The switch from 3G to 4G was a big transformation for operators, but the next stage will be different. As Lester Thomas, chief systems architect at Vodafone, puts it: “4G’s technical term is LTE – long term evolution – so 4G was always designed to be something that would evolve.”
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
While 5G trials are ongoing and December 2017 saw the approval of the non-standalone 5G new radio specifications, 4G networks are still reaching maturity, and although some US and Asian operators are touting roll-outs as early as later this year or next, 5G networks are not expected to launch in the UK until 2020.
Research from mobile industry association GSMA, published in February this year, finds that in 2019, 10 years after launch, 4G will have the most connections (more than 3 billion) worldwide.
By 2025, there are expected to be 1.2 billion 5G connections globally, but 4G is still predicted to account for 53% of total mobile SIMs, compared to 29% today, and 5G will account for just 14%.
Estimates suggest 5G will cost tens or hundreds of billions for a network across the US or Europe. Operators have already sunk billions into 4G networks and need to squeeze the maximum return from them, which demands ongoing investment.
Howard Jones, head of network communications at EE, says: “Our 4G investment will continue right through to the middle of the next decade, both from a coverage perspective and from a capacity perspective.”
The company is focused on enhancing 4G in high-density places, rural areas with poor coverage and indoor spaces.
We can do it without 5G
Operators say many of the features associated with 5G, like large multiple input/multiple output (MIMO), beamforming and network slicing, can be achieved to some extent with 4G.
Network slicing allows a physical network to be divided into multiple virtual networks so the operator can provision the right “slice” depending on the requirements of the use case.
But 4G networks can already support it, particularly through network functions virtualisation (NFV) and software-defined networking (SDN) which allow operators to “spin up” and move network capacity around. 5G adds the capability to the radio access network (RAN), to create slices at the edge for low-power applications where sensors have limited battery power or for a very low-latency, high bandwidth application like augmented reality.
“In essence, the way we’ve built the Emergency Services Network (ESN),” says Jones. “It behaves as though it’s a network slice, so it’s LTE, but it’s a fairly early look at what it would be like to have a 5G network where you had a dedicated slice for a particular customer, because we give them top priority on the network and a dedicated core network of their own.”
Further, operators such as Vodafone in the UK have already deployed massive MIMO and beamforming across its network, using multiple antennae to send and receive data more efficiently and boost capacity where lots of people connect to the network at the same time.
The company noted: “While building a full 5G network will take time and 5G isn’t expected to roll out globally until 2020, networks are already being enhanced to keep ahead of demand and bring some of the benefits much sooner.”
4G and 5G working together
Despite advances in 4G, capacity is likely to be the initial driver for implementing 5G. According to Cisco’s latest Visual Networking Index, in 2016, global mobile data traffic amounted to 7 exabytes per month.
In 2021, mobile data traffic worldwide is expected to reach 49 exabytes per month at a compound annual growth rate of 47%. GiffGaff expects global mobile data usage to increase by 720% by 2021, projecting data usage of “a staggering” 98.34 GB by 2025, per SIM, compared to just 1.26 GB in 2016.
To manage these demands and maintain customer experience, 5G will need to support billions of simultaneously connected devices and up to a million connections per square kilometre.
“5G, to some extent will be used to relieve capacity issues in densely populated areas. If you don’t have a lot of roadmap for 4G from a capacity perspective, 5G is the answer,” says Jones.
“The challenge, then, is getting enough users with their compatible devices onto those sites so that you see the benefit of that capacity, so you can shift the user from 4G to 5G, which means more 4G capacity for everybody else,” he says.
While data and network demands are rising, telco revenues for traditional services are flat-lining. Recent analysis from IHS Markit found that global telecom revenue grew just 1.1% in 2017, “despite unabated network usage.” The report noted that every region showed either revenue decline or low single-digit growth at most.
Clearly telcos need to find some radical new source of revenue, and 5G offers hope, although not everyone is convinced that the financials stack up yet.
At a Huawei mobile broadband event in November, BT CEO Gavin Patterson said: “We need to finish the job on 4G – we need to make sure we get return on investment, we need to make sure we truly get the use of mobile networks up.
“Then we need to build the use case for 5G – clearly the innovation is there but ultimately, as carriers, we have to make a significant investment and put capex down, and the business case for that still needs to be built in many ways.”
The transition from 3G to 4G was underpinned by going from poor internet experience to one that opened up the internet for mobile – the industry hasn’t found that push yet for 5G.
“5G will be a better experience, but it is opening up the verticals, particularly around the IoT, that will open up new revenue streams,” says Patterson. “Finding use cases is the biggest challenge we have at the moment.”
Ericsson forecasts that 5G will generate $1.3 trillion in business value by 2026, and mobile operators could benefit from an additional 34% bump in revenues. The biggest opportunities are seen to be in energy, manufacturing, public safety and healthcare, plus transportation, media and automotive.
Many are “critical” services, which rely on ultra-high reliability and availability (99.9999% availability is anticipated in 5G networks) and ultra-low latency (on a 4G network it’s typically about 50-100 milliseconds but on a 5G network it will be 1 millisecond or less).
“The differrence between 4G at a 100-millisecond delay and 5G at 1 millisecond is in the order of half a car length. Half a car length can mean the difference between life and death,” Kenneth Budka, senior partner at Bell Labs Consulting, said at the Brooklyn 5G Summit in April 2017.
As 5G is a service-oriented architecture (SOA), it can slice the network more easily to support unique use case requirements.
“You can support things that you can’t support today where 4G is like a single architecture,” says Thomas.
Jones at EE notes the potential in new areas too. “We have had a lot of interest from broadcasters looking at 4G to provide them with outside broadcast capabilities,” he says. “However, the nature of 4G means we can’t necessarily prioritise a single customer and provide a guaranteed SLA (service-level agreement).
“That changes with 5G. We can make promises on particular SLAs for a given time period if need be, so the impact of 5G on the broadcast industry could be significant.
“It takes a lot of investment to get there because you’ve got to have coverage,” he says. “But it could bring to life what 5G can do in terms of throughput capabilities and reliability.”
Many operators are also looking at the platform business model to boost revenue – that is, offering digital ecosystems connecting producers of goods and/or services with consumers and/or the platform-based IT architecture which supports these models.
The platform business model could be the potential “killer app” for 5G, says Thomas.
Research in 2017 from trade association TM Forum found that more than a third of communications service providers offer digital ecosystems or platform marketplaces already, and another 40% intend to within the next two years.
Vodafone, for example, offers an IoT platform for simple services, such as smart cities with end devices like smart lighting and waste management, with little networking requirement.
In the future, 5G will be needed to cope with the volume, variety and velocity of use cases, with applications from augmented and virtual reality to connected cars and digital health. Also, a central principle of platforms is that its applications keep evolving.
“With a 4G network it was almost like a compromise between bandwidth, latency and cost,” says Thomas. “But 5G, because it has the network slicing and the service-oriented approach, you can tailor the network for each different use cases on the same shared network platform.”
TM Forum’s report concludes that while 5G may not necessarily be critical to the success of platform business models, platforms combined with IoT will be essential to 5G and unlocking new IoT revenue streams that 4G can’t, through combining higher bandwidth with network slicing.