Sunday 29 July 2018

Automating the 5G Core using Machine Learning and Data Analytics

One of the new entities introduced by 3GPP in the 5G Core SBA (see tutorial here) is Network Data Analytics Function, NWDAF.
3GPP TR 23.791: Study of Enablers for Network Automation for 5G (Release 16) describes the following 5G Network Architecture Assumptions:

1 The NWDAF (Network Data Analytics Function) as defined in TS 23.503 is used for data collection and data analytics in centralized manner. An NWDAF may be used for analytics for one or more Network Slice.
2 For instances where certain analytics can be performed by a 5GS NF independently, a NWDAF instance specific to that analytic maybe collocated with the 5GS NF. The data utilized by the 5GS NF as input to analytics in this case should also be made available to allow for the centralized NWDAF deployment option.
3 5GS Network Functions and OAM decide how to use the data analytics provided by NWDAF to improve the network performance.
4 NWDAF utilizes the existing service based interfaces to communicate with other 5GC Network Functions and OAM.
5 A 5GC NF may expose the result of the data analytics to any consumer NF utilizing a service based interface.
6 The interactions between NF(s) and the NWDAF take place in the local PLMN (the reporting NF and the NWDAF belong to the same PLMN).
7 Solutions shall neither assume NWDAF knowledge about NF application logic. The NWDAF may use subscription data but only for statistical purpose.

Picture SourceApplication of Data Mining in the 5G Network Architecture by Alexandros Kaloxylos

Continuing from 3GPP TR 23.791:

The NWDAF may serve use cases belonging to one or several domains, e.g. QoS, traffic steering, dimensioning, security.
The input data of the NWDAF may come from multiple sources, and the resulting actions undertaken by the consuming NF or AF may concern several domains (e.g. Mobility management, Session Management, QoS management, Application layer, Security management, NF life cycle management).
Use case descriptions should include the following aspects:
1. General characteristics (domain: performance, QoS, resilience, security; time scale).
2. Nature of input data (e.g. logs, KPI, events).
3. Types of NF consuming the NWDAF output data, how data is conveyed and nature of consumed analytics.
4. Output data.
5. Possible examples of actions undertaken by the consuming NF or AF, resulting from these analytics.
6. Benefits, e.g. revenue, resource saving, QoE, service assurance, reputation.

Picture SourceApplication of Data Mining in the 5G Network Architecture by Alexandros Kaloxylos

3GPP TS 23.501 V15.2.0 (2018-06) Section 6.2.18 says:

NWDAF represents operator managed network analytics logical function. NWDAF provides slice specific network data analytics to a NF. NWDAF provides network analytics information (i.e., load level information) to a NF on a network slice instance level and the NWDAF is not required to be aware of the current subscribers using the slice. NWDAF notifies slice specific network status analytic information to the NFs that are subscribed to it. NF may collect directly slice specific network status analytic information from NWDAF. This information is not subscriber specific.

In this Release of the specification, both PCF and NSSF are consumers of network analytics. The PCF may use that data in its policy decisions. NSSF may use the load level information provided by NWDAF for slice selection.

NOTE 1: NWDAF functionality beyond its support for Nnwdaf is out of scope of 3GPP.
NOTE 2: NWDAF functionality for non-slice-specific analytics information is not supported in this Release of the specification.

3GPP Release-16 is focusing on 5G Expansion and 5G Efficiency, SON and Big Data are part of 5G Efficiency.
Light Reading Artificial Intelligence and Machine Learning section has a news item on this topic from Layer123's Zero Touch & Carrier Automation Congress:

The 3GPP standards group is developing a machine learning function that could allow 5G operators to monitor the status of a network slice or third-party application performance.

The network data analytics function (NWDAF) forms a part of the 3GPP's 5G standardization efforts and could become a central point for analytics in the 5G core network, said Serge Manning, a senior technology strategist at Sprint Corp.

Speaking here in Madrid, Manning said the NWDAF was still in the "early stages" of standardization but could become "an interesting place for innovation."

The 3rd Generation Partnership Project (3GPP) froze the specifications for a 5G new radio standard at the end of 2017 and is due to freeze another set of 5G specifications, covering some of the core network and non-radio features, in June this year as part of its "Release 15" update.

Manning says that Release 15 considers the network slice selection function (NSSF) and the policy control function (PCF) as potential "consumers" of the NWDAF. "Anything else is open to being a consumer," he says. "We have things like monitoring the status of the load of a network slice, or looking at the behavior of mobile devices if you wanted to make adjustments. You could also look at application performance."

In principle, the NWDAF would be able to make use of any data in the core network. The 3GPP does not plan on standardizing the algorithms that will be used but rather the types of raw information the NWDAF will examine. The format of the analytics information that it produces might also be standardized, says Manning.

Such technical developments might help operators to provide network slices more dynamically on their future 5G networks.

Generally seen as one of the most game-changing aspects of 5G, the technique of network slicing would essentially allow an operator to provide a number of virtual network services over the same physical infrastructure.

For example, an operator could provide very high-speed connectivity for mobile gaming over one slice and a low-latency service for factory automation on another -- both reliant on the same underlying hardware.

However, there is concern that without greater automation operators will have less freedom to innovate through network slicing. "If operators don't automate they will be providing capacity-based slices that are relatively large and static and undifferentiated and certainly not on a per-customer basis," says Caroline Chappell, an analyst with Analysys Mason .

In a Madrid presentation, Chappell said that more granular slicing would require "highly agile end-to-end automation" that takes advantage of progress on software-defined networking and network functions virtualization.

"Slices could be very dynamic and perhaps last for only five minutes," she says. "In the very long term, applications could create their own slices."

Despite the talk of standardization, and signs of good progress within the 3GPP, concern emerged this week in Madrid that standards bodies are not moving quickly enough to address operators' needs.

Caroline Chappell's talk is available here whereas Serge Manning's talk is embedded below:



I am helping CW organise the annual CW TEC conference on the topic The inevitable automation of Next Generation Networks
Communications networks are perhaps the most complex machines on the planet. They use vast amounts of hardware, rely on complex software, and are physically distributed over land, underwater, and in orbit. They increasingly provide essential services that underpin almost every aspect of life. Managing networks and optimising their performance is a vast challenge, and will become many times harder with the advent of 5G. The 4th Annual CW Technology Conference will explore this challenge and how Machine Learning and AI may be applied to build more reliable, secure and better performing networks.

Is the AI community aware of the challenges facing network providers? Are the network operators and providers aware of how the very latest developments in AI may provide solutions? The conference will aim to bridge the gap between AI/ML and communications network communities, making each more aware of the nature and scale of the problems and the potential solutions.

I am hoping to see some of this blog readers at the conference. Looking forward to learning more on this topic amongst others for network automation.

Related Post:

Tuesday 24 July 2018

Multicast Operation on Demand (MooD) and Service Continuity for eMBMS


Many regular readers of this blog are aware that back in 2014 I wrote a post looking critically at LTE-Broadcast business case and suggested a few approaches to make it a success. Back in those days, 2014 was being billed as the year of LTE-Broadcast or eMBMS (see here and here for example). I was just cautioning people against jumping on the LTE-B bandwagon.

According to a recent GSA report 'LTE Broadcast (eMBMS) Market Update – March 2018':

  • thirty-nine operators are known to have been investing in eMBMS demonstrations, trials, deployments or launches
  • five operators have now deployed eMBMS or launched some sort of commercial service using eMBMS

Its good to see some operators now getting ready to deploy eMBMS for broadcast TV scenarios. eMBMS will also be used in Mission Critical Communications for the features described here.

In a recent news from the Australian operator Telstra:

Telstra is now streaming live sports content to a massive base of around 1.2 million devices each weekend and sports fans consume 37 million minutes of live content over our apps on any given weekend.

This increase brings new challenges to the way traffic on our mobile network is managed. Even though a large group of people might be streaming the same real-time content at the same time, we still need to ensure a high quality streaming experience for our customers.

This challenge makes our sporting apps a prime use case for LTE-Broadcast (LTE-B).

Earlier this year, we announced we would be turning on LTE-B functionality on the AFL Live Official app for Telstra customers with Samsung Galaxy S8 and Galaxy S9 devices. Following extensive testing, Telstra is the only operator in Australia – and one of the first in the world – to deploy LTE-B into its mobile network.

At a live demonstration in Sydney, over 100 Samsung Galaxy S8 and Galaxy S9 devices were on display showing simultaneous high definition content from the AFL Live Official app using LTE-B.

Its interesting to note here that the broadcast functionality (and probably intelligence) is built into the app.

According to another Telstra news item (emphasis mine):

The use of LTE-Broadcast technology changes the underlying efficiency of live video delivery as each cell can now support an unlimited number of users watching the same content with improved overall quality. To date though, LTE-B technology has required that a dedicated part of each cell’s capacity be set aside for broadcasting. This had made the LTE-B business case harder to prove in for lower streaming demand rates.

This has now changed as Telstra and our partners have enabled the world’s first implementation of the Multicast Operation on Demand (MooD) feature whereby cells in the network only need to configure for LTE-B when there are multiple users watching the same content.

This combined with the Service Continuity feature allows mobile users to move around the network seamlessly between cells configured for LTE-B and those which are not.

Earlier this year we announced our intention to enable LTE-Broadcast (LTE-B) across our entire mobile network in 2018. With MooD and service continuity we are one step closer to that goal as we head into another year of major growth in sporting content demand.

Supported by technology partners Ericsson and Qualcomm, Telstra has now delivered world first capability to ensure LTE-B can be delivered as efficiently as possible.

Service Continuity will allow devices to transition in and out of LTE-B coverage areas without interruption. For instance, you might be at a music festival streaming an event on your phone but need to leave the venue and make your way back home (where LTE-B is not in use). Service Continuity means you can continue to watch the stream and the transition will be seamless – even though you have the left the broadcast area.

Taking that a step further, MooD allows the network to determine how many LTE-B compatible devices in any given area are consuming the same content. MooD then intelligently activates or deactivates LTE-B, ensuring the mobile network is as efficient as possible in that location.

For example, if a die-hard football fan is streaming a match we will likely service that one user with unicast, as that is the most efficient way of delivering the content. However if more users in the same cell decide to watch the match, MooD makes the decision automatically as to whether it is more efficient to service those users by switching the stream to broadcasting instead of individual unicast streams.

Its good to see Ericsson & Qualcomm finally taking eMBMS to commercial deployment. Back in 2015, I added their videos from MWC that year. See post here.
I think the Telstra post already provides info on why MooD is needed but this picture from Qualcomm whitepaper above makes it much clearer. Back in 3G MBMS and early days or eMBMS, there used to be a feature called counting, MooD is effectively doing the same thing.
For Service Continuity, this paper 'Service Continuity for eMBMS in LTE/LTE-Advanced Network: Standard Analysis and Supplement' by Ngoc-Duy Nguyen and Christian Bonnet has interesting proposal on how it should be done. I cannot be sure if this is correct as per the latest specifications but its interesting to learn how this would be done when the user moves out of coverage area in Idle or connected mode.

Note that this Expway paper also refers to Service continuity as Session continuity.

Related posts:



Thursday 19 July 2018

5G Synchronisation Requirements


5G will probably introduce tighter synchronization requirements than LTE. A recent presentation from Ericsson provides more details.

In frequencies below 6GHz (referred to as frequency range 1 or FR1 in standards), there is a probability to use both FDD and TDD bands, especially in case of re-farming of existing bands. In frequencies above 6GHz (referred to as frequency range 2 or FR2 in standards, even though FR2 starts from 24.25 GHz), it is expected that all bands would be TDD.

Interesting to see that the cell phase synchronization accuracy measured at BS antenna connectors is specified to be better than 3 μs in 3GPP TS 38 133. This translates into a network-wide requirements of +/-1.5 microseconds and is applicable to both FR1 and FR2, regardless of the cell size.

Frequency Error for NR specified in 3GPP TS 38.104 states that the base station (BS) shall be accurate to within the following accuracy range observed over 1 ms:
Wide Area BS → ±0.05 ppm
Medium Range BS → ±0.1 ppm
Local Area BS → ±0.1 ppm

The presentation specifies that based on request by some operators, studies in ITU-T on the feasibility of solutions targeting end-to-end time synchronization requirements on the order of +/-100 ns to +/-300 ns

There is also a challenge of how the sync information is transported within the network. The conclusion is that while the current LTE sync requirements would work in the short term, new solutions would be required in the longer term.

If this is an area of interest, you will also enjoy watching CW Heritage SIG talk by Prof. Andy Sutton, "The history of synchronisation in digital cellular networks". Its available here.

Thursday 12 July 2018

Minimum Bandwidth Requirement for 5G Non-Standalone (NSA) Deployment

I was attending the IEEE 5G World Forum live-stream, courtesy of IEEE Tv and happen to hear Egil Gronstad, Senior Director of Technology Development and Strategy at T-Mobile USA. He said that they will be building a nationwide 5G network that will initially be based on 600 MHz band.


During the Q&A, Egil mentioned that because of the way the USA has different markets, on average they have 31 MHz of 600 MHz (Band 71). The minimum is 20 MHz and the maximum is 50 MHz.

So I started wondering how would they launch 4G & 5G in the same band for nationwide coverage? They have a good video on their 5G vision but that is of course probably going to come few years down the line.

In simple terms, they will first deploy what is known as Option 3 or EN-DC. If you want a quick refresher on different options, you may want to jump to my tutorial on this topic at 3G4G here.

The Master Node (recall dual connectivity for LTE, Release-12. See here) is an eNodeB. As with any LTE node, it can take bandwidths from 1.4 MHz to 20 MHz. So the minimum bandwidth for LTE node is 1.4 MHz.

The Secondary Node is a gNodeB. Looking at 3GPP TS 38.101-1, Table 5.3.5-1 Channel bandwidths for each NR band, I can see that for band 71


NR band / SCS / UE Channel bandwidth
NR Band
SCS
kHz
5 MHz
101,2 MHz
152 MHz
202 MHz
252 MHz
30 MHz
40 MHz
50 MHz
60 MHz
80 MHz
90 MHz
100 MHz
n71
15
Yes
Yes
Yes
Yes








30

Yes
Yes
Yes








60













The minimum bandwidth is 5MHz. Of course this is paired spectrum for FDD band but the point I am making here is that you need just 6.4 MHz minimum to be able to support the Non-Standalone 5G option.

I am sure you can guess that the speeds will not really be 5G speeds with this amount of bandwidth but I am looking forward to all these kind of complaints in the initial phase of 5G network rollout.

I dont know what bandwidths T-Mobile will be using but we will see at least 10MHz of NR in case where the total spectrum is 20 MHz and 20 MHz of NR where the total spectrum is 50 MHz.

If you look at the earlier requirements list, the number being thrown about for bandwidth was 100 MHz for below 6 GHz and up to 1 GHz bandwidth for spectrum above 6 GHz. Don't think there was a hard and fast requirement though.

Happy to hear your thoughts.

Tuesday 3 July 2018

Terahertz and Beyond 100 GHz progress

There seems to be a good amount of research going on in higher frequencies to see how a lot more spectrum with a lot more bandwidth can be used in future radio communications. NTT recently released information about "Ultra high-speed IC capable of wireless transmission of 100 gigabits per second in a 300 GHz band". Before we discuss anything, lets look at what Terahertz means from this article.

Terahertz wave: Just as we use the phrase ‘kilo’ to mean 103 , so we use the term ‘giga’ to mean 109 and the term ‘tera’ to mean 1012 . “Hertz (Hz)” is a unit of a physical quantity called frequency. It indicates how many times alternating electric signals and electromagnetic waves change polarity (plus and minus) per second. That is, one terahertz (1 THz = 1,000 GHz) is the frequency of the electromagnetic wave changing the polarity by 1 × 1012 times per second. In general, a terahertz wave often indicates an electromagnetic wave of 0.3 THz to 3 THz.

While there are quite a few different numbers, this is the one that is most commonly being used. The following is the details of research NTT did.

In this research, we realized 100 Gbps wireless transmission with one wave (one carrier), so in the future, we can extend to multiple carriers by making use of the wide frequency band of 300 GHz band, and use spatial multiplexing technology such as MIMO and OAM. It is expected to be an ultra high-speed IC technology that enables high-capacity wireless transmission of 400 gigabits per second. This is about 400 times the current LTE and Wi-Fi, and 40 times 5G, the next-generation mobile communication technology. It is also expected to be a technology that opens up utilization of the unused terahertz wave frequency band in the communications field and non-communication fields.

Complete article and paper available here.

Huawei has also been doing research in W (92 - 114.5 GHz) and D (130 - 174.5 GHz) bands.


A recent presentation by Debora Gentina, ETSI ISG mWT WI#8 Rapporteur at the UK Spectrum Policy Forum is embedded below.



This presentation can be downloaded from UK SPF site here. Another event on beyond 100GHz that took place last year has some interesting presentations too. Again, on UKSPF site here.


Ericsson has an interesting article in Technology Review, looking at beyond 100GHz from backhaul point of view. Its available here.

If 5G is going to start using the frequencies traditionally used by backhaul then backhaul will have to start looking at other options too.

Happy to listen to your thoughts and insights on this topic.

Monday 25 June 2018

Free Apps for Field Testing - Part 2

The last time I wrote about the free apps for field testing, many people came back and suggested additional apps that are much more commonly used. In fact we got the following comment when 3G4G re-posted this

As I have used both these apps frequently, here is a small summary on them.

Network Signal Guru: This is surprisingly very popular and is quite useful. The only issue is that you need to have a rooted phone with Qualcomm chipset. I know many testers have their favourite phones and quite a few testers buy the latest phones, root them and start testing using NSG (Network Signal Guru).

I prefer using Motorola Moto Gx series phones. They are cheap, not too difficult to root (YouTube have quite a few tutorials and Google search works too) and I find that their receivers are better than others. Have detected cells that other phones cant and have even camped and speed tested on them too.

So what can NSG do?

It can provide lots of useful information on the physical layer, cell configurations, neighbor cell lists, MIMO, etc.
You can even RAT lock to LTE / WCDMA / GSM and band lock to use a specific band. It can be very useful during surveys when you want to check if you can see particular frequency anywhere in an area. You can also see Codecs, RACH information, Data information, etc.

Finally, one of the best things I find is the signalling information. Some of the details are only available for purchased option, its nevertheless very useful. Just in case you are wondering how much does it cost, its roughly £50 per month license in UK.


Cell Mapper: I find this much more helpful as it can be used without rooting. CellMapper is a crowd-sourced cellular tower and coverage mapping service. Its simple and only used for basic testing but nevertheless very useful. To give you an idea, the other day I was camped on a cell with very good signal quality but very poor data rates and there weren't many people so congestion didn't seem like a factor. On investigation I found out that I was camped on 800MHz band that has limited bandwidth per operator and there was no CA.

Cell mapper, as you can see provides information about the cell you are camped on, the cell tower location, what other sectors and frequencies are there, etc.


Do you have a favorite testing app that I missed? Let me know in comments.

Friday 22 June 2018

5G and IoT Security Update from ETSI Security Week 2018

ETSI Security Week 2018 (link) was held at ETSI's Headquarters in Sophia Antipolis, South of France last week. It covered wide variety of topics including 5G, IoT, Cybersecurity, Middlebox, Distributed Ledger Technology (DLT), etc. As 5G and IoT is of interest to the readers of this blog, I am providing links to the presentations so anyone interested can check them out at leisure.


Before we look at the presentations, what exactly was the point of looking at 5G Security? Here is an explanation from ETSI:

5G phase 1 specifications are now done, and the world is preparing for the arrival of 5G networks. A major design goal of 5G is a high degree of flexibility to better cater for specific needs of actors from outside the telecom sector (e.g. automotive industry, mission-critical organisations). During this workshop, we will review how well 5G networks can provide security for different trust models, security policies, and deployment scenarios – not least for ongoing threats in the IoT world. 5G provides higher flexibility than legacy networks by network slicing and virtualization of functions. The workshop aims to discuss how network slicing could help in fulfilling needs for different users of 5G networks.

5G will allow the use of different authentication methods. This raises many interesting questions. How are these authentication methods supported in devices via the new secure element defined in ETSI SCP, or vendor-specific concepts? How can mission-critical and low-cost IoT use cases coexist side-by-side on the same network?

The 5G promise of higher flexibility is also delivered via its Service-Based Architecture (SBA). SBA provides open 3rd party interfaces to support new business models which allow direct impact on network functions. Another consequence of SBA is a paradigm shift for inter-operator networks: modern APIs will replace legacy signaling protocols between networks. What are the relevant security measures to protect the SBA and all parties involved? What is the role of international carrier networks like IPX in 5G?

Event Objectives
The workshop intends to:

  • Gather different actors involved in the development of 5G, not only telecom, and discuss together how all their views have shaped phase 1 of 5G, to understand how security requirements were met, and what challenges remain;
  • Discuss slicing as a means to implement separate security policies and compartments for independent tenants on the same infrastructure;
  • Give an update of what is happening in 3GPP 5G security;
  • Explain to IoT players what 5G security can (and cannot) do for them, including risks and opportunities related to alternative access credentials;
  • Understand stakeholders' (PMNs, carriers, GSMA, vendors) needs to make SBA both secure and successful. How can SBA tackle existing issues in interconnect networks like fraud, tracking, privacy breaches;
  • Allow vendors to present interesting proposals for open security questions in 5G: secure credential store, firewalling SBA's RESTful APIs;
  • Debate about hot topics such as: IoT security, Slicing security, Privacy, Secure storage and processing and Security of the interconnection network.


So here are the relevant presentations:

Session 1: Input to 5G: Views from Different Stakeholders
Session Chair: Bengt Sahlin, Ericsson

Hardening a Mission Critical Service Using 5G, Peter Haigh, NCSC

Security in the Automotive Electronics Area, Alexios Lekidis, SecurityMatters

Integrating the SIM (iUICC), Adrian Escott, QUALCOMM

Smart Secure Platform, Klaus Vedder, Giesecke & Devrient, ETSI SCP Chairman

Network Slicing, Anne-Marie Praden, Gemalto

Don't build on Sand: Validating the Security Requirements of NFV Infrastructure to Confidently Run Slices, Nicolas Thomas, Fortinet

5G Enhancements to Non-3GPP Access Security, Andreas Kunz, Lenovo

Security and Privacy of IoT in 5G, Marcus Wong, Huawei Technologies

ITU-T activities and Action Plan on 5G Security, Yang Xiaoya, ITU-T SG17

Wrap up: 5G Overview from 3GPP SA3 Perspective and What is There to Be Done for Phase 2, Sander Kievit, TNO


Session 2: Security in 5G Inter-Network Signalling
Session Chair: Stefan Schroeder, T-Systems

Presentation on SBA: Introduction of the Topic and Current Status in SA3, Stefan Schroeder, T-Systems

5G Inter-PLMN Security: The Trade-off Between Security and the Existing IPX Business Model, Ewout Pronk, KPN on behalf of GSMA Diameter End to End Security Subgroup

Secure Interworking Between Networks in 5G Service Based Architecture, Silke Holtmanns, Nokia Bell Labs

Security Best Practises using RESTful APIs, Sven Walther, CA Technologies

Identifying and Managing the Issues around 5G Interconnect Security, Stephen Buck, Evolved Intelligence

Zero Trust Security Posture in 5G Architecture, Galina Pildush, Palo Alto Networks (Missing)


Session 1 & 2 Workshop Wrap up: 5G Phase 1 Conclusions and Outlook Towards Phase 2 - Stefan Schroeder, T-Systems and Bengt Sahlin, Ericsson


Session 5: Benefits and Challenges of 5G and IoT From a Security Perspective
Session Chair: Arthur van der Wees, Arthur's Legal

Setting the Scene, Franck Boissière, European Commission

ENISA's View on Security Implications of IoT and 5G, Apostolos Malatras, ENISA

Smart City Aspects, Bram Reinders, Institute for Future of Living

The Network Operators Perspective on IoT Security, Ian Smith, GSMA


Related Links:

Saturday 16 June 2018

Summary and Analysis of Ericsson Mobility Report 2018

Ericsson Mobility reports always make a fantastic reading. Its been a while since I wrote anything on this topic so I thought lets summarize it and also provide my personal analysis. Please feel free to disagree as this is just a blog post.

Before we start, the official site for the report is here. You can jump directly to the PDF here. Ericsson will also be holding a webinar on this topic on 19 June, you can register here.

A short summary of some of the highlights are in the table above but lets look at more in detail.

Mobile subscriptions 



  • The total number of mobile subscriptions was around 7.9 billion in Q1 2018.
  • There are now 5.5 billion mobile broadband subscriptions.
  • Global subscription penetration in Q1 2018 was 104 percent.
  • The number of LTE subscriptions increased by 210 million during the quarter to reach a total of 2.9 billion.
  • Over the same period, GSM/EDGE-only subscriptions declined by 90 million. Other technologies declined by around 32 million.
  • Subscriptions associated with smartphones now account for around 60 percent of all mobile phone subscriptions.

Many things to note above. There is still a big part of the world which is unconnected and most of the connectivity being talked about is population based coverage. While GSM/EDGE-only subscriptions are declining, many smartphone users are still camped on to GSM/EDGE for significant time.

While smartphones are growing, feature phones are not far behind. Surprisingly, Reliance Jio has become a leader of 4G feature phones.

My analysis from the developing world shows that many users are getting a GSM feature phone as a backup for when smartphone runs out of power.


Mobile subscriptions worldwide outlook


  • 1 billion 5G subscriptions for enhanced mobile broadband by the end of 2023, accounting for 12 percent of all mobile subscriptions.
  • LTE subscriptions continues to grow strongly and is forecast to reach 5.5 billion by the end of 2023
  • In 2023, there will be 8.9 billion mobile subscriptions, 8.3 billion mobile broadband subscriptions and 6.1 billion unique mobile subscribers.
  • The number of smartphone subscriptions is forecast to reach 7.2 billion in 2023.

The report describes "A 5G subscription is counted as such when associated with a device that supports NR as specified in 3GPP Release 15, connected to a 5G-enabled network." which is a good approach but does not talk about 5G availability. My old question (tweet below) on "How many 5G sites does an operator have to deploy so that they can say they have 5G?" is still waiting for an answer.


5G device outlook



  • First 5G data-only devices are expected from the second half of 2018.
  • The first 3GPP smartphones supporting 5G are expected in early 2019.
  • From 2020, when third-generation chipsets will be introduced, large numbers of 5G devices are forecast.
  • By 2023, 1 billion 5G devices for enhanced mobile broadband are expected to be connected worldwide.

Qualcomm has made a good progress (video) on this front and there are already test modems available for 5G. I wont be surprised with the launch. It would remain to be seen what will be the price point and demand for these 5G data-only devices. The Register put it quite bluntly about guinea pigs here. I am also worried about the misleading 5G claims (see here).


Voice over LTE (VoLTE) outlook



  • At the end of 2017, VoLTE subscriptions exceeded 610 million.
  • The number of VoLTE subscriptions is projected to reach 5.4 billion by the end of 2023.
  • VoLTE technology will be the foundation for enabling 5G voice calls.
  • New use cases in a 5G context are being explored, such as augmented reality (AR) and virtual reality (VR).

Back in 2011, I suggested the following (tweet below)
Looks like things haven't changed significantly. There are still many low end devices that do not support VoLTE and many operators dont support VoLTE on BYOD. VoLTE has been much harder than everyone imagined it to be.


Mobile subscriptions worldwide by region



  • Globally, mobile broadband subscriptions now make up 68 percent of all mobile subscriptions.
  • 5G subscriptions will be available in all regions in 2023.
  • In 2023, 48 percent of subscriptions in North America and 34 percent in North East Asia are expected to be for 5G.

I think that for some regions these predictions may be a bit optimistic. Many operators are struggling with finance and revenue, especially as the pricing going down due to intense competition. It would be interesting to see how these numbers hold up next year.

While China has been added to North-East Asia, it may be a useful exercise to separate it. Similarly Middle East should be separated from Africa as the speed of change is going to be significantly different.


Mobile data Traffic Growth and Outlook

  • In Q1 2018, mobile data traffic grew around 54 percent year-on-year.
  • The quarter-on-quarter growth was around 11 percent.
  • In 2023, 20 percent of mobile data traffic will be carried by 5G networks.
  • North America has the highest monthly usage of mobile data per smartphone at 7.2 gigabytes (GB), anticipated to increase to 49GB in 2023.
  • Total mobile data traffic is expected to increase by nearly eight times by the end of 2023.
  • In 2023, 95 percent of total mobile data traffic is expected to be generated by smartphones, increasing from 85 percent today.
  • North East Asia has the largest share of mobile data traffic – set to reach 25EB per month in 2023.

This is one of the toughest areas of prediction as there are a large number of factors affecting this from pricing to devices and applications.

Quiz question: Do you remember which year did data traffic overtake voice traffic? Answer here (external link to avoid spoilers)


Mobile traffic by application category



  • In 2023, video will account for around 73 percent of mobile data traffic.
  • Traffic from social networking is also expected to rise – increasing by 31 percent annually over the next 6 years.
  • The relative share of social networking traffic will decline over the same period, due to the stronger growth of video.
  • Streaming videos in different resolutions can impact data traffic consumption to a high degree. Watching HD video (720p) rather than standard resolution video (480p) typically doubles the data traffic volume, while moving to full HD (1080p) doubles it yet again.
  • Increased streaming of immersive video formats would also impact data traffic consumption.

It would have been interesting if games were a separate category. Not sure if it has been lumped with Video/Audio or in Other segments.


IoT connections outlook


  • The number of cellular IoT connections is expected to reach 3.5 billion in 2023. This is almost double our last forecast, due to ongoing large-scale deployments in China.
  • Of the 3.5 billion cellular IoT connections forecast for 2023, North East Asia is anticipated to account for 2.2 billion.
  • New massive cellular IoT technologies, such as NB-IoT and Cat-M1, are taking off and driving growth in the number of cellular IoT connections.
  • Mobile operators have commercially launched more than 60 cellular IoT networks worldwide using Cat-M1 and NB-IoT.

It is important to look at the following 2 definitions though.

Short-range IoT: Segment that largely consists of devices connected by unlicensed radio technologies, with a typical range of up to 100 meters, such as Wi-Fi, Bluetooth and Zigbee. This category also includes devices connected over fixed-line local area networks and powerline technologies

Wide-area IoT: Segment consisting of devices using cellular connections, as well as unlicensed low-power technologies, such as Sigfox and LoRa

The Wide-area IoT in the table above includes cellular IoT. If you are a regular reader of this blog, you will know that I think LoRa has a bright future and my belief is that this report ignores some of the reasons behind the popularity of LoRa and its growth story. 


Network coverage

  • In 2023, more than 20 percent of the world’s population will be covered by 5G.
  • 5G is expected to be deployed first in dense urban areas to support enhanced mobile broadband.
  • Another early use case for 5G will be fixed wireless access.
  • Today, 3GPP cellular networks cover around 95 percent of the world’s population.

A lot of work needs to be done in this area to improve coverage in rural and remote locations.

I will leave this post at this point. The report also contains details on Network Evolution, Network Performance, Smart Manufacturing, etc. You can read it from the report.