TMN Quarterly | 2015 | Issue 09 by The Mobile Network

The Mobile Network // www.the-mobile-network.com

ALSO FEATURING Making sense of the world’s mobile networks

42 C EM: CARING ABOUT THE CUSTOMER 50 COUNTRY PROFILE: GERMANY // MORE... 2015 // Issue 09

d l r o W e l i Mob gress ///// // Con Edition

Double wo ////////////// tT r a P / / / / / ///

N I N I F E 22 D

for? d o o g t i hat is W G 5

ING M A O R E 30 VOLT

/////

g n i t u p Com e g d E e ING l i b E GETT R o A O H M W 16 TWORK ATORS THE NE E OPER H RATING SIDE: T RCHEST ALSO IN O // V F HN ON WIT

ISSUE

REGULARS GLOBAL CORRESPONDENTS // ANATOMY OF A MOBILE OPERATOR: AT&T

#09

Trillium速 Small Cell Software

T-Series Platforms for SDN/NFV

CONTENTS

MOBILE WORLD CONGRESS EDITION PART 2 //////////////////

FEATURE

EDITOR

Hi!

In this issue we run the second parts of our features on 5G, VoLTE and Mobile Edge Computing. The 5G feature moves into a detailed look at potential use cases for 5G. In other words, just what are the applications and activities that will require the capabilities that 5G can offer? From massive machine type communications to massive broadband - many of the 5G justifications are gathered here.

Anatomy of a Mobile Operator: AT&T

More intelligent life at the edge of the network. Our second MEC piece raises the issues of 5G, security and applications.

Making the use case for 5G use cases. We are used to seeing 5G talked about in terms of use cases. We look into the key ones.

But how? Operators may have little choice whether to enable VoLTE roaming, but they do have choices on how to do it.

You may have heard about orchestration. What does it mean in terms of managing services over networks containing virtual functions?

Mobile Edge Computing: Part II

Exploring 5G: Part II

VoLTE Part II: Roaming

LTE, LTE-A, VoLTE, and some non-LTE news.

NFV Orchestration

Our VoLTE feature looks at how you might go about enabling roaming for VoLTE. We don’t ask why you might do this, though, because it’s going to happen whether you think it’s worth it or not.

Country profile: Germany Leading in tech, leading in consolidation.

Picture Story: Up in the skies.

Customer Experience Management Do operators really care about the customer? (Actually, yes)

NFV Update Who is doing what at the major operators.

Finally, we ask the slightly cheeky question - do operators really care about the customer? The answer, really, is yes they do. But how can they put together a meaningful CEM strategy? By the way, for much more on this make sure you get your hands on our CEM Market Update. So there we have it, the second of our two-hander Mobile World Congress 2015 Issues. Welcome to Issue 9 of The Mobile Network Quarterly.

* 36

In this MWC issue, we nod to the prevailing hype and take a two part look at NFV, the term that has a specific use but also stands in as a catch-all for the cloudification of mobile network technology and services. We catch up with what operators have been up to, and with how services and infrastructure can be managed.

Commercial Director: Shahid Ramzan // shahid@the-mobile-network.com Editorial Director: Keith Dyer // keith@the-mobile-network.com Creative Direction and Design: Shona Gow // hello@shonagow.co.uk // www.shonagow.co.uk

Keith Dyer keith@the-mobile-network.com

4 TMNQUARTERLY

TMNQUARTERLY 5

FROM OUR CORRESPONDENT

“From Our Correspondent” collects writing from around the world, covering

CORRESPONDENT : Anon

FROM OUR CORRESPONDENT If you would like to contribute to this section, mail TMN’s editor keith@the-mobile-network.com

mobile network stories in local markets.

USA...

and the money rolls in

And so the AWS-3 auction - or auction 97 as it is thrillingly known - closed with bidding ending just shy of $45 billion. As we wrote in the previous issue this auction of paired spectrum in the 1700MHz and 2.GHz bands exceeded all expectations at least in terms of raising money. These are eye-raising sums for mobile operators to be paying out. Yet these numbers were put in context by just a single company that announced a quarterly net profit of $18 billion on

6 TMNQUARTERLY

quarterly revenues of $75 billion, bringing its cash pile to nearly $180 billion. Yes, the great god of the iPhone again amazed analysts, who boggled at the sheer numbers involved. Of course, Apple’s ability to extract massively high prices for its phones is absolutely not its fault, people pay what they are willing to pay after all. And if they deem the phone, the apps, the environment Apple has created worthy of the the price then they will pay it. Certainly the operators are willing to meet Apple’s prices up front, knowing that having the device is still a big customer draw. That’s capitalism, baby. But how soon will it be before serious questions are asked about this company’s profitability. Apple’s profitability is not just a function of its ability to ask more for its phones than anyone else. It also beats down on its supply chain. Yes, no more so than anyone else in the consumer electronics sector. But is it time to ask, “Heck, our government might even ask how many businesses Apple has created in Asia could we do anything to create our own supply industries, or does competing with companies paying a few dollars a day wages make Western companies necessarily uncompetitive?”

INTO SPACE The mobiles in space dream refuses to die. As regular readers of this authoritative journal will know, there’s a long history of companies trying, failing, trying again, succeeding in a limited way, of launching global mobile coverage from the sky. The latest cunning wheeze comes with the backing of Qualcomm and the famously shy Richard Branson. Not like Branson to attach his name to an attention-grabbing project just for the brand enhancement, so there must be something in it. Qualcomm, however, is a name that we in this industry have learnt to take very

seriously indeed. Sure, the company has backed a few schemes over the years with “Loser” stamped firmly on the file as it was stowed back in the cupboard. But in general, the San Diego beast is usually pretty keen to expand its addressable market, and takes sensible bets. So is Space one of them? Well, this time around there is a slight difference to previous launches. In the past, the vision was of low earth orbit satellites connecting directly to phones. This time around, OneWeb’s vision is to put a great many satellites into the sky (more than currently exist up in the heavens, in fact) but instead of connecting directly to satellite phones, the idea is to hook up to access points that will sit on the ground, and in turn provide bubbles of coverage. Essentially it’s using satellite as one massive backhaul network to distributed cellular and WiFi base stations. A proviso, though. It’s not actually clear what Qualcomm’s involvement is, nor what the economics will be of using Virgin’s OneLauncher (yet to get off the ground) to fire 6-700 birds into the near-heavens. File under:

“don’t hold your breath”.

THE NETHERLANDS While the rest of the world comes to grips with net neutrality as a concept, here in The Netherlands we are seeing the reality. We’re used to being pioneers, of course. After all we got rich on the exploits of the merchant-tyrants of the VOC - the Dutch East India Company - as it opened up trading routes for spices, silks and more across South Asia. In those days the VOC was actually granted a monopoly to trade in the East. These days our competition authorities have rather changed their stance, and would take a rather dim view of such a position. So now it is your turn to watch a slightly less thrilling battle play out as our competition authority makes sure that the internet is delivered fairly, without certain companies being given priority, or consumers being stopped from doing things they would like to do. We are aware that this makes us closely folllowed, especially as the USA’s FCC and its carriers wrestle with net neutrality. It’s far from a purely US-centric discussion, of course. Closer to home our EU politicians too are grappling with what role net neutrality can play in the digital single market, and if the imposition of such a thing might hamper operators from investing. Of course the operators say that if they are stopped from applying certain controls to traffic, such as blocking it, or giving this one or that one preferential access, then their ability to monetise networks will be greatly hampered. And if they cannot monetise their networks then why should they invest in their networks. And if they don’t invest in their networks, then where does that leave the poor old governments that are, after all, trusted with providing the best infrastructure for their people and businesses? Well, in January we saw our Competition authority actually fine two mobile

operators for such blocking and preferential actions. KPN was fined for stopping users from making VoIP calls over open WiFi networks. Vodafone was given a rap for zero-rating data within an app providing content from HBO. “It is forbidden for Internet providers to determine what their customers can and cannot do on the Internet,” said the regulator. The first is perhaps more clear cut in terms of being unsupportable, despite KPN’s protestations. The second use-case, that of zero-rating HBO content, relates far more to a great many business models being proposed right now. Operators are rather keen on the two-sidedness of this model - getting customer acquisition or loyalty (confidence that you are not nixing your data limit by watching video) whilst keeping your content provider happy. The sums in questions were not small, nor huge - fines of EUR250,000 and EUR200,000 in each case being annoying rather than punitive, but not something the operators would want to see racking up. So, if you are interested in the practical application of network neutrality and what it means for operators’ business models, it might be worth keeping an eye on The Netherlands. “Overly prescriptive or intrusive rules would risk degrading today’s Internet experience and limiting its development potential,” said industry body ETNO, as it surveyed the possibility of the Dutch experiment being broadened out across Europe. “Policymakers should favour a forward-looking approach, recognising that the evolution of the Internet is an ongoing process.”

In other words - lay off, who are you to tell us what to do?

TMNQUARTERLY 7

FROM OUR CORRESPONDENT

MEXICO We reported from Mexico six months ago, as Carlos Slim’s playground was broken up. Who would come in for Lusacell, we asked? Well that answer turned out to be AT&T, as you surely know, with the US giant completing on a $2.5 billion purchase of our operator in January 2014. That deal was then quickly followed up by AT&T doing a two for one move on the market and taking our Nextel Mexico, subsidiary of the inbankruptcy protected NII Holdings. Sure enough the company said it had an agreement to take Nextel Mexico off NII’s hands for $1.8 billion giving it 3 million customers and a wide area mobile network, a fair chunk of spectrum and a distribution network.

PAKISTAN We are to join the ranks of LTE nations with more than one LTE operator, as our fifth and until very recently most-struggling operator Warid Telecom came roaring back into the market with an announcement of LTE services in its 1800MHz spectrum. The company joins ChinaMobile subsidiary Zong in doing so, with Zong already offering 1800MHz LTE. Warid, which is using Ericsson as its kit supplier, has said it will invest $500 million in its 8 TMNQUARTERLY

ISRAEL So Nextel Mexico will become part of AT&T’s North American service area encompassing the U.S. and Mexico, mashed up alongside Lusacell. “Combining Nextel Mexico with Lusacell will allow AT&T to more quickly improve and expand its mobile Internet service to the benefit of millions of Mexicans, particularly those who live outside major metropolitan areas, than it could otherwise do without the transaction,” said AT&T, in a statement to the market. Some now think AT&T could even go after America Movil assets. Hold on, I thought we were supposed to be breaking this market up - not finding a new consolidator?

GERMANY LTE network, and instantly offered customers a free week’s 4G goodness. Warid says it is good to go for now, and has enough spectrum at 1800MHz, but the Pakistain government is still looking to auction off two slots, one actually in the 1800MHz band and another at 850MHz, for LTE use. These are bands that went unsold at last year’s auction The 850MHz block had a reserve price of $291 million and was only available to new operators. It didn’t sell, nor did 10MHz of 1.8GHz spectrum, with a $210 million base price per 10MHz. Now, with the prospect of this new spectrum coming back onto the market, it looks like we may be seeing more LTE capacity in the country.

As bodies such as the WRC and others discuss harmonising 700MHz spectrum (in the UK they are still discussing what operators get and what broadcasters get) we’re up and running in Germany as our regulator, Bundesnetzagentur, pressed go on a frequency auction for spectrum in the 700 MHz band. The big B said that moving broadcasters to a DVB-T2 platform will free up 2x30 MHz in the 700 MHz band for the mobile operators to get their hands on. Predicted start date for that is 2017. Will these be the same bands as other operators will be offered in other markets? Or will other countries follow in our footsteps in allocating these bands? More like the latter, as the German government said it would be the first in Europe to be making 700MHz spectrum available.

If you want to see what a mobile market with a relatively small number of subscribers and lots of operators looks like, come to Israel. Here, six operators (yes, six) have just been awarded LTE spectrum in the 1800MHz band. Some of the awards seemed hardly worth bothering with, with Cellcom awarded a whole 3 MHz of bandwidth and Partner Communications, Golan Telecom and HOT Telecom all getting a measly 5 MHz. Pelephone at least got 15 MHz. The additional spectrum the companies have bought brings each of them up to a total 15MHz dedicated to LTE. Cellcom, Partner and Pelephone, our three largest operators, have been offering 4G services in the major cities for some time - and in fact do so for free as a result of their currently limited spectrum holdings. Reaching 15MHz bandwidth may enable each of them to develop some more compelling commercial offers.

A Scottish outfit that has been developing a system to transmit data over the visible light spectrum has raised more money. Start-up pureLiFi, which was formed out of an Edinburgh University research project, has raised $2.25 million in its latest round of investment. The company uses the term Li-Fi for its Visible Light Communication (VLC) technology. In the fourth quarter of 2014, the pureLiFi team launched and shipped what it describes as the world’s first Li-Fi network product - Li-Flame - to industry customers worldwide. “Li-Fi is increasingly viewed as a transformative technology that can change the way we use the mobile Internet as part of future 5G cellular networks and at

5G action 5G initiatives are popping up all over the world, with universities, governments and major vendors all bonding together to create research and feasibility projects into certain 5G technologies. Not to be outdone, UK regulator Ofcom raised its hand and said it would be looking into what it should be doing with spectrum above 6GHz - and wanted to hear what others thought.

the same time be an enabler of the emerging Internet of Things,” pureLiFi’s Chief Science Officer Harald Haas said in a press release. The system developed by pureLiFi, formerly known as PureVLC, turns off-the-shelf light fixtures into Li-Fi access points that can simultaneously communicate to multiple users in a bi-directional fashion, the company explained in the release. It also consists of the world’s first battery powered Li-Fi mobile unit that is attached to a laptop screen and allows user roaming within a room - or an entire building. Readers of TMN Q may remember the company being featured previously, in our Picture Story section. (Issue 4, December 2013)

The regulator somewhat cautiously described potential use cases for “socalled 5G communications” - barely concealing its disdain for the muddy signifier this term has become. So what real-world use case did Ofcom itself proposed as possible candidate for this new “so-called” 5G world. Why, only virtual reality holographic projection. Good timing, perhaps, as Microsoft quickly followed up with its holograph technology announcement at its Windows 10 launch.

TMNQUARTERLY 9

FEATURE: ANATOMY OF AN OPERATOR

NETWORKS EXECUTIVE MANAGEMENT:

IN NUMBERS

$25Billion

KRISH PRABHU CTO

investment in network and spectrum (2013)

JOHN DONOVAN SENIOR EXECUTIVE VICE PRESIDENT, ARCHITECTURE, TECHNOLOGY, OPERATIONS TOM KEATHLEY SVP NETWORK AND PRODUCT PLANNING

AT&T IS A BIT OF A BEAST. CHECK OUT THE NUMBERS OPPOSITE.

OPERATOR PROFILE

AT&T

In 2012, just after the carrier had been forced to exit from its doomed bid to become even more of a beast by buying T-Mobile, AT&T announced Project Velocity IP (VIP), a multi-year plan to invest for growth right across its network, and in 2013 its network investment was still hovering around the $20 billion mark - if you include spectrum acquisitions. Just in early 2015 the carrier added what is likely to be another several billion dollars in spectrum costs as it bid in the AWS-3 auction. In 2015, the FCC expects to hold an auction for spectrum in the 600MHz frequency. As a result of such a concerted investment effort its LTE network now covers 300 million by population, with 10x10 MHz deployment in most markets at 700MHz and with future plans to deploy LTE on AWS spectrum. The carrier says 67% of its 55 million smartphone postpaid customers now have an LTE-capable phone. However, it doesnâ&#x20AC;&#x2122;t break out how many of those have actually activated an LTE subscription.

Here is a breakdown of the technical elements AT&T is putting together as it expands, densifies and cloudifies its networks.

LTE-A

The operator has activated Carrier Aggregation in the network in a limited manner - hooking together 700MHz and 2100MHz spectrum in a few areas, including Chicago where it launched in May 2014. Through 2015 the market expects AT&T to deploy LTE on 2.3 GHz spectrum and on refarmed 1900 MHz (ex-PCS) spectrum. That might give it additional bands to play with in a carrier aggregation mode. Carrier Aggregation demands device support, of course, with users needing the right category of device to support 2 carrier or 3 carrier aggregation.

SCOTT MAIR SVP NETWORK PLANNING & ENGINEERING AT

34,000 AT&T Wi-Fi Hot Spots

Billion 2013 revenues

JERRIE JENSON KERTZ SVP OPERATIONS PLANNING AND OPTIMISATION ANDRE FUETSCH SVP ARCHITECTURE & DESIGN, HEAD OF DOMAIN 2.0 ARCHITECTURE JOHN MEDAMANA VP, NETWORK PLATFORMS DON HOUGHTON VP, RAN PERFORMANCE MANAGEMENT AND OPERATIONS

40,000 small cells in Project VIP

MO KATIBEH VP, GLOBAL NETWORK PLANNING

HANK KAFKA VP, NETWORK ARCHITECTURE DAVID ORLOFF DIRECTOR, SMALL CELLS PLATFORMS ANGELA LIM DIRECTOR, RAN ENGINEERING

118.7m AT&T Wireless customers

56%

DAVID SASKA VP, RADIO ACCESS NETWORK ENGINEERING

300 million coverage of LTE networks

Revenue from Wireless

ROBERT KLEIN DIRECTOR, RADIO ACCESS NETWORK MARGARET CHIOSI DISTINGUISHED NETWORK ARCHITECT, (ALSO PRESIDENT, OPNFV)

150% Increase in mobile data traffic November 2012-July 2014.

10 TMNQUARTERLY

$128.8

Proportion of traffic on the network that is video. TMNQUARTERLY 11

FEATURE: ANATOMY OF AN OPERATOR

SMALL CELLS

AT&T’s strategic goal is to place small cells and DAS in a cost effective way to provide wireless coverage in difficult to serve areas. It is unusual amongst carriers in that it has long had a publicly stated commitment to a defined number of small cells, stating back in 2012 that as part of its Project VIP it expected to install 40,000 small cells across the network. It has deployed this technology in 48 states so far. Project VIP includes the deployment of small cell technology, macro cells and additional distributed antenna systems (DAS), which will increase the density of the network. This network densification will further improve network quality and increase spectrum efficiency. The carrier has said publicly it sees multi-standard small cells solutions being introduced in the 1st half of 2015.

FEATURE: ANATOMY OF AN OPERATOR

HIGH LEVEL CLOUD NETWORKING ARCHITECTURE

GET WITH THE PROGRAMME

Commercial Cloud Computing Environments

So far, AT&T has named 10 vendors in the program. These are Affirmed Networks, Alcatel-Lucent, Amdocs, Brocade Networks, Ciena, Cisco, Ericsson, Tail-F Systems (now Cisco), Juniper Networks, and Metaswitch Networks.

Network Function Virtualisation Infrastructure a cloud distributed where needed to optimise characteristics such as latency, costs, etc, and with control, orchestration, and management capabilities for real-time, automated operations

“The goal is to create a user-defined network cloud that uses standardised hardware and specialised software to make it easier to offer customers new features and functionality.”

AT&T is one of the few global operators to have launched a public programme dedicated to cloudification of its networks and SDN. The programme, launched in February 2014 and called SDN DOMAIN 2.0, includes a named list of suppliers with which AT&T wants to work on development, as well as target outcomes for the operator. It wants, by 2020, to create a user defined all-IP network, and to get there it is launching several beachhead programmes, to make progress in various areas. An introductory whitepaper from AT&T said: “In a nutshell, Domain 2.0 seeks to transform AT&T’s networking businesses from their current state to a future state where they are provided in a manner very similar to cloud computing services,

12 TMNQUARTERLY

and to transform our infrastructure from the current state to a future state where common infrastructure is purchased and provisioned in a manner similar to the PODs used to support cloud data center services.” This infrastructure is expected to be comprised of several types of substrate. The most typical type of substrate being servers that support NFV, followed by packet forwarding capabilities based on merchant silicon. However it’s envisioned that other specialised network technologies are also brought to bear when general purpose processors or merchant silicon are not appropriate. As an example, an edge router might be purchased as a monolithic device today – where the hardware, feature functions,

and a specific applicable scale of use are pre-integrated into a single device. Often a variety of device sizes need to be purchased in order to support variances in workload from one location to another. In Domain 2.0, such a router is composed of NFV software modules, merchant silicon, and associated controllers. The software is written so that increasing workload consumes incremental resources from the common pool, and moreover so that it’s elastic: so the resources are only consumed when needed. Different locations are provisioned with appropriate amounts of network substrate, and all the routers, switches, edge caches, and middle-boxes are instantiated from the common resource pool. Such sharing of infrastructure across a broad set of uses makes planning and growing that infrastructure easier to manage.

TENANT APPLICATIONS & VIRTUAL MACHINE

APis and Dynamic Policy Control

Network Function Virtualisation Infrastructure Cloud VIRTUAL APPLIANCE

VIRTUAL APPLIANCE

FUNCTIONS DESIGNED FOR CLOUD

Wireless Access

Broadband/ Fiber Access

SDN DOMAIN 2.0 PROGRAMME

TENANT APPLICATIONS & VIRTUAL MACHINE

WIFI

The carrier has been instrumental in the Carrier Grade WiFi vision. Something that started as a necessary infill solution to cover up for the performance of the 3G network - most notably exposed years ago after the launch of the iPhone onto the network - has become more of a strategic property. The carrier has well over 30,000 access points under direct management and has lead efforts to enable greater ease of WiFi roaming for its customers, with single sign-on. AT&T was at the forefront of the WBA’s development and demonstrations of Next-Generation Wi-Fi Hotspot technology at Mobile World Congress in 2014. Here, customers of the participating mobile operators with the latest compatible phones were able to authenticate onto

Elastic Network Capabilities for customers, partners, 3rd party provider tenants of commercial clouds

Network Function Software that will be evolving from the current form embedded in network appliances to software (re)designed for cloud computing

Packet & Optical Transport that will be evolving from current integrated TCP/IP control/data phone routers towards SDV where controllers external to a packet switch provide the forwarding rules

Packet and Optical

the Hotspot 2.0 Wi-Fi network as if they had roamed onto the mobile network by the enablement of SIM authorisation onto the unlicensed spectrum network. “As the Wi-Fi and cellular worlds merge, developing a common authentication mechanism with standard roaming agreements is a natural next step,” said JR Wilson, Vice President, Partnerships and Alliances, AT&T Mobility and Chairman of the Wireless Broadband Alliance (WBA), at the time. More recently, AT&T has said it plans to launch VoWiFi in 2015, as long as it can ensure the user experience. This places it behind others, such as T-Mobile, which have had a stronger go-to-market policy on VoWiFI. “We’re very focused on making sure it’s a great experience for customers, but we see it as a complement, not a

Credit: modified from AT&T original

replacement,” head of AT&T mobility and enterprise Ralph de la Vega said. “We feel good about a great nationwide network with unlimited talk and text,” De La Vega added.

VOLTE

AT&T launched commercial VoLTE services in May 2014 in a handful of markets. That number increased by the end of the year to a few dozen areas, and the operator plans expansion throughout 2015. That’s in contrast to Verizon, again, which looked to first densify its LTE network (ensuring fewer handoffs) and then launch on a national basis. AT&T came out well during tests carried out in 2014 by Signals Research - which tested service in the Minneapolis- St Pauls area. It is committed to establishing VoLTE interworking with Verizon during 2015.

TMNQUARTERLY 13

SPONSORED FEATURE

SEVEN THINGS I KNOW ABOUT TRUE CUSTOMER EXPERIENCE MANAGEMENT

a company that brings customers and operators closer together by crowd-sourcing user experience data, enabling operators to be transparent about their networks.

With deployments racking up at Tier 1 mobile operators, Blake

Customers like being talked to - so engage, commit, close the loop

If you tell a customer why you have a problem, if you have a problem, where you have a problem, how long you think it will last - you are creating network transparency. Your customers will love you for it because you are no longer leaving them in the dark about when service might resume or improve. And if you don’t know why a customer is having a problem, then it’s OK to say so and promise to update them when you do know. Just make sure that you actually do get back to them with relevant information. That closes the loop and makes the customer feel that you have acknowledged their concerns and done something about it.

tells us what he has learnt about building network transparency, and the benefits of it for operators and their customers.

2 1

From Customer Experience to Customer Engagement

There are a lot of things out there that claim to be about CEM, but although they may give the operator some technical information about the network very often they are not about interacting at all with the customer. Customer engagement is about communicating with a customer so that the operator can find out where and when customers are experiencing problems, and then take action to deal with that and remain in touch with the customer. That’s something that many operators’ CEM systems currently don’t have the capability to do. The way you create that engagement is to give customers the ability to communicate with the operator in the manner they want to - through an app, web form, or call centre - about their network experience. The operator then responds to the customer immediately and transparently - creating that ongoing engagement. 28 TMNQUARTERLY

You must create a unified view for a transparent network

Not everyone likes to call a contact centre, not everyone likes to fill in a form; people like to report problems and issues in different ways. But the way the operator analyses and then uses that data must be consistent. This means that what you are telling your customers is based on the exact same data as your internal teams access. So your CRM teams see the same information as you would share to a customer through an on-device app, and your network management teams uses the same data to take action in the network. That creates a transparent network, where your customers access the same data you do - that’s their reward and their incentive for reporting their experiences in the first place.

It’s not all about managing faults

CEM isn’t just for when things go wrong. If you have crowd-sourced data that tells you about people’s experience of the network and the services they are consuming, then you can use that in a variety of ways across the business. You can provide point of sale validation to a customer, reassuring them that they will be able to access 4G at their office and home, for example. Some operators are using the data in their network expansion planning to know where to go next with their rollout after they have covered the obvious locations. This actually creates a network rollout tied to potential revenue, shortening the return on investment cycle on new network deployments. Engineering or service operations teams can also use the data to prevent outages - picking up on real time data, as reported by customers, to access visibility into a specific network node that may be coming under stress, long before element and network management alarms would present a ticket to the system.

We can’t do this without using the power of the cloud

If something does go wrong, crowd-sourcing customer experience data means being able to process, analyse and correlate hundreds of thousands of customer reports in a very short space of time. When an operator sees a major outage span one or two days, the number of queries coming onto the system will be constant and massive. A server farm would have to be built for that peak demand which is very inefficient. Because we have built a redundant, scaleable, elastic architecture in the Cloud, with SpatialBuzz operators can flexibly scale their discovery and reporting capability up and down, and always have the availability they require.

Transparency works internally as well as externally. We can break down communication barriers within the operator

One thing that happens when you start to share data is that you create transparency not just between the operator and subscriber, but also within and between operator departments. Unfortunately, CRM and network engineering teams normally only ever talk to each other if there’s an escalation of an issue. However, by creating a common view of the network based on actual customer reports, we break down the usual barriers to the flow of information between departments bringing together service management, network management, CRM and marketing teams. For example, a network engineering team may take action on an event before they would normally receive a job ticket. Or, because customers have stopped reporting issues, a CRM team can see a problem has been resolved before officially receiving the all-clear from the networks team.

Crowd sourcing and crowd shaping data forms a virtuous circle between operator and customers

When individual customers share their network experience, the operator can use that information, in near real time, to communicate back to the customer – creating not just the impression of openness, but real transparency. With increased transparency comes improved customer loyalty, lower likelihood of churn and an improved Net Promoter Score. This virtuous circle turns the one-way street of Customer Experience Management into a circular conversation between operator and customer, to the benefit of both.

WWW.SPATIALBUZZ.COM TMNQUARTERLY 29

FEATURE: VOLTE: PART II

I I T

R A P

FEATURE: VOLTE: PART II

LET’S LOOK AT THE REASONS WHY THEY WILL. Current or ex-CDMA operators definitely will because they don’t have the GSM circuit switch network to fall back on - hence they will be moving to a VoLTE-only model. If their users go abroad, they will probably want them to be able to make a call. LTE operators that sell VoLTE as being better quality will want to gain an advantage by extending that quality premium to both inbound and outbound roamers

Can VoLTE roaming break out? Slowly, operators are just beginning to introduce VoLTE and provide interconnect between themselves to enable LTE data roaming. The next step will be to introduce VoLTE roaming, but it is not yet a done

LTE operators that have enabled enriched communications services, things like video calling, group calling, IP messaging, file sharing, will in time want to be able to offer that functionality whilst their users are roaming. LTE operators that are turning off 2G and/or 3G to exploit the greater spectral efficiencies of LTE will (like their CDMA counterparts) be expected by their roaming customers to be able to provide a working voice service. It’s going to happen anyway, the standards are there, so let’s not get left behind.

deal how this will be implemented technically. Why does VoLTE roaming matter? At the moment most users would not expect to have access to LTE data services when roaming, let alone VoLTE. If they are even aware of how such things work most users would be happy, when roaming, to be able to place a call over 2G or 3G. This is what happens currently, of course. If you want to make or receive a call whilst you’re roaming then you are good to go. VoLTE roaming would mean being able to make a VoLTE call over the LTE network local to the country you are visiting, so that functionality enabled by the IMS-based service you are using remains supported.

30 TMNQUARTERLY

Is that such a benefit? A call is a call, right? Currently, if you have an LTE phone and can make VoLTE calls whilst on your home network, you are already in a minority. If you can make a VoLTE call to someone on another network you are in a tiny minority. If you can make a VoLTE call while roaming you are in the future. But not for long. Analysts differ on this but there are likely to be 100-200 million subscribers with VoLTE by 2018 or so - and you don’t have to be a genius to see that in general, most LTE operators are going to to follow up on LTE launches by enabling VoLTE. And then once they have enabled VoLTE, they will move to make sure that they can support both inbound and outbound roamers to use the service.

modelling an IP exchange on the same system as the GRX or GPRS eXchange that enabled 3G data roaming. One example of such an IPX provider is BICS, which currently has about 170 IPX customers. BICS’ Jorn Vercamert, who is VP of the company’s voice business, says, “When it comes to connecting voice we see that operators are now moving to 4G, and we are forecasting that by the end of 2016 50 mobile operators will have commercially launched voice over LTE services. The next step to move into international interoperability.” Now note this, Vercamert says this shift will “change the wholesale ecosystem” because with VoLTE the proposition is to offer enriched, engaging services where you can combine other media with voice, such as messaging and video. At the same time, operators can do this securely and ensure privacy, building what they hope will be a compelling counterproposition to OTT services. “That means when we move to VoLTE roaming this will reinvent the whole

ecosystem whereby the traffic will be going more transparently and more securely, with more privacy,” Vercamert says. It is this focus on VoLTE’s enhanced experience - better quality and security with enhanced latency-sensitive services - that has brought some to the conclusion that a different technical architecture might be required. One proposed method of ensuring this better experience is to enable something called Local BreakOut (LBO). LBO has actually been around for a long, long time as a roaming concept. As its names suggests it posits a direct connection to local services, thereby avoiding the “roundtrip” aspect of media connectivity. It has also been standardised into LTE roaming standards, with policy-to-policy connections on an interface known as S9, rather than the S8 interface that supports home routing (SEE DIAGRAM).

So VoLTE roaming is going to happen, just as VoLTE interworking is going to happen, just as VoLTE is going to happen. But how? You might expect VoLTE roaming to follow the same technical route as LTE data roaming, which is nearly always enabled by something called Home Data Routing, which as the name suggests routes control of the call back to the “home” network, which then hooks up the user to the service they are looking to access. This routing can be done through interconnect providers that provide both the control plane and signalling interconnect and the actual data plane connection. These providers operate something called the IPX, TMNQUARTERLY 31

FEATURE: VOLTE: PART II

Vercamert says that if it were a purely technical question, operators might well support LBO for VoLTE roaming - but there are commercial considerations for the operators. As an IPX, BICS’ job is to have the signalling support in place for both models. Another IPX provider, Syniverse, is willing to come down more firmly in the home routing camp. Its head of networks John Wick points out that LBO was specified for GPRS roaming, but was never used because home carriers preferred to have visibility into their customers’ traffic. “That same requirement still exists,” Wick says. “A home operators still wants a degree of visibility into what people are doing and with LBO that kind of disappears.” With LBO, as an IPX Syniverse would still see the signalling traffic but not the bearer traffic, and that impacts on operators’ ability to do anything about customer experience. “Having signalling traffic is nice but to be able to do any sort of VAS in the network we need to be session aware and that disappears with LBO,” he says. Wick adds that with multiple data centre availability across a region, an IPX provider can create a sort of de facto LBO, with traffic being served as locally as possible to its origination - pushing out functionality from the core to a more distributed edge. One proponent of LBO, and someone who even thinks it could be accretive for IPX providers is F5’s Peter Nas. Nas writes, There are opportunities around LBO for IPX carriers which must be looked at as LBO will reduce revenues from backhauled data. One interesting aspect of LBO is that the signaling for two additional Diameter interfaces, S9 for policy and Gy for charging, could be exchanged between visited and home networks, and if so, this will be done via an IPX network as per GSMA guidelines (IR.88). There are different views on whether or not using the S9 interface to exchange policy information between the visited PCRF and home PCRF, will be massively 32 TMNQUARTERLY

used once LBO is offered, but let’s assume it will be used. In this case, an IPX carrier can offer various services around Diameter interworking, security and perhaps also screening, overload control, prioritisation and potentially adapting policy rules and more. People who doubt the uptake of using the S9 interface might be more interested if IPX carriers could offer services that help visited networks control which policies can be applied in the home network or in the visited network. The offering around S9 signaling can be the same as what is currently offered for transporting and managing S6a signaling (for authentication and mobility management), but offers security, quality, and many more differentiating features. Another very interesting Diameter interface is the Gy, related to charging. Once LBO is offered, the visited network will have primary control of charging the visitor (for example to offer various ways how a customer can be charged, like per credit card,

voucher, scratch card, loyalty points, other credits from OTT applications, etc.) but there is also interest by the home network to receive charging information. Nas’s vision extends the role of the IPX beyond just transporting charing information, to build services such as security, priority, firewalling, interworking and overload control.

CONCLUSION LTE roaming is coming. VoLTE roaming is coming in its wake - not immediately but certainly some time after 2015. With that comes the potential for operators to offer a much better voice and communications experience whilst roaming. But to do that may require a new technical architecture, and may bring a big shake-up in the wholesale market. Yet opportunities still exist for IPX providers that could offer new services that take advantage of either home routing or LBO architectures.

LTE Roaming Scenarios:

Home Routed x Local Breakout (LBO) - Main Concepts

HOME PLMN VISITED PLMN

2G/3G

HOME ROUTED

LOCAL BREAKOUT (LBO)

IP NETWORKS

5Gi

PDN GW

SGi IP NETWORKS

SERV GW

Home Routed: “similiar” (in topology terms) that we have today in 2G/3G networks.

hPCRF S9

LTE

OTHER ACCESSES

2G/3G

PDN GW SERV GW

LTE

vPCRF

OTHER ACCESSES

LBO: procedure that allows to access locally, services in the visited network, given opportunities for optimise latency and transport network usage between HPLMN and VPLMN.

The mobile communications revolution is driving the world's major technology breakthroughs. From wearable devices to connected cars and homes, mobile technology is at the heart of worldwide innovation. As an industry, we are connecting billions of men and women to the transformative power of the Internet and mobilising every device that we use in our daily lives. The 2015 GSMA Mobile World Congress will convene industry leaders, visionaries and innovators to explore the trends that will shape mobile in the years ahead. We’ll see you in Barcelona at The Edge of Innovation.

WWW.MOBILEWORLDCONGRESS.COM

AN EVENT OF

LOY P E D

2018

413M

2013 = 1.5% VoLTE

‡

1.5%

34 TMNQUARTERLY

TMNQUARTERLY 35

NFV trials and demos, and what are the key challenges that NFV throws up?

NTT DoCoMo and Super Six

Korea Telecom and Alcatel-Lucent

Telefonica and Unica

NTT got right on the case, as you might expect, and moved into trial stages with a variety of vendors, in fact six pretty major ones in Alcatel-Lucent, Cisco, Ericsson, Huawei, NEC and Nokia Networks. Each vendor was given a role to play in Proof of Concept trials incorporating vEPC, so that the trials verified the interoperability of each vendor’s EPC software with a different vendor’s equipment, such as scalable data processing capability for congested data traffic and immediate network recovery in the event of hardware failure.

KT has begun an NFV proof-of-concept trial which will include Alcatel-Lucent’s vEPC, CloudBand NFV platform and Virtualised Services Platform (VSP) from the company’s subsidiary Nuage.

A year ago Telefonica was one of the first operators to state in a very public way its ambitions for NFV - labelling its activites in SDN-NFV under its Unica name. The operator garnered a lot of publicity for its vision, announced a whole clutch of major vendor parters and demonstrated some NFV use cases and concepts on its stand at MWC. Since then it has announced that it is building an NFV reference platform and lab with Red Hat and Intel, and hopes to have virtual private clouds and vIMS deployed commercially within 2015.

DOCOMO aims to have in place elements of a virtualised mobile network for the commercial deployment of services by the end of March 2016.

36 TMNQUARTERLY

AT&T and Domain 2.0 As you can read in this issue (Anatomy of an Operator, p10) AT&T has a whole programme underway dedicated to building a cloud-based network infrastructure. Its named vendors include the likes of Metaswitch, which has developed its open source Project Clearwater project to provide a version of the IMS designed specifically for the cloud.

A year ago Telefonica was one of the first operators to state in a very public way its ambitions for NFV

What have operators been doing with

NFV TRIALS & DEMOS

& s s e r g o r P > > n o i s s e r prog >>

FEATURE: NFV UPDATE

SKT and Ericsson, Samsung

Deutsche Telekom

Ericsson-LG and SK Telecom signed an MOU for R&D collaboration to verify Cloud Orchestration technology. Ericsson’s Cloud Orchestration is an NFV-based cloud solution that enables central detection, control, operation and optimisation of separately running applications and hardware infrastructure.

Deutsche Telekom carried out an NFV proof of concept in 2014 that demonstrated the integration of an IMS application provided by Huawei, with the CloudBand Platform provided by Alcatel-Lucent over a standard cloud infrastructure implemented in Deutsche Telekom facilities. The NFV POC demonstrated automated onboarding of VNFs provided by a single vendor and its integration into NFVI (NFV Infrastructure) in a multi-vendor setup.

SK Telecom has moved beyond a trial and said it will use Samsung to “supply a nationwide network function virtualisation (NFV) platform for its dedicated mobile internet of things (IoT) network”. The operator’s IoT network will be controlled by Samsung’s AdaptiV Core and is scheduled to go live in the first half of the year. SKT said initial applications will target package delivery tracking, CCTV monitoring and citywide sensor monitoring.

Telekom Austria The company is pushing NFV trials through its subsidiaries in the Balkans. Croatian operator VIPNet demonstrated HD video streaming made possible by combining LTE technology and NFV. In Bulgaria, Mobiltel also used HD video streaming as a use case for a vEPC trial.

OPENSOURCE NFV Opensource NFV projects on the rise

The initial project objectives are to:

At the end of September 2014 the Linux Foundation announced the founding of the Open Platform for NFV Project (OPNFV). OPNFV is a joint initiative with over 40 carrier and vendor partners. OPNFV’s aim is to “establish a carriergrade, integrated, open source reference platform that industry peers will build together to advance the evolution of NFV and ensure consistency, performance and interoperability among multiple open source components.” The initial scope of OPNFV will be on building NFV infrastructure (NFVI) and Virtualised Infrastructure Management (VIM) leveraging existing open source components where possible.

>> develop an integrated and tested open source platform that can be used to investigate and demonstrate core NFV functionality; >> include proactive participation of leading end users to validate that OPNFV meets the needs of the end user community; >> contribute to and participate in relevant open source projects that will be leveraged in the OPNFV reference platform; >> establish an open ecosystem for NFV solutions based on open standards and open source software; and >> promote OPNFV as the preferred open reference platform.

TMNQUARTERLY 37

ETSI is now into the second phase of its work on NFV, having completed 11 specfications documents during its first phase.

ETSI

ETSI enters NFV Phase 2

C-RAN trial results

Work has started on producing 28 new documents and more are expected over the coming 2 years. The main objectives of this second phase of work are to encourage interoperability, building upon the achievements made in the first two years of the ISG. This new phase of work will include both normative and informative content. Adding normative requirements will provide the clarity required for interoperability and formal testing. Another key element of Phase 2 will be to work more closely with other standards bodies to help focus their work and to avoid duplication, and to collaborate within open source projects developing NFV implementations.

Power consumption down 41%

Rental expenses down 71%

Capex down 15% overall

Opex down 50% overall

d u o l C > n a R

Overview of China Mobile Cloud RAN Activities When it comes to virtualisation of the radio access network, no operator other than China Mobile (CMCC) has perhaps been more supportive of Cloud RAN; and for good reasons. China Mobile boasts over 800 million subscribers served by 900,000 GSM and 500,000 TD-SCDMA base stations. By mid 2014, China Mobile operated 410,000 TD-LTE base stations as part of a plan to roll out 500,000 by the end of 2014. However, like many wireless network operators, EBITDA margins are eroding and stand at 38% in 2013, down from 51% in 2009. Cloud RAN provides a solution to control costs by centralising baseband function in data centers to reduce energy, site rental and operational expenses. Cell sites account for 72% of CMCC energy consumption, 46% of which is due to HVAC. In one trial in Changdun where 506 GSM base stations were aggregated in a few centers, CMCC reported 41% reduction in power consumption due to shared air conditioning. In a second trial in Zhuhai where 18 new TD-SCDMA sites were rolled out in centralised configuration, site rental expenses are reduced by 71%. Today, 40% of the Zhuhai 3G network is centralised. In general, CMCC estimates that C-RAN saves up to 15% in capex and 50% in opex. As cell sites only have remote radios and antennas on premises, site roll out

time is reduced by a third. In addition to cost savings, Cloud RAN enhances performance especially at the cell edge. Field trials by CMCC show uplink CoMP gain for a 6-macrocells cluster is between 40~100% in weak coverage areas corresponding to increase in throughput between 1 â&#x20AC;&#x201C; 2 Mbps. Similar if not higher gains are expected for heterogeneous network deployment where low power remote radios are used as small cells. CMCC continues to develop its strategy for Cloud RAN by conducting a number of field and market trials. In Jilin, 1400 GSM sites were centralised. In Guangzhou, CMCC deployed the first 3G/4G dual-mode network where the same remote radio is shared between the two technologies. Today, there are at least 5 field trials for TD-LTE in centralised configuration to evaluate different aspect of Cloud RAN. In these and other trials, CMCC relies heavily on its fiber assets for fronthaul. These assets will prove pivotal for the implementation of Cloud RAN as a single 20 MHz TD-LTE channel with 8x8 MIMO configuration requires 9.8 Gbps fronthaul data rate. Optimising the architecture of Cloud RAN to maximise its advantages while leveraging existing fibre infrastructure to minimise additional investments, is an evolving topic of study for the months to come. By Frank Rayal, XONA Partners

TMNQUARTERLY 39

FEATURE: EXECUTIVE INTERVIEW

&A INTERVIEW QEXECUTIVE What’s coming up for the SCF at MWC? This is an important year for us and and our MWC plans reflect our priorities. We’ve got a packed speaker programme, events for operators and important announcements to make. Everything we’re doing right now is about evangelizing how we’ve established the drivers and business cases for all the main use cases. We’ve dealt with the technical challenges so now it’s about getting carriers to seize the opportunities small cells enable. MWC is about getting the word out there to carriers worldwide.If companies have a strong interest in small cells then come to our stand. (Stand 7F61) Where is SCF with the Release Programme? What is next? We’re launching Release 5 at MWC with a whole new set of documents looking at the deployment of small cells in rural and remote environments. This covers a wide variety of potential use cases that can be extremely valuable to carriers – from classic rural locations, to isolated locations such as cells on planes, trains, remote industrial, disaster recovery and that kind of environment. It’s about connecting the unconnected.

40 TMNQUARTERLY

Alan Law, Chair of the Small Cell Forum on: SCF Release Programme Market momentum Deployment challenges New technology developments 5G

With the release program we really have cracked the small cell deployment challenge. We’ve looked at all the barriers, presented business cases and given the operators everything they’ve asked for. What the whole process has done is create a huge opportunity for anyone deploying small cells anywhere in the world. The focus now is taking the work out to carriers worldwide with an immediate focus on Asia, Latin America and the Middle East. There’s real momentum to our work and industry right now. Our next big piece of work is on virtualization of the small cell layer and the first phase of that work is now underway. We also have a fairly lengthy list of operator requirements including

everything from smarter WiFi integration, to M2M and more work on SON. This is what we’ll be talking about at MWC. How many Release downloads have you had to date? The programme from R1-4 has had 80,000 downloads in total and 52,000 in the last 12 months alone. The areas that tend to be popular are use cases for deployment that are effectively testimonials from operators and vendors discussing scenarios that have worked for them. With a focus on applying small cells in a particular environment, the Releases give everyone the tools they need to make the business decisions to get the most out of small cells.

What do you think have been the more significant small cell deployments in the past year? There have been a lot of developments that open up the potential for small cells. If you look at the market as a whole for example, FierceWireless are suggesting that Sprint could spend up to $500 million per year on small cells. So it’s great to see small cells being heartily embraced by operators that have understood they provide the capability needed to drive a new wave of capacity into networks. In the last 12 months we’ve also seen a real strengthening of plans and deployments on urban and enterprise. It is good to start to see large material deployments as operators grasp the importance small cells have on their business. It’s also interesting to see signs of market maturity as companies come together, such as the announcement between ip.access and Amdocs where they are bringing planning capability as well as small cells together to enrich delivery. We are still very early in the deployment journey, though: if you take the Sprint example it is looking at more activity in 2016 than 2015. What do you think we will see this year in terms of deployments? Many operators are working out their plans privately and looking at what’s the best way to materially realise capability in their market. Over the next 12 months we will see more and more announcements from operators on those plans and ambitions. What do operators still see as “challenging” around small cells? Operators have a strong understanding of aspects like the products themselves, backhaul, site rental. The individual components are there, the challenge is effectively bringing that together, setting up the business processes to support

large scale rollouts. It’s less about technology and more organisational. Operators are used to the macro deployment model and the prospect of deploying very large numbers of small cells means reconsidering how you deploy assets and your working processes. This is precisely what the Small Cell Forum sets out to address. Then as the technology continues to evolve operators have got decisions

to make on things like LTE-A and the ability use unlicensed spectrum to complement licensed assets. Virtualisation, the vision of extending flexibility through increased use of software functions deeper into the access network, means we need to understand how to embrace small cells with that model. In these areas as a Forum we provide guidance to make sure we maintain and develop market understanding.

Given that today small cells comprise discreet femtocells, picocells, clustered cells under a controller, integrated cellular-WiFi access points, Centralised-RAN, Cloud-RAN (virtualised BBU), “DAS-like” indoor/outdoor solutions, does the term “small cell” have any value anymore? And who for? There’s value in defining small cells in the sense that a lot of the challenges apply equally whether deployments are femtocell or a classic small cell or a DAS-like small cell. There’s real value in being a centre of knowledge to understand the challenges of all different types of technologies, to use the Forum to provide insights from one type of solution to another type of solution. As 2020 network discussions really take off, what is role of the SCF? Many of the 5G discussions seem relevant to Forum outputs already – do you have a hook into any of the research and/or standards bodies looking at 5G? We are building formal and informal relationships with a number of bodies working with 5G, with two thoughts in mind. First, to understand the way 5G develops and the certain parts that are relevant to small cells. Secondly, we have learnt a lot about the challenges you face in deploying small cells and there is strong value in the experiences we can reflect back into 5G bodies. Typically with very high throughput rates, and as you move up frequency bands into the tens of GHz bands, you need a much better channel. And if you need a better channel you need to be closer to the radiating element, and that brings us full circle back towards small cells. So we are not trying to be a 5G body, but there’s certainly knowledge we could share.

You can continue the discussion with the Small Cell Forum at Stand 7F61, Mobile World Congress and at www.smallcellforum.org

TMNQUARTERLY 41

FEATURE: CUSTOMER EXPERIENCE MANAGEMENT

Do Operators Really Care About Their Customers.... Really?

Telcos may have a poor record on customer satisfaction, but they are trying to get better. The issue is, you don’t just buy “Customer Experience Management“ – there’s a complex value chain involved. By Sandra O’Boyle.

42 TMNQUARTERLY

orgive the cynical headline, but it’s no secret that the telecom industry scores poorly on customer satisfaction and even caring about the customer experience, compared to other industries. There are signs of this changing with Customer Experience Management (CEM), big data analytics, real-time application and network performance and service personalisation fast becoming hot topics. Vodafone’s CEO Vittorio Colao is on record as being firmly behind delivering the best customer experience, something he believes will make the difference as Vodafone seeks to reverse revenue declines in highly competitive European mobile markets. Vodafone will invest in new IT systems, including a new billing system, to ensure a consistent customer experience in store, on the phone or online. Vodafone will measure success in providing the best customer experience based on Net Promoter Score (NPS) i.e. would the customer recommend an operator to friends and family. Anil Rao, Senior Analyst at Analysys Mason, says, “CEM is definitely a mainstream topic in mature mobile markets such as North America, Western Europe and the Middle East and is gaining prominence in some developing markets,

where senior executives are driving a focus on customer experience, top down. With the massive growth of mobile data, triggered by LTE and Wi-Fi, and fuelled by the OTT‘s IP messaging, voice and video applications, together with the proliferation of smartphones, consumers now demand high levels of quality of experience, service personalisation and control, raising the bar in a way for the mobile operators. CEM is increasingly becoming a critical driver for both service differentation and churn prevention.“ Operators continue to invest heavily in their core network infrastructure but an increasing share of spend is going towards IT systems, OSS and BSS and real-time network and application performance insight as well as online charging and personalised customer support. “European operators appear to be ahead of operators in the U.S. when it comes to CEM investments,” according to Shira Levine, Directing Analyst at Infonetics. “Customer churn rates remain very low and stable in the U.S. so operators are not feeling the same pressure. There is momentum behind CEM analytics to identify new services and revenue opportunities, though regulatory uncertainty around customer data is an inhibitor“.

Breaking CEM Down

ne of the biggest headaches is complexity – CEM is hard to do - an operator cannot go out and buy a CEM system that will accurately report on customer experience. In reality, getting to a single truth of whether you have a satisfied or dissatisfied customer, involves a vast value chain of technology and tracking events in a customer’s journey from the network RAN, core, IMS, OSS, BSS, and even performance monitoring on the device itself. IP probes are being deployed to scan network traffic and use deep packet inspection (DPI) techniques to gain application level sessions, this requires fast processing of data at a near real-time level. Operators can then build a customer profile of behaviour to learn more about their most valuable customers, what time of day are they using certain applications, what are they using their smartphones for - Spotify, Netflix, WhatsApp - at what locations which can drive support for future network investments that will improve performance and boost customer satisfaction. A richer picture of a mobile customer emerges when the content and user plane data from probes is combined with actual performance data, and then linked with CRM systems. Analytics can happen on ARPU, bundles, billing, patterns and frequency of calls to customer care. Another aspect could be social network analysis to identify influential users’ propensity to churn. Operators such as Orange are investing in heavy duty real-time analytics engines such as Hadoop to crunch all of their structured and unstructured data and develop their own customer experience

TMNQUARTERLY 43

FEATURE: CUSTOMER EXPERIENCE MANAGEMENT

analytics. These KPIs and reporting interfaces can be tailored to provide insight for different business units such as customer care, marketing and network operations and engineering and executives. Eventually KPI thresholds can be linked to executive actions that can be automated and triggered. This is still for the future, in reality the majority of mobile network operators are sitting on vast amounts of data that they are not adequately monetising and gaining customer insight from.

CEM requires a Human Effort

ven after choosing the best technology, operators can still fail to deliver a great customer experience, if the entire organisation and culture is not behind it. For many operators, this requires a culture change and complete re-think on how services are created and launched, even redesigning processes to get an end-to-end view of the customer. Top management support is critical, as well as hiring new talent in customer experience and big data analytics and involving IT and system integrators. “Aligning the organisation around customer experience is still a work in progress for many operators“ says Thomas Vasen, Senior Director of Subscriber Experience Marketing at Procera. “An operator will hire a Chief Experience Officer but their immediate concern is brand consistency rather than looking at the customer experience journey. This starts from the network experience when they use the service right up to enabling customer support to see that subscriber’s experience when they do call in. This insight can have a real bottom-line business impact e.g. by reducing average call length to customer support.“ Vasen believes the operators are waking up to the value of CEM. OTT content guys such as Netflix are turning the tables and measuring the network operators, scoring and ranking their networks on streaming

44 TMNQUARTERLY

video performance. That’s another strategic reason why operators have to care about the end customer network experience. Increasingly operators are signing commercial partnerships with these OTT content providers so the customer experience matters and a consistently poor experience can be used to negotiate better terms.”

So what’s next for operators that care? Things are moving fast in the mobile world and operators realise they cannot afford to be complacent about customers or content partners. Leading from the front and innovating with CEM and driving business change with analytics will be just as important as launching innovative new services and having the fastest network. Net Promoter Scores (NPS) will remain a key external source of truth of what the customer really thinks about an operator i.e. would they happily recommend them? The next frontier will be bridging the gap between what an operator thinks based on all the CEM analytics and data collected and how the customer actually feels about their experiences in NPS ratings. Operators will want to understand how to use their CEM toolkit to make customers influencers, change perceptions and positively impact the NPS. Operators will also look to take action in real-time based on having richer contextual marketing data around what subscribers are doing and their location. This could be in the form of customized contractual offers, upsell offers, locationbased applications and promotions.

FEATURE: NFV ORCHESTRATION

How to get the data you need for the Customer Experience you want to deliver: 1. IP probes deployed to scan network traffic and use deep packet inspection (DPI) techniques to gain application level sessions, this requires fast processing of data at a near real-time level 2. Performance data from element traces, NMS and network monitoring probes. Is the network working as it should. Is the operator isolating location and time and impact of issues on the network 3. Customer data on-device and customer

reporting. Operators can build a richer picture when content and user plane data from probes is combined with actual performance data and then linked with CRM systems. Analytics can happen on ARPU, bundles, billing, patterns and frequency of calls to customer care 4. Analyse and produce KPIs and reports. Heavy duty real-time analytics engines

such as Hadoop crunch structured and unstructured data and develop customer experience analytics.

CEM MARKET UPDATE:

TMN has produced a full-length report on the CEM landscape. Scan here to view.

NFV ORCHESTRATION PI C K I N G U P T H E If NFV is to achieve more than moving code from appliances into common standard hardware, it requires a great deal of proof of scaleability and the ability to expand and dedicate resources as required. That’s the whole point of virtualised network functions, of course. To date, much of the noise around NFV has been about this or that vEPC or even vRAN element. Of course that’s great, but it doesn’t do much to move to the wider questions of NFV - how these elements will be managed and controlled as part of a whole. The issue is that SDN and NFV, taken together, break apart the existing network model, which means the existing network management model is also broken. To put it back together, the industry has reached for the term Orchestration - taken from the computing world, of course. This network orchestration, which can manage the virtual functions themselves as well as the physical elements underlying the network, and also control the service architecture (including service chaining), is being defined within ETSI’s MANO (Management and Network Orchestration).

B A T O N

Orchestration is the automated management of networks. In NFV, it is becoming a key part of determining how NFV elements, and the services accessing them, can be deployed and managed.

TMNQUARTERLY 45

FEATURE: NFV ORCHESTRATION

Michael Howard, principal analyst, Infonetics Research, says that the importance of orchestration is that it can unlock the potential benefits of cloudified networks. “While operators are certainly attuned to the capital and operational benefits of these transformative technologies, it is the orchestration of virtual compute, storage, and network, combined with the automation of existing network resources that will lead to new, higher-margin revenue streams and happier customers.” Tom Nolle of CIMICorp, who has probably spent as long thinking about NFV as anyone on the planet, points out that you can organise services across both horizontal federations of controllers

and vertically down through the layers of network protocol/technology in three ways; policy-linked structures, domain federation requests, and orchestration. In his view: “The orchestration model says that the responsibility for creating a service doesn’t lie in any layer at all, but in an external process (which, for example, NFV would call “MANO”). The service request from the user invokes an orchestration process that commits resources. This process can “see” across layers and commit the resources where and when needed. The continuity of the service and the cooperative behavior of the layers or controller domains is guaranteed by the orchestration and not by interaction among the domains.”

NFV MANAGEMENT AND ORCHESTRATION ARCHITECTURE

OSS/BSS

NFV ORCHESTRATOR (NFVO)

NS CATALOG

VNF CATALOG

NFV INSTANCES

NFVI RESOURCES

VeNf-Vnfm EMS VNF

VNF Manager (VNFM) VeNf-Vnfm

Vn-Nf NFVI

Nf-VI

Vnfm-VI Virtualised Infrastructure Manager (VIM)

Or-VI NFV-MANO

EXECUTION REFERENCE POINTS OTHER REFERENCE POINTS MAIN NFV REFERENCE POINTS

46 TMNQUARTERLY

Operators will feel that they have, intrinsically, some of this functionality already in their OSS and network management systems. But the difference is in how these will behave in networks of virtualised functions controlled by software controllers. NFV Orchestration allows for: the configuration, provisioning, and chaining of virtual network functions in addition to other resources required for a service, the determination and selection of an optimal physical location and platform on which to place VNFs, mapping of the instantiation of virtual network functions against real-time demand and full lifecycle management of the VNFs, ie. the creation, instantiation, and monitoring of the VNF until it is decommissioned. The ETSI MANO framework breaks down the management and orchestration needs for the NFV architecture into three functional layers: Virtualised Infrastructure Managers (VIMs): This layer handles the virtualisation of physical hardware by integrating with virtual-machine managers. Using a hypervisor, the virtual-machine manager provides the ability to create multiple virtual compute, network, and storage elements. The virtual machines provide lifecycle management functions (create, edit, delete, start, and stop) for the virtual data center elements related to compute, network, and storage functions. VNF Managers (VNFMs): The VNF managers handle the configuration, lifecycle management, and element management of the virtualised network functions. NFV Orchestrator (NFVO): The orchestrator provides lifecycle management of the network services that includes instantiation, scaling,, performance measurements, event

correlation, resource management, validation and authorisation for resource requests, and policy management. So in ETSI’s MANO (Management and Orchestration) architecture, the Orchestrator is responsible for onboarding network services and service lifecycle management (instantiation, scale-out/in, performance measurements, event correlation, termination). The VNF Manager, as it sounds, manages VNF instances, coordinating between the VNF Infrastructure and element management systems.

VENDOR APPROACHES Amdocs views its NSC Orchestrator as forming both the NFV Orchestrator functional block and that of the Virtual Network Function (VNF) Manager (although the orchestrator can also work with other VNF Managers). Shaul Rozen, Director Of Product Strategy, Network Cloud Service Orchestration, Amdocs, said, that the Amdocs Orchestrator has created a “unique” network modelling scheme that enables carriers to model choice into the network diagram from the network service catalogue. For example a user could create a model by choosing either a Checkpoint or Juniper firewall, Rozen said, and then define in the model the security services to be applied (firewall, DDOS protection etc). Rozen said Amdocs is also working to build a partner ecosystem, with 22 partners from VNFs (vEPC, vIMS, vCPE) and from NFV infrastructure and hardware providers. This will provide pre-integrated NFV Proof of Concepts, Rozen said, to enable carriers to move forward quickly with NFV deployments. In one example Rozen cited, a call recording service was created on a vIMS (from Metaswitch) that

was modelled into the Amdocs catalogue and instantiated within five days. So if you are open, why the need for a pre-integrated partner approach? Rozen said it is really about benefitting from shared experience, and to get around the lack of standards available so far. “Every partner comes with a different VNF Descriptor (NFV-D), and most don’t even have a VNF-D so we are creating best practices to drive the industry forward. Another issue is that our partners are looking for Proof of Concept and it’s hard to show proof of Orchestration by itself and so for that go to market we needed to bring partners on board.” As for the issue of next generation OSS and network management in NFV-SDN networks, Rozen said that he thinks Amdocs is ahead of the curve. “We see ourselves as covering the orchestration and the VNF manager, as well as some even more northbound functions bordering next gen OSS because we are very much serviceorientated. I can tell you we will hear more and more talk in ETSI within the next year about real time OSS and next gen OSS, because if has not been tackled enough. So we are ahead of ETSI.” Amdocs is certainly not alone in targeting service orchestration. There have been a clutch of product launches over the past months.

CYAN Cyan’s Planet Orchestrate, launched before the Amdocs product and commercially available towards the end of 2014, is also designed to fulfill the roles of the orchestrator and VNF manager functionality described in the ETSI NFV ISG architecture. Cyan’s vision is to be able to support VNF management and orchestration across one or more data centers, as well as distributed architectures where VNFs may be

deployed at the customer premise edge. Cyan took part in two ETSI-approved operator-sponsored proof-of-concept demonstrations in 2014 including “End-toEnd vEPC Orchestration in a Multi-vendor Open NFVI Environment” and “Multivendor Distributed NFV.”

OVERTURE Orchestrator is an NFV service lifecycle management and orchestration system that coordinates virtual resources and physical network elements to create, activate and assure services using one or more virtual network functions. It uses the OpenStack cloud controller – bundled with ESO – to manage the virtual compute environment, including virtual machines, virtual switches and top-of-rack data center switches. For management of the physical wide area network traffic flows, ESO leverages Overture’s Ensemble Network Controller, but it can also be integrated with other third-party network controllers. Overture is also part of an ecosystem approach being delivered by WindRiver. NetCracker’s Service Orchestrator is described as “unique” by NetCracker because “it enables end-to-end orchestration of service provisioning and assurance for hybrid networks made of both virtualized SDN/NFV-based components and traditional network technologies.” This is clearly designed with the migration market in mind, and also potentially as a protective play of the company’s widely deployed OSS and telecoms management software. The product provides configuration management, service chaining, capacity management and real-time service provisioning as well as policy and analytics capabilities that orchestrate and execute network and service policies in response to live analyses of network and service conditions and events.

TMNQUARTERLY 47

FEATURE: GERMANY

software development for radio applications.

he story of Germany’s mobile industry in the operator space has been one of acquisition and now consolidation. First came the acquisitions as Europe’s major Group operators entered the country to do battle with incumbent Telekom. Famously there was Vodafone Airtouch’s mega-deal to buy Mannesmann in 2000 - at the time the largest corporate deal in history. O2 Germany, formed out of the ex BT Wireless, was then acquired in 2005 by Telefonica and in 2011 Telefonica merged that holding with broadband player Hansenet (owned by Telecom Italia). All the while KPN had been building E-Plus out from its base as the second oldest operator in the country.

50 TMNQUARTERLY

And so the country arrived at a situation where four main operators were established - with Telekom and Vodafone as leaders in terms of market share, and E-Plus and O2 as roughly equal third and fourth players. And then in late 2014 came the deal that would see Europe’s largest mobile market move from four major national operators to three - the merger of Telefonica and O2. At a stroke, this deal would move the two lowest ranked players to a position where they lead the market, relegating Telekom and Vodafone in the process. The combined entity would have approximately a 39% mobile customer and 32% mobile service revenue market share in Germany. The deal would also see the new entity become established as a major spectrum holder in the country.

VODAFONE

SUBSCRIBERS

34MILLION SUBSCRIBERS

EPLUS

upon a strong foundation of test, silicon development and

37MILLION

25MILLION SUBSCRIBERS

O2 TELEFONICA

Germany’s mobile network technology sector is grounded

in looking starved of the ability to generate increased service revenues, this merger, and with it a shake up of the market dynamic, may offer an opportunity of sorts. One boost could come from market opportunities opened up by new spectrum coming to market. In February 2015, Germany became the first country in Europe to have announced plans to auction 700MHz spectrum, and the sale is expected to raise as much as EUR4.5 billion considerably more than the EUR€3.58 billion garnered from the sale of 800MHz licenses in 2010. Away from the service provider sector, Germany’s mobile network technology companies have continued to provide innovation to the market, fostered by the large academic and research facilities in the country, and by encouragement from, especially, Telekom’s T-Mobile, which has always proved a rigorous proving ground for any technology company. Telekom’s commitment to developing new technologies - for example in OSS transformation, in NFV, in access technology - has given it a very attractive status as a key reference customer in the industry. Get through the labs at Telekom, and you will attract knowing nods from the industry sages. It’s hard not to see a link between this commercial-academic nexus and the sheer number of companies (see listings following this article) that represent Germany’s mobile network tech sector on a global stage. If Germany “lost” its main equipment provider a while back, with the exit of Siemens Networks, it retains a presence from semiconductor, to a host of test and certification players, to companies developing software for network operation and management.

MOBILE OPERATORS

FOUR INTO THREE, A MOBILE TECH SECTOR CONTINUES TO THRIVE

Before merger, E-Plus held 10MHz spectrum in the 900MHz band; 54.8MHz in the 1800MHz band; 44.6MHz at 2.1GHz band, 30MHz at 2.6GHz and 84MHz in the 3.5GHz band. O2 Telefonica was proposing to add to that its own 20MHz in the 800MHz band; 10MHz in the 900MHz band; 34.8MHz at 1800MHz; 48.9MHz at 2.1GHz and 50MHz in the 2.6GHz spectrum band. According to Germany’s regulator, a company formed by a Telefonica and E-Plus merger would therefore own 28.8% of the spectrum available in the 900MHz band and 63.8% of 1800MHz spectrum - over two times Deutsche Telekom’s 1800MHz holdings and eight times more than Vodafone’s. In the 2.6GHz LTE band, it would hold 42% of the spectrum, although it would remain an equal minority holder in sub-1GHZ bands with 33% of the 800MHz and 29% of the 900MHz spectrum. With such a position to work out, it was not surprising that the deal took around 14 months from initial firm proposal to clearance of an EUR8.6 billion deal. Telefonica said it would derive around EUR5 billion in “synergies” from the merger. To get the deal done the company had to sell back to the market around 30% of its overall spectrum holdings with Drillisch, up until then a MVNO, taking ownership of that spectrum. The merger is significant not just internally in the German market, but as a marker for how other countries in Europe might enable consolidation in their respective markets. For example the UK seems likely to see similar demands made of the operators if a merger with Three and O2 comes together. Certainly, with organic growth hard to come by and the overall market tracking most Western European markets

TELEKOM

FEATURE: GERMANY

20MILLION SUBSCRIBERS

TMNQUARTERLY 51

FEATURE: GERMANY

GERMANY’S MOBILE NETWORK 5GPPP As its name might suggest this is a joint research body looking to drive 5G development. It is not, though, a standards setting body in the same way as 3GPP. Initiated by the EU Commission, 5GPPP includes manufacturers, operators, service providers, SMEs and researchers. www.5g-ppp.eu EANTC EANTC provides vendor-neutral testing of network equipment, systems and software. Tests include prototyping, verification and certification. Clients might include vendors who want independent verification of a new iteration of software, to work for standards bodies proofing use cases for new technologies such as NFV. The company was formed as a spin-off from Technical university Berlin and services include areas such as Multi-Protocol Label Switching (MPLS), IP (Internet Protocol) switching, Triple Play, Voice over IP and mobile network technologies. www.eantc.de

Adva Optical Adva Optical’s AnyCell connectivity provides precise timing and sychnronisation to C-RAN and fronthaul architecture. The company’s range of networking solutions include, timing and synch for mobile backhaul, Carrier Ethernet and optical transport solutions, as well as service and network management. www.advaoptical.com

Avinotec The company provides mobile video technology for multiple platforms. It claims it compresses video streams so small they can be delivered over GSM and 3G networks, as well as LTE networks. On GSM it says it can deliver 15 pictures per second, with only 64 kbit/s. Applications include video-assisted living, traffic and security. www.avinotec.de Infineon One of a few companies amongst Germany’s Semiconductor company. Infineon makes many solutions targeted at the wireless sector in terms of performance and security requirements. Customers include device manufacturers, operators, design houses and smart card suppliers. www.infineon.com

Cetecom A test specialist, CETECOM was first established in 1994 in Essen, Germany as RWTÜV’s telecommunications division with the development of GSM technology as its objective. Known as the leading test laboratory in the telecommunications and information technology industries, cetecom expanded its portfolio to cover the entire life cycle of a mobile communication product. www.cetecom.com Fraunhofer Heinrich Hertz Institute The HHI carries out research and development into advanced communications technologies, including photonics and mobile broadband systems. R&D-activities in the area of mobile broadband systems focus on design aspects of future cellular systems such as multi-antenna systems (MIMO) up to 1 Gbit/s, resource management for both cellular mobile, radio systems and WPAN-systems up to 60 GHz. www.hhi.fraunhofer.de

52 TMNQUARTERLY

MIMOon MIMOon provides software implementation and system design for SDR (Software Defined Radio) platforms, specialising in LTE PHY layer and software Stack for devices and for small cells. Customers in small cells include recent wins NuRAN and Nash Technologies. www.mimoon.de

Coriant Optical communciations specialist Coriant places much of its R&D and manufacturing capability in Germany. The company is a coalition of Nokia Siemens Networks Optical Networks Business, Sycamore Networks Solutions and Tellabs - brought together by private equity house Marlin. Coriant solutions include smart edge routers, and multiservice aggregation (SDH/SONET, OTN, IP/Ethernet, MPLS-TP, PTN), switching and transport solutions powered by Software-Defined Networking (SDN) intelligence, automation and control. It enables network operators to build resilient, flexible and scalable (100G and beyond) transport networks optimised for the challenges and opportunities of the hypergrowth era. www.coriant.com

ipoque ipoque’s expertise lies in application recognition and classification, via DPI technology. ipoque’s protocol and application classification and analysis engine (PACE) enables network infrastructure and security vendors to develop products with intelligent bandwidth control, prioritised quality of service delivery and reliable network security. It is now a subsidiary of Rhode & Shwarz, where it is being increasingly integrated into the company’s strategic direction, but was developed as a separate, standalone company. www.ipoque.com

Node-H Node-H develops software for UMTS, LTE and dual-mode small cells, giving small cell manufacturers the flexibility to add their own code to their products, or go to market with the proven solution. For example the Node-H HSPA+ solution is currently available for the UMTS/HSPA+ SoC solutions of Broadcom and Qualcomm, and available for Broadcom’s LTE version. www.node-h.net

Orga Systems Charging systems provider that provides real time billing, policy and charging control capability to operators. This Policy, Charging and Control functionality becomes increasingly critical to operators to develop offers such as shared data tariffs, tiered and metered charging, and responsive, personalised offers. www.orga-systems.com Kathrein One of the leading antenna manufacturers worldwide, Kathrein is known in the mobile industry for its high quality filters, combiners, amplifiers and antennas. It is a supplier to many of the world’s leading OEMs for applications such as indoor coverage via a micro-CRAN, as well as the full range of spectrum support for macro and micro outdoor applications. www.kathrein.de

Nokia Not to be forgotten that despite a good portion of Nokia’s cuts falling within Germany, where of course the exSiemens Networks expertise resided, there are still major competency centres in Germany. Most recent, and notable, is the opening of Nokia’s Security centre in the country, providing a global test hub and competency centre for all things related to network security and integrity. www.networks.nokia.com

Rohde & Schwarz One of the industry’s major test and assurance companies - with a wide range of solutions for wireless network and equipment conformance and assurance. Solutions range from network drive test and automation, to infrastructure test to conformance and verification. Headquartered in Germany, this is one company with a true global presence. www.rohde-schwarz.co.uk

Lantiq Another semiconductor company, this one has just been bought by Intel, attracted by the potential of the company’s solutions for the connected home market. Lantiq’s System-on-Chip solutions address a wide variety of technologies, including WiFi and LTE gateways. Intel said having Lantiq on board would expand its success in the cable residential gateway market and broaden its offering to other gateway markets, including DSL, Fiber, LTE, retail and IoT smart routers. www.lantiq.com

Opticom A spin-off of the Fraunhofer-Institute, Opticom is a developer of standardised audio quality testing technology and methodologies - providing a means of perceptual measurement. Its POLQA, Perceptual Objective Listening Quality Analysis (FR) is, according to ITU-T Rec. P.863, the quality testing standard for HDVoice, VoIP, 3G and 4G/LTE. www.opticom.de

Qosmotec Qosmotec supplies operators with test automation systems for call-generation, load-tests, international-roaming, and handover tests and have been employed to test specific network technologies like GSM-R, TETRA and Car-2-X communication in Intelligent Transport Systems (ITS). www.qosmotec.com

TMNQUARTERLY 53

Virgin, Qualcomm, plan to cover the world with broadband Plan will see WorldVu launch 648 satellites to provide service to LTE/WiFi access points (for more details see page 6)

TMNQUARTERLY 55