the argument that 5G RAN cannot be separated from the industrial applications and might constitute a risk not present in 4G. However, based on 3GPP specifications 5G RAN can be perfectly detached from the edge network and the core functions hosted within. The biggest challenge does not lie in separating the RAN but in securing the applications running on 5G networks. This is the task which deserves much more attention.
To summarise:
5G exploits additional spectrum through innovative antenna technology. It does not automatically (out of the box) unfolds its potential, i.e. super-fast download, millions of connected devices, low latency, which needs additional investments.
5G is not a monolithic block. It comprise sub-networks, i.e. access, transport and core, and many components provided by multiple vendors and integrated and managed by operators.
Industrial applications need not only the network, but also edge computing and intelligent software. Diversification of vendors and competition will drive this eco-system. Operators must manage complexities.
When addressing security, policy makers should pay more attention to securing the core through crypto-graphic means and the applications.
A gradual upgrade of 4G to 5G makes business sense. It helps operators to manage the increasing data load generated by video and gaming. At some point, they need to deploy stand-alone 5G and engage more deeply with verticals.
We are in a phase where 5G is a consumer business, with still limited incremental utility for consumers. Many ‘fancy’ user applications, such as virtual reality gaming, are rather stationary applications.
Making use of 5G as an enabler for industry 4.0 is no automatism, and policy makers should help by providing more spectrum and procurement for public use.
|
Epilogue: 6G
Although operators still have a lot to do with 5G deployment and will use 4G as the work horse for years to come, research to develop 6G is underway. Politically, this has quickly become a priority, again, mainly because of geo-politics. The US would like to re-gain a strong position in mobile communications. China aims to build on the progress it has made with 4G and pursues a vision of a hi-tech infrastructure. Europe is anxious not to lose its two champions, Nokia and Ericsson. |
|
The EU aspires to become digitally more sovereign. It has launched a major 6G research project without participation of a Chinese vendor. Germany has earmarked 100 million € for 6G research.
Thus, 6G has very early on become entangled in political rivalry and re-positioning of industry policies. Will this endanger a common global 6G standard? Will 6G turn out to become an example of technological decoupling?
4G has become a win-win for the world. With 5G we are observing how industry policy and geo-politics have become a mélange that delays applications and raises mistrust. To change this course with 6G we need a better geo-political climate, one which only optimists can see right now.
What could 6G technically become? Research teams have already posted some surprisingly concrete ideas. First of all, you need to start with spectrum. Broadly speaking 6G will exploit the spectrum between 100 GHz and 1,000 GHz (1 THz). The World Radio Conferences (WRC) in 2023 and in 2027 will set the direction which and how much spectrum could be made available for 6G.
This spectrum will allow much higher data rates, but at considerably shorter distances. This allows significantly smaller antennas. Researchers have demonstrated chips capable to handle these extremely short frequencies. Unlike 5G, the characteristics of such a 6G network will be remarkably different from the previous generations of mobile communication.
For instance, the cell structure will be replaced by immersive radio access. Antennas will be tiny and there will be lots of them. This will allow cell-free, ultra-massive MIMO, not necessarily visible as antennas but embedded everywhere, for instance, entire building facades could be used as radio reflectors. We will end up with extreme connectivity. Such an architecture will lead to a merger of radio access network with the core network. Devices could be powered wireless, i.e. over the air interface, replacing batteries.
The properties of the spectrum will allow the network to function as a sensor (sensing as a service). The radio could, for instance, measure particles to manage air quality or to carry out precision localisation controlling indoor drones or other vehicles.
Researchers agree that the air interface and network management will be controlled by artificial intelligence. Recall, 5G was introduced as a cloud native network, 6G will be AI native. AI would also support services, for instance, algorithms to identify objects. In the extreme you could generate a digital (hologram) twin of any place by combining the network sensing capabilities with AI object recognition.
Some expect that 6G will also use visible light for communication. Another advanced concept is to secure 6G communication by quantum cryptography, which might be needed given, for instance, the sensing capabilities.
If you ask for applications or which problems 6G could tackle, there is considerable scope for speculation. Some refer to high-fidelity holograms or turning entire walls into an 8K video screen. Yet, as shown above with 5G, mobile networks are platforms that, on the one hand, will be used in ways not initially predicted but, on the other hand, expected applications may not materialise. One thing, however, is certain, by pushing the technology frontier 6G will open avenues for innovation.
When you look at the envisaged technology and architecture of 6G, it is not difficult to predict that European techno skepticism will emerge soon. NGOs and concerned citizens will mount resistance and voice concerns about radiation and privacy intrusion as soon as the specifications of 6G become clearer. Keeping up entrepreneurial spirit and promoting enthusiasm for cutting edge mobile communications will be at least as important as funding research projects.
Thus, 6G has very early on become entangled in political rivalry and re-positioning of industry policies. Will this endanger a common global 6G standard? Will 6G turn out to become an example of technological decoupling?
4G has become a win-win for the world. With 5G we are observing how industry policy and geo-politics have become a mélange that delays applications and raises mistrust. To change this course with 6G we need a better geo-political climate, one which only optimists can see right now.
What could 6G technically become? Research teams have already posted some surprisingly concrete ideas. First of all, you need to start with spectrum. Broadly speaking 6G will exploit the spectrum between 100 GHz and 1,000 GHz (1 THz). The World Radio Conferences (WRC) in 2023 and in 2027 will set the direction which and how much spectrum could be made available for 6G.
This spectrum will allow much higher data rates, but at considerably shorter distances. This allows significantly smaller antennas. Researchers have demonstrated chips capable to handle these extremely short frequencies. Unlike 5G, the characteristics of such a 6G network will be remarkably different from the previous generations of mobile communication.
For instance, the cell structure will be replaced by immersive radio access. Antennas will be tiny and there will be lots of them. This will allow cell-free, ultra-massive MIMO, not necessarily visible as antennas but embedded everywhere, for instance, entire building facades could be used as radio reflectors. We will end up with extreme connectivity. Such an architecture will lead to a merger of radio access network with the core network. Devices could be powered wireless, i.e. over the air interface, replacing batteries.
The properties of the spectrum will allow the network to function as a sensor (sensing as a service). The radio could, for instance, measure particles to manage air quality or to carry out precision localisation controlling indoor drones or other vehicles.
Researchers agree that the air interface and network management will be controlled by artificial intelligence. Recall, 5G was introduced as a cloud native network, 6G will be AI native. AI would also support services, for instance, algorithms to identify objects. In the extreme you could generate a digital (hologram) twin of any place by combining the network sensing capabilities with AI object recognition.
Some expect that 6G will also use visible light for communication. Another advanced concept is to secure 6G communication by quantum cryptography, which might be needed given, for instance, the sensing capabilities.
If you ask for applications or which problems 6G could tackle, there is considerable scope for speculation. Some refer to high-fidelity holograms or turning entire walls into an 8K video screen. Yet, as shown above with 5G, mobile networks are platforms that, on the one hand, will be used in ways not initially predicted but, on the other hand, expected applications may not materialise. One thing, however, is certain, by pushing the technology frontier 6G will open avenues for innovation.
When you look at the envisaged technology and architecture of 6G, it is not difficult to predict that European techno skepticism will emerge soon. NGOs and concerned citizens will mount resistance and voice concerns about radiation and privacy intrusion as soon as the specifications of 6G become clearer. Keeping up entrepreneurial spirit and promoting enthusiasm for cutting edge mobile communications will be at least as important as funding research projects.