Neil McRae: What’s next for the telecom industry
In a talk at the FutureNet World conference, held in London on May 3-4, Neil McRae explains why he is upbeat about the telecoms industry’s prospects
Neil McRae is tasked with giving the final talk of the two-day FutureNet World conference.
“Yeah, I’m on the graveyard shift,” he quips.
McRae, the former chief network architect at BT, is now chief network strategist at Juniper Network.
The talk’s title is “What’s Next”, McRae’s take on the telecom industry and how it can grow.
McRae starts with how, as a 15-year-old, he had attended an Apple Macintosh computer event at a Novotel Hotel in Hammersmith, London, possibly even this one hosting this conference.
An Apple representative had asked for his feedback as a Macintosh programmer. McRae then listed all the shortfalls programming the PC. Later, he learnt that he had been talking to Steve Jobs.
Perhaps this explains his continual focus on customers and meeting their needs.
What customers care about, says McRae, is ‘new stuff’ that makes a difference in their lives. “Quite often in telecoms, we accidentally change the world without even thinking about it,” he says
McRae cites as an example using FaceTime to watch a newborn grandchild halfway across the world.
“We do it all the time; it is a phenomenal thing about our industry,” says McRae.
The Unvarnished Truth
McRae moves to showing several market and telco survey charts from IDC and Analysys Mason, what he calls ‘The unvarnished truth’.
The first slide shows how the European enterprise communication service market is set to grow at a compound annual growth rate (CAGR) of 3% between 2020 and 2025.
“Three per cent growth, who thinks that is a great business for telcos?” says McRae. “And enterprise is what we are all depending on for big growth and change because [the] consumer [market] is pretty much flat,” says McRae.
Another chart shows similar minimal growth: a forecast that Western Europe’s mobile retail service market will grow from $102 billion in 2016 to $109 billion by 2026. Yet mobile is where the telcos spend a ton of money, he says.
“So, who thinks we should continue doing what we are doing?” McRae asks the audience.
Another forecast showing global fixed and mobile service revenues is marginally better since it includes developing nations that still lack telecommunications services.
In the UK, 95 per cent of the population is on the internet, in Europe it is 84%, says McRae: “The UK is a tough place to be to grow business.”
Telco transformation
Another slide (see above), the results of a telco survey, shows a list of topics and their impact on telco transformation. McRae asks the audience to respond to those they think will ‘save’ the industry.
He goes through the list: cloud and cloudification, artificial intelligence (AI) and machine learning, 5G, and data analytics. The audience remains muted.
The next item is application programming interfaces (APIs). Again the audience is quiet. “You have been talking about APIs for two days!” says McRae.
“The right APIs,” shouts an audience member. “Ah, yes, the right APIs,” says McRae.
McRae continues down the list, virtualisation and software-defined infrastructure, OpenRAN elements – “not sure what the elements mean” – orchestration platforms, advanced process automation (OSS/BSS), micro-services, and blockchain.
McRae says he has spent the equivalent of a small nation’s budget over his career on OSS and BSS. “Nothing is automated, and I can’t get the data I need,” he says.
McRae gives his view. He believes the cloud will help telcos, but what most excites him is AI and machine learning, and data analytics.
“Learning the insights the data tells us and using them, putting a pound sign on them,” says McRae. “We have done some of that, but there is much more to do.”
He puts up a second survey showing the priorities of European operators: customer experience and increasing operational agility.
“Finally, after years, telcos realise that customers are important,” he says.
Opportunities
The survey also highlights the telcos’ belief that they can deliver solutions for industries and enterprise customers.
“This is a massive opportunity for telcos that allows us to grow revenues, create cool technology and hire amazing engineers,” says McRae.
The transformation needed in telecoms is about customers and taking risks with customers, he says.
One opportunity is digitalisation. McRae points outs that digitalisation is a process that never stops.
The three leading Chinese operators are keenly pursuing what they call industrial digitalisation or industrial internet. For China Telecom, industrial digitalisation now accounts for a quarter of its service revenues.
“Today, it is about cloud, cloud technologies, and smartphones, but tomorrow it could be about wearables or technology that is tracking what you are doing and making your life easier,” says McRae.
Digitalisation is an expertise that the telecom industry is not putting enough effort into, he says: “And as telcos, we have a massive right to play here.”
Another opportunity is AI and data, learning from the insights present in data to grow revenue.
“We have more data than most organisations, we haven’t used it very well, and we can build upon it,” says McRae, adding that AI needs the network to be valuable and improve our lives.
With data and AI, trust is vital. “If we are not trusted as an industry, we are dead,” says McRae. But because telcos are trusted entities, they can help other organisations improve trustworthiness.
Another opportunity is using the network for humans to interact in advanced ways. Since telecoms is a resource-heavy industry, such network-aided interaction would be immediately beneficial.
For this, what is needed is a cloud-native platform that integrates well with the network, and cloud platforms are generally poorly integrated with the network, he says.
He ends his talk by returning to customers and what they want: customers expect networks and services to be always present.
This explains the telcos’ continual marginal growth, he says: “The reason we have this is because there is a big chunk of customers’ lives where they can’t rely upon the network.”
Different thinking is needed if the network is to grow beyond the smartphone. Population coverage is not enough; what is needed is total coverage.
“Wherever I am, I want to use my device, to be connected, for the things that I don’t even know is doing stuff to be able to do them without worrying about connectivity,” he says.
And that is why 6G must be about 100 per cent connectivity,” says McRae: “Either we can do it, or someone else is going to.”
With that, FutureNet comes to ends, and McRae quickly departs to embark on the next chapter in his career. `
Neil McRae will be one of the speakers at the DSP Leaders World Forum, May 23-24, 2023.
2012: A year of unique change
The third and final part on what CEOs, executives and industry analysts expect during the new year, and their reflections on 2011.
Karen Liu, principal analyst, components telecoms, Ovum @girlgeekanalyst
"We’ve entered the next decade for real: the mobile world is unified around LTE and moving to LTE Advanced, complete with small cells and heterogenous networks including Wi-Fi."
Last year was a long one. Looking back, it is hard to believe that only one year has elapsed between January 2011 and now.
In fact, looking back it is hard to remember how things looked a year ago: natural disasters were considered rare occurrences. WiMAX’s role was still being discussed, some viewed TDD LTE as a Chinese peculiarity. For that matter, cloud-RAN was another weird Chinese idea. But no matter, China could do anything given its immunity to economics and need for a return-on-investment.
Femtocells were consumer electronics for the occasional indoor coverage fix, and Wi-Fi was not for carriers.
Only optical could do 100Mbps to the subscriber, who, by the way, was moving on to 10 Gig PON in short order. Flexible spectrum ROADMS meant only Finisar could play, and high port-count wavelength-selective switches had come and gone. 100 Gigabit DWDM took several slots, hadn’t shipped for real, and even the client-side interface was a problem.
As for modules, 40 Gigabit Ethernet (GbE) client was CFP-sized, and high-density 100GbE looked so far away that the non-standard 10x10 MSA was welcomed.
NeoPhotonics was a private company, doing that wacky planar integration thing that works OK for passives but not actives.
Now it feels like we’ve entered the next decade for real: the mobile world is unified around LTE and moving to LTE Advanced, complete with small cells and heterogenous networks including Wi-Fi.
Optical is one of several ways to do backhaul or PC peripherals. 40GbE, even single-mode, comes in a QSFP package, tunable comes in an SFP — both of which, by the way, use optical integration.
Most optical transport vendors, even metro specialists, have 100 Gigabit coherent in trial stage at least. Thousands of 100 Gig ports and tens of thousands of 40 Gig have shipped.
Flexible spectrum is being standardised and CoAdna went public. The tunable laser start-up phase concluded with Santur finding a home in NeoPhotonics, now a public company.
But we also have a new feeling of vulnerability.
Optical components revenues and margins slid back down. Bad luck can strike twice, with Opnext taking the hit from both the spring earthquake and the fall floods. China turns out not to be immune after all, and time hasn’t automatically healed Europe.
What will happen this year? At this rate, I think we’ll see a lot of news at OFC in a couple of months' time. By then I’ll probably think: "Was it as recently as January when the world looked so different?"
Brian Protiva, CEO of ADVA Optical Networking @ADVAOpticalNews
Last year was an incredible year for networks. In many respects it was a watershed moment. Optical transport took a huge step forward with the genuine availability of 100 Gigabit technologies.
What's even more incredible is that 100 Gigabit emerged in more than the core: we saw 100 Gig metro solutions enter the marketplace. This means that for the first time enterprises and service providers have the opportunity to deploy 100 Gig solutions that fit their needs. Thanks to the development of direct-detection 100 Gig technology, cost is becoming less and less of an issue. This is a game changer.
In 2012, 100 Gig deployments will continue to be a key topic, with more available choices and maturing systems. However, I firmly believe the central focus of 2012 will be automation and multi-layer network intelligence.
"We need to see networks that can effectively govern and optimise themselves."
Talking to our customers and the industry, it is clear that more needs to be done to develop true network automation. There are very few companies that have successfully addressed this issue.
We need to see networks that can effectively govern and optimise themselves. That can automatically deliver bandwidth on demand, monitor and resolve problems before they become service disrupting, and drive dramatically increased efficiency.
The future of our networks is all about simplicity. The continued fierce bandwidth growth can no longer be supported by today's complex operational inefficiencies. Streamlined operations are essential if operators are to drive for further profitable growth.
I'm excited about helping to make this happen.
Arie Melamed, head of marketing, ECI Telecom @ecitelecom
The existing momentum of major traffic growth with no proportional revenue increase has continued - even intensified - in 2011. This means that operators have to invest in their networks without being able to generate the proportional revenue increase from this investment. We expect to see new business models crop up as operators cope with over-the-top (OTT) services.
To differentiate themselves from competition, operators must make the network a core part of the end-customer experience. To do so, we expect operators to introduce application-awareness in the network – optimising service delivery to avoid network expansions and introduce new revenues.
We also expect operators to offer quality-of-service assurance to end users and content application providers, turning a lose-lose situation around.
Larry Schwerin, CEO of Capella Intelligent Subsystems @CapellaPhotonic
Over 2011, we witnessed the demand for broadband access increase at an accelerated rate. Much of this has been fueled by the continuation of mass deployments of broadband access - PON/FTTH, wireless LTE, HFC, to name a few - as well as the ever-increasing implementation of cloud computing, requiring instantaneous broadband access. Video and rich media are a small but growing piece of this equation.
The ultimate of this is yet to be felt, as people start to draw more narrowcast versus broadcast content. The final element will be when upstream content via appliances similar to Sling Media, as well as the various forms of video conferencing, become more widespread. This will lead to more symmetrical bandwidth from an upstream perspective.
'Change is definitely in order for the optical ecosystem. The question is how and when?'
Along with this, the issue of falling revenue-per-bit is forcing network operators to develop more cost-effective ways for managing this traffic.
All of aforementioned is driving demand for higher capacity and more flexible support at the fundamental optical layer.
I believe this will work to translate into more bits-per-wavelength, more wavelengths-per-fibre, and finally more flexibility for network operators. They will be able to more easily manage the traffic at the optical layer. This points to good news for transponder, tunable and ROADM/ WSS suppliers.
2011 also pointed out certain issues within the optical communications sector. Most notably, entering 2011, the financial marketplace was bullish on the optical sector following rapid quarter-on-quarter growth of certain larger optical players. Then, the “Ides of March” came and optical stocks lost as much as 40% of their value when it was deemed there was a pull back in demand by a very few, but nonetheless important players in the sector.
Later in the year came the flooding in Thailand, which hampered the production capabilities of many of the optical components players.
Overall margins in the sector remain at unacceptable levels furthering the speculation that things need to change in order for a more robust environment to exist.
What will 2012 bring?
I believe demand for bandwidth will continue to grow. Data centres will gain more focus as cloud computing continues to gain traction. This will lead to more demand for fundamental technologies in the area of optical transmission and management.
The next phase of wavelength management solutions will start to emerge - both at the high port count (1x20) as well as low-port count (1x2, 1x4) for edge applications. More emphasis will be placed on monitoring and control as more complex optical networks are built.
Change is definitely in order for the optical ecosystem. The question is how and when? Will it simply be consolidation? How will vertical integration take shape? How will new technologies influence potential outcomes?
2012 should be a year of unique change.
Terry Unter, president and general manager, optical networks solutions, Oclaro
Discussion and progress on defining next-generation ROADM network architectures was a very important development in 2011. In particular, consensus on feature requirements and technology choices to enable a more cost-efficient optical network layer was generally agreed amongst the major network equipment manufacturers. Colourless, directionless and, to a significant degree, contentionless are clear goals, while we continue to drive down the cost of the network.
"We expect to see a host of system manufacturers making decisions on 100 Gig supply partners. This should be an exciting year."
Coherent detection transponder technology is a critical piece of the puzzle ensuring scalability of network capacity while leveraging a common technology platform. We succeeded in volume production shipments of a 40 Gig coherent transponder and we announced our 100 Gig transponder.
2012 will be an important year for 100 Gig. The availability of 100 Gig transponder modules for deployment will enable a much wider list of system manufacturers to offer their customers more spectrally-efficient network solutions. The interest is universal from metro applications to the long haul and ultra-long haul market segments.
While there is much discussion about 400 Gig and higher rates, standards are in very early stages. The industry as a whole expects 100 Gig to be a key line rate for several years.
As we enter 2012, we expect to see a host of system manufacturers making decisions on 100 Gig supply partners. This should be an exciting year.
For Part 1, click here
For Part 2, click here