Ce diaporama a bien été signalé.
Nous utilisons votre profil LinkedIn et vos données d’activité pour vous proposer des publicités personnalisées et pertinentes. Vous pouvez changer vos préférences de publicités à tout moment.

Ericsson Technology Review, issue #2, 2016

504 vues

Publié le

The latest issue of Ericsson Technology Review covers a wide range of topics including narrowband Internet of Things, the next-generation central office, telco-grade platform as a service, 4G/5G RAN architecture, and cloud robotics enabled by 5G. The feature story – Five trends shaping innovation in ICT – presents what I consider to be the major technology trends that will stimulate innovation in the coming year. Do you agree with me? I’d love to hear from you with any feedback you might have.

If I were to suggest one takeaway from all of the articles included in this issue, I would say it is speed. Device processing is getting faster, data speeds are constantly increasing and radio speeds are approaching those of fiber. More people are becoming subscribers, more things are becoming connected and more applications are running constantly. Developers of new technologies are working hard to enhance responsiveness by reducing latency, a key performance parameter. The capability to determine which functions can be virtualized to maximize ideal placement in the network and ensure low latency is one of the primary driving factors behind the proposed split of radio-access architecture discussed in this issue.

As always, I hope you find our stories relevant and inspiring.

Publié dans : Technologie
  • Soyez le premier à commenter

  • Soyez le premier à aimer ceci

Ericsson Technology Review, issue #2, 2016

  1. 1. #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEW 1 STANDARDIZING NARROWBAND ✱ ERICSSON TECHNOLOGY C H A R T I N G T H E F U T U R E O F I N N O V A T I O N V O L U M E 9 3 | 2 0 1 6 – 0 2 FIVETECHNOLOGY TRENDS SHAPINGICTINNOVATION CLOUDROBOTICS ENABLEDBY5G NB-IOT: SUSTAINABLE TECHNOLOGY
  2. 2. ✱ STANDARDIZING NARROWBAND 2 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 3 STANDARDIZING NARROWBAND ✱
  3. 3. ✱ STANDARDIZING NARROWBAND 4 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 5 CONTENTS ✱ 08 NB-IOT: A SUSTAINABLE TECHNOLOGY FOR CONNECTING BILLIONS OF DEVICES Under the umbrella of 3gpp, radio- access technologies for mobile broadband have evolved effectively to provide connectivity to billions of subscribers and things. Within this ecosystem, the standardization of a radio technology for massive mtc applications – narrowband iot (nb- iot) – is also evolving. The aim is to provide cost-effective connectivity to billions of iot devices, supporting low power consumption, the use of low-cost devices, and provision of excellent coverage – all rolled out as software on top of existing lte infrastructure. 18 THE CENTRAL OFFICE OF THE ICT ERA: AGILE, SMART AND AUTONOMOUS Enabled primarily by virtualization and sdn technologies, network architectures are becoming more flexible, with improved programmability and a greater degree of automated behavior. In combination with technology enablers such as the increased reach offered by fiber, automation of provisioning and orchestration, and improvements in the performance of generic hardware, network transformation has provided operators with the opportunity to rationalize and consolidate infrastructure. The next generation central office will introduce intelligence and service agility into the network through disaggregation. 30 FIVE TRENDS SHAPING INNOVATION IN ICT Tech companies often gain competitive advantage by causing market disruption through their ability to understand and act on technology trends. Like waves in the ocean, it’s much easier to ride these trends if you can see them coming and read them correctly. Our cto points to the five trends he expects to have the most impact on ict development in the year ahead. 42 PAVING THE WAY TO TELCO- GRADE PAAS The concepts of platform as a service (paas) and microservices – which have been gaining traction in the it world – are deeply rooted in the need to cut development times. The benefits are equally important in the telco domain, but there are gaps that need to be closed before paas is suitable for telco. Most of the challenges relate to the need for additional features that telco applications typically require. 52 4G/5G RAN ARCHITECTURE: HOW A SPLIT CAN MAKE THE DIFFERENCE In line with the evolution of 4g and the introduction of 5g, ran architecture is undergoing a transformation. The proposed future-proof software-configurable split architecture will be able to support new services, deployed on general-purpose and specialized hardware, with functions ideally placed to maximize scalability, spectrum, and energy efficiency – all while supporting the concept of network slicing. 66 CLOUD ROBOTICS: 5G PAVES THE WAY FOR MASS- MARKET AUTOMATION Robotics has shifted from the floor of the research lab to becoming a crucial cost-, time-, and energy- saving element of modern industry. By adding mobility to the mix, the possibilities to include system automation in almost any process in almost any industry increase dramatically. But there is a challenge. How do you build smart robotic systems that are affordable? The answer: cloud robotics enabled by 5g. 66 RCF RUHW SPP GPP GPP Antenna location RBS site Functions/ software configuration 1st level CO 2nd level CO RDC RF RF BPF PPF MME SGW PGW 52 42 Control Fabric Subscribers Apps TOR Fabri c f vP/S-GW Line cards 18 08 30
  4. 4. ✱ STANDARDIZING NARROWBAND 6 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 7 EDITORIAL ✱ and then. And so the need has arisen for a radio- access technology that can meet the connectivity requirements for massive mtc applications. This technology is narrowband iot (nb-iot), a solution designed to be deployed in the gsm spectrum, within an lte carrier, or in an lte or wcdma guard band. Robotics is one of the clear winners when it comes to the cross-fertilization of developments from different disciplines. The field of robotics brings materials, communication, and manufacturing together. The result: highly sophisticated production processes that are adapted to just-in-time methodologies, zero-waste policies, minimum use of raw materials, and low energy consumption. 5gis a key ingredient that will help to make the robotics industry mass market and affordable. By providing the connectivity that will support even the most demanding applications, 5g will enable system intelligence to be transferred to a cloud where computational capacity is greatest, and put simplified – more affordable – robots on the ground. And once people start to rely on robots – just as they depend on their smartphones – to carry out the practical tasks of daily life, we can expect this sector to boom. Security will continue to present new challenges. But as security issues continue to dominate headline news, developments are shifting from fire-fighter mode to prevention. You can read more of my thoughts on the shift in technology in the Tech Trends section of this issue of Ericsson Technology Review. Apart from security, my trends for the coming year include: the ability of the cloud to spread intelligence, self- managing devices, communication beyond sight and sound, and the influence of other sectors. If I was to suggest one takeaway from all of the articles included in this issue, I would say it is speed. Device processing is getting faster, data speeds are constantly increasing, radio speeds are approaching those of fiber, more people are becoming subscribers, more things are becoming connected, more applications are running constantly. Developers of new technologies are working hard to enhance responsiveness by reducing latency, a key performance parameter. The capability to determine what functions can be virtualized to maximize ideal placement in the network and ensure low latency is one of the primary driving factors behind a proposed split of radio-access architecture, which is detailed in the article 4g/5g ran architecture: how a split can make the difference. As always, I hope you find our stories relevant and inspiring. All of our content is available at www.ericsson.com/ericsson-technology-review, through the Ericsson Technology Insights app, and on SlideShare. ULF EWALDSSON SENIOR VICE PRESIDENT, GROUP CTO AND HEAD OF GROUP FUNCTION TECHNOLOGY ■ developments in technology have contributed to the launch of innumerable products and solutions designed to make our lives easier. A smartphone can help us find out where we are, conduct research, watch a favorite movie, make a video call to a friend, read a magazine article, find the nearest restaurant, book tickets, or send someone a picture – even while flying. Advancements in technology have helped to reunite refugees with their families, and combat terrorism. They have brought medical care into the living room, tearing down the obstacles and boundaries of traditional business models. The benefits brought about through research not only apply to telecoms, but to all modern industries. But the significant change that has resulted in the rapid deployment of innovation is the way industries collaborate today, with technology developments in one industry rapidly providing benefits to other market sectors. According to the June 2016 Ericsson Mobility Report, connected iot devices will outnumber mobile phones by 2018. This forecast reminds me of the milestone we witnessed at the end of 2009, when data traffic surpassed voice traffic in mobile networks. Since then, voice traffic has remained more or less constant, yet data traffic has continued to demonstrate strong growth. The smartphone revolution that followed brought with it a massive amount of network adaptation, so that networks built for voice could be transformed into data carriers. Today, the iot presents a similar need for change. With features like billing, in-app purchasing, and video streaming, most mobile-broadband networks have been designed to support the traffic generated by typical subscribers. But the iot, with its wide range of applications, needs customized connectivity: in other words, connectivity that suits each application in terms of cost, reach, bandwidth, and latency. The bottom line is cost. It simply doesn’t make economic sense to use broadband networks and valuable spectrum for applications that transmit just a few kbs of data now THE IOT, SPEED, AND DEEP COLLABORATION E R I C S S O N T E C H N O L O G Y R E V I E W Bringing you insights into some of the key emerging innovations that are shaping the future of ict. Our aim is to encourage an open discussion on the potential, practicalities, and benefits of a wide range of technical developments, and help provide an insight into what the future has to offer. a d d r e s s Ericsson se-164 83 Stockholm, Sweden Phone: +46 8 719 00 00 p u b l i s h i n g All material and articles are published on the Ericsson Technology Review website: www.ericsson.com/ ericsson-technology-review. Additionally, content can be accessed on the Ericsson Technology Insights app, which is available for Android and ios devices. The download links can be found on the Ericsson Technology Review website. p u b l i s h e r Ulf Ewaldsson e d i t o r Deirdre P. Doyle (Sitrus) deirdre.doyle@sitrus.com e d i t o r i a l b o a r d Aniruddho Basu, Joakim Cerwall, Stefan Dahlfort, Deirdre P. Doyle, Björn Ekelund, Dan Fahrman, Geoff Hollingworth, Jonas Högberg, Cenk Kirbas, Sara Kullman, Börje Lundwall, Hans Mickelsson, Ulf Olsson, Patrik Roseen, Robert Skog, Gunnar Thrysin, Tonny Uhlin, Javier Garcia Visiedo, and Erik Westerberg t e c h n o l o g y t r e n d s Ulf Ewaldsson and Kristina Gold a r t d i r e c t o r Kajsa Dahlberg (Sitrus) l ay o u t Jade Birke (Sitrus) i l l u s t r at i o n s Claes-Göran Andersson cg@cga.se c h i e f s u b e d i t o r Birgitte van den Muyzenberg (Sitrus) s u b e d i t o r s Paul Eade and Ian Nicholson (Sitrus) issn: 0014-0171 Volume: 93, 2016 TECHNOLOGY DEVELOPMENTS IN ONE INDUSTRY CAN RAPIDLY PROVIDE BENEFIT TO OTHER MARKET SECTORS.
  5. 5. 8 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 9 STANDARDIZING NARROWBAND IoT ✱✱ STANDARDIZING NARROWBAND IoT SARA LANDSTRÖM JOAKIM BERGSTRÖM ERIK WESTERBERG DAVID HAMMARWALL deployedintypicalscenarios,per-deviceand life-cyclecostsneedtobekepttoaminimum, andmeasuresthatpromotebatterylongevityare essentialforensuringtheoverallcost-effectiveness ofthesystem. Thecoverageandthroughputneedsformassive mtc applicationsarequitedifferentfromthoseof mbb.Theneedtosupporthighbitrates,forexample, appliestombb scenarios,butseldomtomassive mtc.Theprecisenatureofmassivemtc allowsfor asignificantdegreeofoptimizationinthedesignof radioaccess. Standardizationofnb-iot beganin2014with a3gpp study.Theobjectiveofthisstudywasto determinetherequirementsformassivemtc, tochooseanevaluationmethodology,andto investigatewhetherproposedradio-accessdesigns couldmeetthesetrequirements.Thisstudyled toworkonthespecificationofnb-iot [1],witha numberofdesigntargets–asillustrated inFigure1. Inadditiontothedesigntargets,extensive deploymentflexibilityanduseofindustry competencetomeettime-to-marketrequirements Under the umbrella of 3gpp, radio-access technologies for mobile broadband have evolved effectively to provide connectivity to billions of subscribers and things. Within this ecosystem, the standardization of a radio technology for massive mtc applications – narrowband iot (nb-iot) – is also evolving. The aim is for this technology to provide cost-effective connectivity to billions of iot devices, supporting low power consumption, the use of low-cost devices, and provision of excellent coverage – all rolled out as software on top of existing lte infrastructure. The design of nb-iot mimics that of lte, facilitating radio network evolution and efficient coexistence with mbb, reducing time to market, and reaping the benefits of standardization and economies of scale. t h e b e s t way to provide mtc applications with cost-effective connectivity is to design the radio-access network accordingly. What is needed is a radio-access network that minimizes battery usage, covers a wide area, and functions with simplified low-cost devices while efficiently matching the varying spectrum allocations of operators. 3gpp release 13 specifications includes the nb-iot feature, with a large degree of deployment flexibility to maximize migration possibilities and allow the technology to be deployed in gsm spectrum, in an lte carrier, or in a wcdma or lte guard band. ■Theiot embedsabroadrangeofmtc applications,andamongthedifferenttypes,itis widelyacceptedthatmassivemtc willbethefirst totakeoff.Thissegmentincludesapplications likesmartmetering,agricultureandrealestate monitoring,aswellasvarioustypesoftracking andfleetmanagement.Oftenreferredtoaslow powerwidearea(lpwa),networksthatprovide connectivitytomassivemtc applicationsrequirea radio-accesstechnologythatcandeliverwidespread coverage,capacity,andlowpowerconsumption. Massivemtc devicestypicallysendsmall amountsofdata,andtendtobeplacedinsignal- challengedlocationslikebasementsandremote ruralareas.Duetothesheernumbersofdevices Terms and abbreviations cs – circuit-switched | dl – downlink | drx – discontinuous reception | edrx – extended DRX | embms – evolved multimedia broadcast multicast service | emtc – enhanced machine-type communications | epc – Evolved Packet Core | e-utra – Evolved Universal Terrestrial Radio Access | iot – Internet of Things | lpwa – low power wide area | mac – media/medium access control | mbb – mobile broadband | mtc – machine-type communications | nb-iot – narrowband Internet of Things | ofdma – Orthogonal Frequency-Division Multiple Access | pa – power amplifier | prb – physical resource block | psm – power save mode | rf – radio frequency | rlc – Radio Link Control | rrc – Radio Resource Control | sc-fdma – single-carrier frequency-division multiple access | tco – total cost of ownership | ue – user equipment | ul – uplink Low device cost: under USD 5 per module Long battery life: more than 10 years Capacity: 40 devices per household Extended coverage: 20dB better than GPRS Report uplink latency: less than 10 seconds Figure 1 iot design targets A SUSTAINABLE TECHNOLOGY FOR CONNECTING BILLIONS OF DEVICES NB-IOT:
  6. 6. 10 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 11 ✱ STANDARDIZING NARROWBAND IoT STANDARDIZING NARROWBAND IoT ✱ havebeenincludedaskeyconsiderationsin thespecificationofnb-iot.Tofuture-proofthe technology,itsdesignexploitssynergieswithlte byreusingthehigherlayers(rlc,mac,andrrc), forexample,andbyaligningnumerology(the foundationofthephysicallayer)inboththeuplink anddownlink.However,theaccessproceduresand controlchannelsfornb-iot arenew. Priortonb-iot specification,workhadalready begunonthedesignofanotherradioaccess formassivemtc tosupportCat-m1 – anew ue category.Withcompletionalsotargetedfor release 13,theresultingstandardizationworkitem –emtc –coversbitrates,forexample,rangingfrom hundredsofkbpsto1mbps.Theserequirementsare broaderthannb-iot which hasbeenstreamlined forapplicationswithwidelyvaryingdeployment characteristics,lowerdatarates,andoperationwith simplifiedlow-costdevices. Withacarrierbandwidthofjust200kHz(the equivalentofagsm carrier),annb-iot carriercan bedeployedwithinanlte carrier,orinanlte or wcdma guardband*.Thelinkbudgetofnb-iot hasa20db improvementoverlte Advanced.Inthe uplink,thespecificationofnb-iot allowsformany devicestosendsmallamountsofdatainparallel. Release13notonlyincludesstandardsfor emtcandnb-iot,italsocontainsimportant refinements,suchasextendeddiscontinuous reception(edrx)andpowersavemode(psm). psm wascompletedinrelease12toensurebattery longevity,andiscomplementedbyedrx foruse casesinvolvingdevicesthatneedtoreceivedata morefrequently. Deploymentflexibilityand migrationscenarios Asafiniteandscarcenaturalresource,spectrum needstobeusedasefficientlyaspossible.And sotechnologiesthatusespectrumtendtobe designedtominimizeusage.Toachievespectrum efficiency,nb-iot hasbeendesignedwithanumber ofdeploymentoptionsforgsm,wcdma,orlte spectrum,whichareillustratedinFigure2. 〉〉 standalone–replacingagsm carrierwithannb-iot carrier 〉〉 in-band–throughflexibleuseofpartofanlte carrier 〉〉 guardband–eitherinwcdma orlte Starting with standalone Thestandalonedeploymentisagoodoptionfor wcdma orlte networksrunninginparallelwith gsm.Bysteeringsomegsm/gprs traffictothe wcdma orlte network,oneormoreofthegsm carrierscanbeusedtocarryiot traffic.Asgsm operatesmainlyinthe900mhz and1,800mhz bands(spectrumthatispresentinallmarkets),this approachacceleratestimetomarket,andmaximizes thebenefitsofaglobal-scaleinfrastructure. Migrationtoin-band Whenthetimingisright,gsm spectrumwill berefarmedforusebymoredemandingmbb traffic.Refarmingspectrumforusebylte isa straightforwardprocess,evenwhennb-iot carriers existinthegsm spectrumbecauserefarmingdoes notimpactnb-iot devices,andanynb-iot carriers ingsm willcontinuetooperatewithinthelte carrieraftermigration.Suchafuture-proofsetupis possible,asthestandaloneandin-bandmodesuse thesamenumerologyaslte,andrf requirements aresettomatchthedifferentdeployments,soall devicesareguaranteedtosupportin-bandoperation atthetimeofmigration. In-band:bestoptionforlte Foroperatorswithmainlylte spectrumavailable, thelte in-bandoptionprovidesthemostspectrum- andcost-efficientdeploymentofnb-iot.Morethan anythingelse,thisparticularoptionsetsnb-iot apartfromanyotherlpwa technology. Annb-iot carrierisaself-containednetwork elementthatusesasinglephysicalresourceblock (prb).Forin-banddeploymentswithnoiottraffic present,theprb canbeusedbylte forother purposes,astheinfrastructureandspectrumusageof lte andnb-iot arefullyintegrated.Thebasestation schedulermultiplexesnb-iot andlte trafficonto thesamespectrum,whichminimizesthetotalcost ofoperationformtc,whichessentiallyscaleswith thevolumeofmtc traffic.Intermsofcapacity,the capabilityofasinglenb-iot carrierisquitesignificant –evaluationshaveshownthatastandarddeployment cansupportadeploymentdensityof200,000 nb-iot deviceswithinacell–foranactivitylevel correspondingtocommonusecases.Naturally,more nb-iotcarrierscanbeaddedifmorecapacityisneeded. Usingguardbandspectrum Athirdalternativeistodeploynb-iot inaguard band,andhere,thefocusisontheuseofsuch bandsinlte.Tooperateinaguardbandwithout causinginterference,nb-iot andlte needto coexist.Incontrasttootherlpwa technologies,the physicalnb-iot layershavebeendesignedwiththe requirementsofin-lte-guard-bandcoexistence specificallytakenintoconsideration.Again,likelte, nb-iot usesofdma inthedownlinkandsc-fdma intheuplink. Thedesignofnb-iot hasfullyadoptedlte numerology,using15khz subcarriersintheuplink anddownlink,withanadditionaloptionfor3.75khz subcarriersintheuplinktoprovidecapacityin signal-strength-limitedscenarios. Longrangeandlongbatterylife Thegeographicalareaforwhichamobilenetwork canprovidecoveragedependsonsitedensityand linkbudget.Comparedwithgprs,wcdma andlte, thelinkbudgetofnb-iot hasa20db margin,anduse casestendtooperatewithlowerdatarates. So,notonlycannb-iot reusethegsm,wcdma, orlte grid,theimprovedlinkbudgetenablesitto reachiot devicesinsignal-challengedlocations suchasbasements,tunnels,andremoteruralareas –placesthatcannotbereachedusingthenetwork’s voiceandmbb services. Intechnicalterms,thecoveragetargetofnb-iot hasalinkbudgetof164db,whereasthecurrent gprs linkbudgetis144db (tr 45.820[2]),andlte is142.7db**(tr 36.888[3]).The20db improvement correspondstoasevenfoldincreaseincoveragearea foranopenenvironment,orroughlythelossthat occurswhenasignalpenetratestheouterwallofa building.Standardizationactivitiesin3gpp have shownthatnb-iot meetsthelinkbudgettarget of164db,whilesimultaneouslymeetingthemtc applicationrequirementsfordatarate,latency,and batterylife. Thebatterylifeofanmtc devicedependsto someextentonthetechnologyusedinthephysical layerfortransmittingandreceivingdata.However, longevitydependstoagreaterextentonhow efficientlyadevicecanutilizevariousidleandsleep GSM LTE LTE LTE Standalone 200kHz 200kHz 200kHz In-band Guard band Figure 2 Spectrum usage deployment options *Guard band is a thin band of spectrum between radio bands that is used to prevent interference. ** The noise figure assumptions in3gpp ts 36.888 [3] used in the link budget calculations are more conservative than in the corresponding link budget for gsm in 3gpp tr 45.820. Using the noise figure assumptions from tr 45.820, the lte link budget becomes 142.7db.
  7. 7. 12 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 13 STANDARDIZING NARROWBAND IoT ✱✱ STANDARDIZING NARROWBAND IoT modesthatallowlargepartsofthedevicetobe powereddownforextendedperiods.Thenb-iot specificationaddressesboththephysical-layer technologyandidlingaspectsofthesystem. Likelte,nb-iot usestwomainrrc protocol states:rrc_idleandrrc_connected.In­­rrc_idle, devicessavepower,andresourcesthatwouldbeused tosendmeasurementreportsanduplinkreference signalsarefreedup.Inrrc_connected,devicescan receiveorsenddatadirectly. Discontinuousreception(drx)istheprocess throughwhichnetworksanddevicesnegotiatewhen devicescansleepandcanbeappliedinbothrrc_idle andrrc_connected.Forrrc_connected, theapplicationofdrxreducesthenumberof measurementreportsdevicessendandthenumber oftimesdownlinkcontrolchannelsaremonitored, leadingtobatterysavings. 3gpprelease12supportsamaximumdrxcycleof 2.56seconds,whichwillbeextendedto10.24seconds inrelease13(edrx).However,anyfurtherlengthening ofthisperiodisasyetnotfeasible,asitwouldnegatively impactanumberofranfunctionsincludingmobility andaccuracyofthesysteminformation.Inrrc_idle, devicestrackareaupdatesandlistentopaging messages.Tosetupaconnectionwithanidledevice, thenetworkpagesit.Powerconsumptionismuchlower foridledevicesthanforconnectedones,aslistening forpagesdoesnotneedtobeperformedasoftenas monitoringthedownlinkcontrolchannel. Whenpsm wasintroducedinrelease12,it enableddevicesinrrc_idletoenteradeepsleep inwhichpagesarenotlistenedfor,noraremobility- relatedmeasurementsperformed.Devicesinpsm performtrackingareaupdatesafterwhichthey directlylistenforpagesbeforesleepingagain. psm andedrx complement each other and can support battery lifetimes in excess of 10 years for different reachability requirements, transmission frequencies of different applications, and mobility. The range of solutions designed to extend battery lifetimes need to be balanced against requirements for reachability, transmission frequency of different applications, and mobility. These relations are illustrated in Figure 3. Superior capacity design Tomeetcapacityrequirements,nb-iot needsto multiplexmanydevicessimultaneously,andprovide connectivityinanefficientmannerforallofthem irrespectiveofcoveragequality.Asaresult,the designofnb-iot supportsarangeofdatarates. Theachievabledataratedependsonthechannel quality(signaltonoiseratio),andthequantityof allocatedresources(bandwidth).Inthedownlink, alldevicessharethesamepowerbudgetand severalmaysimultaneouslyreceivebase-station transmissions.Intheuplink,however,eachdevice hasitsownpowerbudget,andthiscanbeusedto advantagebymultiplexingthetrafficgeneratedby severaldevices,astheircombinedpowerisgreater thanthatofasingledevice. Inmanylocations,nb-iot deviceswillbe limitedbysignalstrengthratherthantransmission bandwidth.Suchdevicescanconcentratetheir transmissionenergytoanarrowerbandwidth withoutlossofperformance,whichfreesupband- widthforothers.Thepossibilityofallocatingsmall amountsofbandwidthtospecificdevicesincreases systemcapacitywithoutlossofperformance. Toenablesuchsmallbandwidthallocations, nb-iot usestonesorsubcarriersinsteadofresource blocks.Thesubcarrierbandwidthfornb-iot is 15khz,comparedwitharesourceblock,whichhas aneffectivebandwidthof180khz.Eachdeviceis scheduledononeormoresubcarriersintheuplink, anddevicescanbepackedevenclosertogetherby decreasingthesubcarrierspacingto3.75khz.Doing so,however,resultsindifferingnumerologyforlte andnb-iot,andsomeresourceswillneedtobe allocatedtoavoidinterferencebetweenthe3.75khz and15khz subcarriersinsteadofutilizingthemfor traffic,whichmayleadtoperformancelosses. Forscenariosthatincludedevicesinbothgood andbadcoverageareas,itispossibletoincrease thedataratebyaddingmorebandwidth.Inthe uplink,dataratescanbeincreasedupto12times byallocatingdeviceswithamulti-toneormulti- subcarrierratherthanasingletone,forexample. Thisapproachimprovescapacityforscenarios wheremanydeviceshavegoodcoverage,asdata transfercompletesquickly.Goodcoverageistypical 10s 30s 1m 3m 5m 10m 30m 1h 3h 6h 12h 24h 2 days 5 days 3 days 5 days 24h 12h 6h 3h 1h 30m 10m 3m 1m Data inter-arrival time Reachability interval PSM eDRX in RRC_Idle eDRX in RRC_Connected Figure 3 Good coverage
  8. 8. 14 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 15 STANDARDIZING NARROWBAND IoT ✱✱ STANDARDIZING NARROWBAND IoT whennb-iot isrolledoutonadensegridand/or whenmostdevicesarewithintheoriginallte cell coveragearea. Datarateisasignificantfactorwhentryingto achievethebestdesignfornb-iot,asitaffectsboth latencyandpowerconsumption.Table1shows theuplinklatencyvaluesforadevicetoconnect andtransmitdata.Thedataratesforworst-case coverage(+20db)arelowerthanthoseformbb atthecelledge(0db),andlatencyincreasesfrom 1.6to7.6seconds.Theuplinkdatarateisthemain causeofthisdegradation,yetevenforworst-case scenarios,nb-iot uplinklatencyisstillunderthe 10-seconddesigntarget.Whenitcomestopower consumption,thedominatingfactoristhespeedat whichdevicestransmitdata,whichincreasesinline withacceleratingdatarates. nb-iot hasbeendesignedwithgoodmultiplexing andadaptabledataratesandsoitwillbeableto meetpredictedcapacityrequirements.Thecapacity requirementstargetin3gpp tr 45.820[1]hasbeen setto40devicesperhousehold,based onassumptionsforLondon,whichcorrespond to52,500devicespercell.Simulationsshow supportfor200,000devicespercell–fourtimes thesettarget. Deviceaspects Affordablemodemsareakeyelementoflarge- scalesensordeployment,sothatprocesses suchastemperatureorwatermeterreporting canbeoptimized.Atthesametime,thedata rateandlatencyrequirementsofsuchsensor- heavyapplicationstendtoberelativelymodest: acharacteristicthatcanbeusedtoadvantageto reducesolutioncomplexity–andcost. nb-iot devicessupportreducedpeakphysical layerdatarates:intherangeof100-200kbpsor significantlylowerforsingle-tonedevices.To facilitatelow-complexitydecodingindevices,turbo codesarereplacedwithconvolutionalcodesfor downlinktransmissions,andlimitsareplacedon maximumtransportblocksize–whichis680bitsfor dl andnotgreaterthan1000bitsforul. Theperformancerequirementssetfornb-iot makeitpossibletoemployasinglereceiverantenna (twoareneededforlte mbb).Asaresult,theradio andbasebanddemodulatorpartsofthedeviceneed onlyasinglereceiverchain.Byoperatingnb-iot deviceshalfduplexsothattheycannotbescheduled tosendandreceivedatasimultaneously,theduplex filterinthedevicecanbereplacedbyasimple switch,andaonlysinglelocaloscillatorforfrequency generationisrequired.Theseoptimizationsreduce costandpowerconsumption. At200khz,thebandwidthofnb-iot is substantiallynarrowerthanotheraccess technologies.lte bandwidths,forexample,range from1.4mhzto20mhz.Thebenefitofanarrowband technologyliesinthereducedcomplexityof analog-to-digital(a/d)anddigital-to-analog(d/a) conversion,buffering,andchannelestimation–allof whichbringbenefitsintermsofpowerconsumption. nb-iot bringsaboutasignificantdesignchange intermsoftheplacementofthedevice'spower amplifier(pa).Integratingthiselementdirectlyonto thechip,insteadofitbeinganexternalcomponent, enablessingle-chipmodemimplementations– whicharecheaper. Reuseofexistingtechnology Thedesignofnb-iot radioaccessreusesanumber oflte designprinciplesandhasthebackingofthe traditionalcellular-networkandchipsetvendors thatmadembb asuccess.nb-iot employsthesame designprinciplesaslte (e-utra),althoughituses aseparatenewcarrier,newchannels,andrandom accessprocedurestomeetthetarget requirements ofiot usecases–suchasimprovedcoverage,lower batteryconsumptionandoperationinnarrow spectrum.Constructingnb-iot inthiswaytakes advantageoflte’swell-establishedglobalreach, economiesofscale,andindustry-leadingecosystem. Thenb-iot downlinkisbasedonofdma and maintainsthesamesubcarrierspacing,ofdm symbolduration,slotformat,slotduration,and subframedurationaslte.Asaresult,nb-iot can providebothin-bandandguardbanddeployment withoutcausinginterferencebetweenitscarriers andthoseusedbylte formbb,makingnb-iot awellintegratediot solutionforlte-focused operatorsinadditiontocat-m1. Useofthesameupperlayersisyetanother similaritybetweenlte andnb-iot,withsome optimizationstosupportoperationwithlow-cost devices.Forexample,asasingletechnologysolution, nb-iot doesnotsupportdualconnectivity;and devicesdonotsupportswitchingbetweenaccess technologies(gsm,wcdma,orwi-fi)inactive mode.Supportforcs voiceserviceshasalsobeen removed.Thesescopesavingsresultinamuchlower requirementformemorycapacityfornb-iot devices comparedwitheventhemostrudimentarymbb lte ones. nb-iot usesans1-basedconnectionbetween theradionetworkandtheepc.Theconnectionto theepc providesnb-iot deviceswithsupportfor roamingandflexiblecharging,meaningthatdevices canbeinstalledanywhereandcanfunctionglobally. Theambitionistoenablecertainclassesofdevices –likesmokedetectors–tobehandledwithpriority toensurethatemergency-situationdatacanbe prioritizedifthenetworkiscongested. Existing3gpp architectureprovidesaglobal, highlyautomatedconnectivitymanagementsolution thatisneededforlarge-scaleiot deployments. nb-iot andlte usethesameo&m framework, runningasasinglenetworkcarryingmbb andmtc traffic,whichreducesoperationalcostsinareas likeprovisioning,monitoring,billing,anddevice management.Similartopresentlte networks, nb-iot supportsstate-of-the-art3gpp security, withauthentication,signalingprotection,and dataencryption. lte featuresthatalreadyexist,likecell-id-based positioning,arestraightforwardenoughfor nb-iot toinherit.Byaligningwithlte evolution, nb-iot couldsupportexistingfeaturesandfuture functionalitydesignedfortheentirecellular ecosystem,includingmbb aswellasiot usecases. Table 1 Maximum uplink latency for a device on the mbb cell border (+0db) and beyond (+ 10db and + 20db) Duration (ms) Coverage Sync MIB PRACH RAmsg2-4 ULgrant ULdata Ack ULdata TOTAL 340 151 324 622 48 39 41 39 1,604 340 151 688 708 45 553 47 553 3,085 520 631 1,440 1,060 49 1,923 77 1,923 7,623 +0dB +10dB +20dB DATA RATE IS A SIGNIFICANT FACTOR WHEN TRYING TO ACHIEVE THE BEST DESIGN FOR NB-IOT, AS IT AFFECTS BOTH LATENCY AND POWER CONSUMPTION nb-iot: the advantages of being part of 3gpp 〉〉useofthelte ecosystem,leadingtofast development,economiesofscale,andglobal roaming 〉〉 canbedeployedasasimpleadditionofnewsoftware toexistinglte infrastructure 〉〉amanagementframeworkexists,enablinglarge- scaledeployments 〉〉frameworkincludesstate-of-the-artsecurity 〉〉futurefeaturegrowthformbb andnb-iot usecases
  9. 9. 16 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 17 STANDARDIZING NARROWBAND IoT ✱✱ STANDARDIZING NARROWBAND IoT Sara Landström ◆ is a strategic product manager in the area of 4gand 5g at Ericsson. She is currently responsible for the iot, v2x, and carrier aggregation radio portfolios. She joined Ericsson in 2008 as a researcher focusing on radio resource management, heterogeneous networks, and radio access for iot. Since then she has been project manager for various proprietary feature development projects and has headed up the Radio Network Algorithms research group. She holds an m.sc. in computer engineering and a ph.d. in computer networking, both from Luleå University of Technology, Sweden. Joakim Bergström ◆ is an expert in new radio- access networks at Design Unit Radio. He has more than 15 years of experience in standardization within the 3gpp ran area working with hspa, lte and 5g. He holds an m.sc. in electrical engineering from kth Royal Institute of Technology, Stockholm. Within the radio area, he has coordinated all of Ericsson’s standardization activities and projects since 2011. Erik Westerberg ◆ joined Ericsson from mit, Massachusetts, us, in 1996 and is a senior expert in system and network architecture. During his first 10 years at Ericsson, he worked with development of the mobile broadband systems before broadening his work to include the full network architecture as he served as Chief Network Architect until 2014. He holds a ph.d. in quantum physics from Stockholm University, Sweden. David Hammarwall ◆ is head of Services and Infrastructure within Product Area 4g/5g ran. A main driver of Ericsson’s strategy and execution within the Cellular Internet of Things, Hammarwall joined Ericsson’s lte product management team in 2013, with primary responsibilities for lte baseband capacity, software architecture, and features developed in device partnerships. He received his ph.d. in telecommunications from kth Royal Institute of Technology in Stockholm in 2007 before joining Ericsson Research to focus primarily on 3gpp standardization. He has acted as a primary standardization delegate in 3gpp, leading Ericsson’s standardization efforts and strategy within multi- antenna technologies, Coordinated Multipoint, and small cell enhancements. Thebroadcastfeatureembms enablesalarge numberofdevicestobeupdatedsimultaneously, andthedevice-to-devicecommunicationfeature thatrelaystransmissionstodevicesinpoorcoverage areexamplesofsynergies.Inthefuture,thesetwo featurescanbespecifiedfornb-iot usingthesame conceptsandexperiencethatwereusedtodevelop themforlte mbb. Conclusions nb-iot isthe3gpp radio-accesstechnology designedtomeettheconnectivityrequirementsfor massivemtc applications.Incontrasttoothermtc standards,nb-iot enjoysallthebenefitsoflicensed spectrum,thefeaturerichnessofepc,andthe overallecosystemspreadof3gpp.Atthesametime, nb-iot hasbeendesignedtomeetthechallenging tco structureoftheiot market,intermsofdevice andran cost,whichscaleswithtransferred datavolumes. Thespecificationfornb-iot ispartof3gpp release 13anditincludesanumberofdesigntargets: devicecostunderusd 5permodule;acoverage areathatisseventimesgreaterthanexisting3gpp technologies;devicebatterylifethatislongerthan 10yearswithsustainedreachability;andmeeta capacitydensityof40devicesperhousehold. Asnb-iot canbedeployedingsm spectrum, withinanlte carrier,orinanlte orwcdma guardband,itprovidesexcellentdeployment flexibilityrelatedtospectrumallocation,whichin turnfacilitatesmigration.Operationinlicensed spectrumensuresthatcapacityandcoverage performancetargetscanbeguaranteedforthe lifetimeofadevice,incontrasttotechnologies thatuseunlicensedspectrum,whichruntherisk ofuncontrolledinterferenceemergingevenyears afterdeployment,potentiallyknockingoutlarge populationsofmtc devices. Thefirststandarddevelopmentof5g radioaccess iscurrentlyunderway,withsystemdeployment targetedfor2020.Inthiscontext,theabilityto future-proofadditionaltechnologieslikenb-iot is atoppriority.Intheongoingdiscussionsin3gpp surrounding5g,lte willcontinuetobeanintegral partofradionetworksbeyond2020,andso,nb-iot's resemblancetolte safeguardsthetechnologyfrom divergingevolutionpaths. References 1. 3gpp, December 2015, NB-IoT work item description RP-152284, available at: http://ow.ly/4mQAfx 2. 3gpp, tr 45.820, Cellular system support for ultra-low complexity and low throughput Internet of Things (clot), available at: http://ow.ly/4mQAny 3. 3gpp, tr 36.888, Study on provision of low-cost Machine-Type Communications (mtc) User Equipments (ues) based on lte (v12.0.0), available at: http://ow.ly/4mQAwn theauthors
  10. 10. 18 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 19 SECOND WAVE CONSOLIDATION ✱✱ SECOND WAVE CONSOLIDATION NAIL KAVAK ANDREW WILKINSON JOHN LARKINS SUNIL PATIL BOB FRAZIER Thefirstwaveofco consolidationand centralizationcameaboutduringthedigitization ofpots.Digitizationresultedinareductionin sizeorfunctionalityofmanycityandruralcos, andinmanyplaces,theywerereplacedwithsmall concentratorsconnectedtoasmallernumberof morecentralizedcos. Thelocationofaco issignificant;when positionedincloseproximitytousers,certain servicescanbeprovidedtolocalgroupsof subscribersinahighlyefficientmanner.This capabilityisoneoftheprimarydifferentiatingassets oftheaccessoperator. Likewiseformobilenetworks,theoptimal placementofanmso takeslocationconstraintsinto consideration,andthecriticalfactorformobileis accesstothebasetransceiverstation(bts).Originally, cablewasusedastheprimarymediaforbts access, shiftinginrecentyearstouseofhigh-capacityfiber tdm circuits.So,forthesamegeographicareaand subscriberbase,msostendtobemorecentralized comparedwithfixedcos. Today,amedium-sizedcitycanbeservedbyjust oneortwomsos,butpossiblyhundredsofcos.In ruralareas,however,msostendtobesparseoreven nonexistent.Operatorsrunningconvergedfixedand mobilenetworkstendtohousemsoswithinexisting cos,rarelyoptingfornewbuildsin dedicatedlocations. Figure1illustratesthelocal,regional,and nationallytieredstructureofcos.Fixedcos have twoormoreprogressivelycentralizedtiers,which originallyprovidedinter-officecallingcapability toavoidtheneedforafullcomesh.Higher-tier cos haveextensivetransmissiontrunkingfrom lower-tierandaccesscos,whichissignificant,as thisarchitecturemaybeutilizedfortheplacement ofnextgenerationcentraloffices.msos canbe colocatedwithasubsetofcosorbedeployed independentlyaslocalandregionalcos.Endsites Network architecture is undergoing a massive transformation, which in turn is having an impact on the role of the central office. Enabled primarily by virtualization and SDN technologies, network architectures are becoming more flexible, with improved programmability and a greater degree of automated behavior. In combination with technology enablers such as the increased reach offered by fiber, automation of provisioning and orchestration, and improvements in the performance of generic hardware, network transformation has provided operators with the opportunity to rationalize and consolidate infrastructure. The next generation central office will introduce intelligence and service agility into the network through disaggregation. t h e c e n t r a l o f f i c e s (cos) of fixed networks and the mobile switch offices (msos) of mobile operators house the networking functionality, management, and compute power needed to provide voice and data services to enterprise and residential subscribers. To route traffic efficiently, cos are distributed throughout the entire geographic region served by the network, and provide operators with a key asset: local proximity to their subscribers. ■ Traditionally,thelocationofafixed-lineco has beendeterminedbythereachconstraintsofthe accesstechnologiesusedinthelastmile–from theco tothesubscriber(residentialorenterprise). Untilrecently,copperwasthepredominantmedia, andsothelocationoftheco hasbeendictatedby themaximumreachofthecopperpairssupporting pots equipmentinthehomeorattheenterprise premises.Althoughcopperisnolongertheprimary choiceforaccessmedia(orevenpresentinmany cases),thelocationofcosstillreflectstheoriginal distanceconstraints.Asaresult,evenmid-sized citieswitharoundamillionsubscribersareserved byhundredsofcos,anditisstillcommonfor thesetobeplacedinagrid-likemanner,spaceda coupleofkilometersapart.Inruralareaswithlow populationdensity,fixed-accesstechnologyreach isalsothemainfactorfordetermininglocation, explainingwhytheratioofsubscriberstocosin ruralareastendstobelow. THE LOCATION OF A CO IS SIGNIFICANT; WHEN POSITIONED IN CLOSE PROXIMITY TO USERS, CERTAIN SERVICES CAN BE PROVIDED TO LOCAL GROUPS OF SUBSCRIBERS IN A HIGHLY EFFICIENT MANNER Termsandabbreviations acl — access control list | api — application programming interface | arpu — average revenue per user | bgp — Border Gateway Protocol | bng — Broadband Network Gateway | bsc — base station controller | bss — business support systems | bts — base transceiver station | cdn — content delivery network | cios – Non-blocking, multistage switch fabric formalized by Charles Clos | cms – cloud management system | co — central office | cots — commercial off-the-shelf | cpu — central processing unit | docsis — Data Over Cable Service Interface Specification | dsl — digital subscriber line | gpon — gigabit passive optical network | hlr — home location register | i/o — input/output | igp — Interior Gateway Protocol | iot — Internet of Things | isp — internet service provider | m2m — machine-to-machine | mac — media/medium access control | mme — Mobility Management Entity | mpls — multi-protocol label switching | mso — mobile switch office | netconf — protocol to install, manipulate, and delete the configuration of network devices | nfv — Network Functions Virtualization | ngco — next generation central office | nic — network interface card | nms — network management system | nvgre — Network Virtualization using Generic Routing Encapsulation | odl — OpenDaylight | olt — Optical Line Termination | onie — Open Network Install Environment | oss — operations support systems | otn — optical transport network | ott — over-the-top | pon — passive optical network | pots — plain old telephone service | p-gw — packet data network gateway | p/s-gw — packet data network/serving gateway | rnc — radio network controller | sdn — software- defined networking | sfp — small form-factor pluggable | s/ggsn — serving/gateway gprs support node | vbng — virtual Broadband Network Gateway | vim — virtual infrastructure manager | vod — video on demand | vswitch — virtual switch | vxlan — Virtual Extensible LAN | xaas — anything as a service | xmpp — Extensible Messaging and Presence Protocol autonomous THE CENTRAL OFFICE OF THE ICT ERA: AGILE, SMART, AND
  11. 11. 20 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 21 SECOND WAVE CONSOLIDATION ✱✱ SECOND WAVE CONSOLIDATION Figure 1 co tiers and distribution LS LS Ag Ag BB BB BB BB BB BBBB BB BB BB BB RA Ac RA Ac Ag Ac Ac DCCO DR DR CO CO CO DC CO NoC CO CO DR DR DR DC DC CO DR Local Local Local Local Local Local Local Regional Regional Regional National connecttocos andmsos throughintermediate transportaggregationsites. Thecoshousepotsanddsl/pon access equipment,andtoalesserextent,ip/Ethernet routingandswitchingcapabilitiesforresidentialand enterpriseservices.msoshouseradioaggregation nodes,suchasbscsandrncs,aswellastransport switches.Someofthemsosmayhouseadditional 3gpp corefunctionssuchass/ggsn andp/s-gw, aswellascontrol-planefunctionssuchasmme and hlr servingmultiplegeographicareas.Othercore functionscan,however,beplacedelsewhere,for exampleinpurpose-builtregionalornationaldcs. Inthissecondwaveofco consolidation,the fundamentalinternalstructureandfunctionality providedateachsitewillchange,anduseofnew technologieswilleitherresultinfewersitesorgreater capacity.Thetermnextgenerationcentraloffice (ngco)hasbeenadoptedbythetelecomindustry torefertothefuturecentralofficesthatwillsupport bothfixedandmobileoperations.Comparedwith itscurrentco counterpart,thengco willbeableto servemoresubscribers,implementaccessfunctions inamoreit-centricway,andsupportandlocally housenew,flexibledataservices.Thengco will functionlikeahighlyautomatedminidatacenter, requiringlessspace,power,andcoolingthantheset oftraditionalcositreplaces. Whytransform? Inadditiontotheconstantneedtoreduceopexand capex,fixedandmobileoperatorscontinuallyface newchallengesastechnologyanduserdemands change.Networktransformationandchanging subscribertrafficpatternshavecreatednew challengesintermsoftheservicesoperatorsoffer, andperhapsmoresignificantly,theservicesthat operatorswouldliketooffer,andhowtoprovide themintheshifttowardthemoreattractiveanything- as-a-service(xaas)businessmodel. Theshiftfromvoicetodataservicesandthe correspondingmassiveincreaseinott traffichave putpressureonnetworks.Changesinuserbehavior, withpreferencesshiftingtouseofbandwidth-hungry dataservices,andvideoconsumptionrequirea revolutionarychangeinthewayexistingco-and mso-basednetworkarchitecturesarestructured. Trafficpatternsanddemands Theannualgrowthrateoftrafficcarriedbymobile andfixednetworkshasrisenmassivelyoverthepast fiveyears.Inadditiontoincreasingtrafficvolumes, meetingtheevermorestringentdemandsplaced onnetworkperformancecharacteristics,suchas latency,isnecessarytosupportemergingindustry applications.Technologyimprovementsmadein fixed-networkaccessandthemobileindustry(as 5g systemsevolve)willenablenetworkstocope withgrowingtrafficvolumesandperformance demands.But,asnetworkcapabilitiesincrease,user expectationsandthedemandformorecapacityand bandwidthwillalsoinevitablyrise. Theincreaseintrafficvolumesandperformance demandscanbepredictedandplannedfor,but changingtrafficpatternsduetochangingsubscriber habitsiscomplicatingnetworkarchitectureinanew way.Asnetworksbecomemoreflexible,user-to- userandmachine-to-machineflowswillbecome morewidespread,addingnewdimensionstothe traditionaluser-to-servertraffic-flowpattern. FactorinthemassiveexpansionoftheInternetof Things(iot)andtheresultwillbeanexplosioninthe numberofflowsandroutesthatnetworkswillneed tosupport. Withstaticordecliningarpu,thequestionfacing manyoperatorsishowtoinvestinnetworkssothey meetconstantlyrisingperformancedemands. Technologyprovidessomeusefulstepsthat canhelpanswerthisquestion.Forexample, wherepossibleandnecessarytomeetlatency requirementsorlowerbackhaulcosts,self-served andpartnercontent,suchasvideo,andsubscriber- associatedip servicedeliverypoints–p-gws, bngs,andmulti-serviceedgerouters–canbe movedclosertotheuser.Trafficnotservedbythe accessoperatorcanbeoffloadedtootherisps, transitcarriers,orott contentprovidersthatare closertotheaccessdomain,ratherthanhaulingit backtomorecentralizedinterconnectionpoints. Similarly,insteadofhubbingenterprisetransport trafficthroughlargecentralizedroutingpoints, amoreoptimalwaytoroutethistypeoftrafficis throughdistributedroutingpointsinthenetwork. Shiftingtrafficaroundlikethiswilldramatically altertheratiooflocallyterminatedtraffictotransit trafficandrequiresthengco toprovidesupportfor routingandservicefunctionalitywellbeyondthe capabilitiesofthetraditionalco. Efficientrolloutofservices Totakeadvantageoftherevenuestreamscreated bymassivetrafficvolumes,toughperformance targetsandnewtrafficpatterns,networksneed tobeabletosupportefficientrolloutofservices. Networkflexibilityiskeyhere,enablingoperators –andindirectlysubscribers–tomodifyservicesto matchtheirevolvingneeds,scalethemeasily,and beabletospecifyandchangethelocationofservice instantiation.Provisioningmechanismsneedtobe highlyefficient,lowopexandcapexareessential, and,astimetomarketiscrucial,highfeaturevelocity isvital. Accessoperatorsofferendservicessuchasweb applications,cdnswiththeirassociatedcontent caches,andbump-in-the-wireservicesincluding parentalcontrolfiltering,aswellastransportservices suchasenterpriseconnectivityorinternetaccess, oracombinationofboth.Moreadvancedservices requiresupportforservicechainingthatcanbe dynamicallycustomizedonaper-subscriberbasis. Publicandprivatecloud-basedxaas isan attractiveofferingforbothenterpriseandnon- enterprisecustomers,butrequiressupportfor multi-tenancyenvironments. Legend: DR: distributed radio; CO: central office; RA: remote access; Ac: access; Ag: aggregation; DC: data center BB: backbone; LS: local switching
  12. 12. 22 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 23 SECOND WAVE CONSOLIDATION ✱✱ SECOND WAVE CONSOLIDATION Byappropriatelylocatingtheseservicesinngcos, carriernetworkswillbecomepartofadistributed andintelligentcloudresource,supplementinglarger, centralizeddatacenters. Softwaredevelopmentanddeployment lifecycle Atypicalservicelifecyclestartswithdevelopment andverificationbeforemovingontowide-scale deploymentinthenetwork. Efficientservicelifecycledependsontwokey factors:shorttimetomarketanddeployment flexibility.Timetomarketcanbeminimizedthrough ahomogeneoussoftwareenvironmentthatenables deploymentonexistingnetworkinfrastructure withouttheneedforhardwaremodification. Deploymentflexibilityisneededtoenableelastic capacityscaling,dynamicservicechaining,andthe deploymentofservicesinnewlocations. Keytoimplementingthesefactorsinthengco isvirtualizationofthecomputeplatformonwhich servicesrun,sothatthetraditionalcouplingof softwaretospecifichardwarecanberemoved. Decouplingprovidesahomogeneousdevelopment anddeploymentenvironmentthatissuitedtoan automatedlifecycle. Technologicalenablers Fiberreach Theincreasedpenetrationoffiberinthelast mileisperhapsthemostsignificantfactorinthe shifttowardfewerandmorecentralizedngcos. Connectivityoverthelastmilemaybedeliveredby apon.Thismightcomeintheformoffiber,orasa hybridsolutioninwhicharelativelyshortcopper extensionusingvdsl ordocsis technologyextends thefiberfromthengco tothecurb. Asanenabler,fiberappliesprimarilytothecentral officesforfixedservices,asmobileofficesalready tendtobepositionedtooperatewithlong-reach accesstechnologies. Virtualization Asitdecouplesapplicationsfromtheunderlying hardwareplatform,virtualizationisoneof thekeyenablersforflexibleserviceand functiondeployment. Withgoodorchestration,virtualization technologiesenablemosttypesofworkloadsto beconsolidatedoncommonmulti-corecompute platforms.Furtherreductionofhardwareinthe ngco canbeachievedbypoolingworkloadson acommoncomputeresource,andadditional powersavingscanbegainedthroughdynamic workloadreassignment. Thesignificanceofvirtualizationinfuturecarrier networksisclearlyreflectedbythemassiveeffort beingputintothisareabyoperators,vendors,and standardizationbodies.Theheightenedfocus onallaspectsofvirtualizationbodeswellforthe accelerationofitsadoption. Automatingthevnf lifecycle Automatedorchestrationofvirtualfunctions’ instantiation,capacityelasticity,andfunction terminationarecriticalnetworkcapabilitiesthat enablefunctionstobedeployedquicklyandflexibly inmultiple,geographicallydistributedngcos. Orchestrationiscentraltotheoperationofany virtualizationenvironmentofferingmulti-tenancy– whetheritisforanoperator’smanyinternaltenants, orexternalresidentialandenterprisetenants. Computeperformance Thecontinuousimprovementsincompute performancecanbeattributedtoanumberof differenttechnologies.Cores,forexample,have becomefaster,thecorepersocketratiohasrisen, on-chipcacheshavebecomebothlargerand faster,andaccesstimestoperipheralmemoryand storagehavedroppeddramatically.Today,itisfairly commonforanindividualcpu tocontaintensof cores,eachrunningat3ghz oncots hardware,with single,dualorquadsockets.Inaddition,i/o speeds haveincreased,enablingmodernserverstosupport dual(andpossiblymore)40gbpsnics. Theincreasesincomputeandi/o performance haveinturnwidenedthesetoffunctionsthat mightbenefitfromvirtualization.Andso,network designisnolongerrestrictedtothevirtualizationof traditionalit andcontrol-planeintensiveworkloads, butcanbeexpandedtoincludetraditionaltelecom networkfunctionsthatdemandhighuser-plane performance,suchasvirtualroutersandvirtual subscribergatewaysincludingvirtualbngs and p/s-gws. Ascomputecapabilitiescontinuetoimprove, anequivalentreductioninthehardwarefootprint ofaccessfunctionswilloccur.Thisnotonlybrings benefitsintermsofcostandenvironmentalimpact, butalsoenablesfunctionsthatbenefitfromproximity totheuser,previouslydeployedinmorespacious dcs,tobedistributedanddeployedinthengco. dc switchingfabric Tovirtualizenetworkfunctionsandother workloadsasfaraspossible,thengco obviously needsappropriatecomputeandstoragecapacity. Emergingdc fabrics–basedonmerchantsilicon leaf-and-spineswitches–thatarescalable,and offerhighcapacityatlowcost,providejusttheright kindofinternalnetworkdesignbetweencompute- and-storagecomponentsandthephysicalwan and accessgateways. Mostngco fabricswillbeconfiguredas non-blockingclos[1]networks,possiblywithunder- subscribeddimensioning,eventhoughsucha structureisnotstrictlyrequired. Software-definednetworking Applyingtheconceptsofsdn toanetworkmakes itcentralized,dynamicallyprovisioned,and programmable.Theagilityandflexibilitysdn offers willbecriticalinprovidingnewandmultiple-service operatorswiththecapabilitytoofferwhatever servicestheyliketotheirsubscribers. Keyarchitecturalcomponents Figure2showsthelocationofthengco andhow itisconnectedtothefixedandmobileservicesit offerstosubscribersthroughthevariousaccess domains.Thediagramalsoincludesanabstract representationoftheinternalstructureofthengco anditsconnectionsdeeperintothenetwork.The orchestrationcomponentmanagesthefunctions andinfrastructureoftheinternal officeaswellascertainexternalentitiessuch asaccessrouters. Infrastructure Thengco infrastructureconsistsofthree majorcomponents: 〉〉 the switching fabric that links all other components together 〉〉 gateways – to the access domain and the wan 〉〉 servers and storage Initially,non-virtualizedbaremetalappliances thatperformspecificfunctionswillalsobepart oftheinfrastructure.Theseappliancesmightbe incorporatedintothegatewaysorbeimplemented onseparatehardwareplatforms,depending onthecapacityofthegatewayandhowwellthe hardwareperforms. Switchingfabric Thestructureofanngco mayuseoverlay/underlay designprinciplesoradoptamoretraditional approach.Inanoverlay/underlaydesign,the switchingfabricformstheunderlayandisagnostic ofserviceendpoints.Intraditionalarchitectures, theswitchingfabricisfullyawareoftheservice endpoints.Thesizeandscaleofthefabricvaries accordingtotherequirementsandlocationofthe office.Forexample,asmallngco servingtensof thousandsofusersmayconsistofjustafewswitches THE INCREASED PENETRATION OF FIBER IN THE LAST MILE IS PERHAPS THE MOST SIGNIFICANT FACTOR IN THE SHIFT TOWARD FEWER AND MORE CENTRALIZED NGCOS
  13. 13. 24 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 25 SECOND WAVE CONSOLIDATION ✱✱ SECOND WAVE CONSOLIDATION andsupportaminimumsetoflocalfunctions, whereaslargerofficesmayincludeaswitchingfabric capableofsupportingextensivelocalservicesfor millionsofsubscribers. Thestructureofthefabric,especiallywhen itcomestolargeroffices,islikelytobebasedon commondata-centerdesignpractices,withan underlayclosarchitecture,usingaclusterofleaf- and-spineswitcheswithsame-lengthlinks,offering potentiallydeterministicdelayandlatency.Inaclos underlay,loadbalancingwithinthefabricisachieved byutilizingthemultiplepathsbetweensourceand destination.Eithercentralizedsdn controllers ordistributedroutingprotocolssuchasbgp or igp willbeusedtobuildtheforwarding,routing, andswitchingtables.Tobuildthefabricunderlay forlarge-scalengcos,theindustrypreference isleaningtowardtheuseofdistributedrouting protocols,astheyaresimplisticandhaveaproven trackrecord. Merchantsilicon-basedwhiteboxescanbeused forfabricswitches,especiallywhenprovidingasimple underlay.Theseboxestendtobelesscapablebutoften havealowerprice-to-bandwidthratiothantraditional switches.Whiteboxesofferentirelydecoupled networkingosandhardware,andbyusingatoolsuch astheOpenNetworkInstallEnvironment(onie), forexample,theinstallednetworkoscanbeeasily swappedoutwithanotherone–allowingoperators toloadtheosoftheirchoiceontoinstalledhardware. So,whiteboxesnotonlycontributetoreducingcosts; theyperhapsmoresignificantlyprovidenetwork programmabilityandflexibility. ShowninFigure3,thengco fabricconceptually representsadisaggregatedrouterthatcanbereadily scaledoutbyaddingleaf-and-spineswitchesas needed.Thefabricmayneedtosupportanumberof underlaytechnologiesincludingip andEthernet,and mpls mayberequired,especiallyincarrierdomains, toensureoperationalsimplicityandseamlessend-to- endinteroperabilitywiththeinstalledbase. Intheeventofaswitchfailure,thefabric automaticallyreroutestrafficthroughtheremaining switchesuntilthefailedswitchhasbeenmanually replacedandauto-configuredbyafabricmanager, allowingthesystemtooperatewithouthavingto waitforamaintenancewindow. Optimumtrafficmanagementrequiresaholisticand real-timeviewoftheavailablenetworkbandwidthand trafficpatterns.Flowstatisticsarecollectedatregular intervals,andwhenanalyzed,providetheinformation neededtodetectandavoidcongestion,guarantee betterutilizationoffabricresources,andadminister prioritizationpolicies. Gateways Accessandwangatewaysactasinfrastructure gateways,andtendtobeconnectedtospecialleaf nodes.Thewangatewayfunctioncouldalternativelybe implementedusingspineswitches. Accessgatewaysthatterminatecustomeraccess linksmayrequireextendedcapabilitiessuchasdeep buffers,trafficmanagementandothermoreadvanced qos capabilities,largeforwardingtables,andacls thatarenotusuallypresentinmerchantsilicon-based whiteboxes.Accessgatewaysterminatedifferentaccess technologiessuchasdocsisandgponolt.olt functionscanbevirtualizedwiththemaclayerandthe opticsseparatedfromthecontrol-andmanagement- planesoftware.Thehardwarepartofthegatewaycan beimplementedonasmallsfpformfactor,whilethe softwarepartcanbevirtualizedandhostedonany serverwithintheco. Usingavarietyofcommunicationprotocols(such asip,mpls, andotn/wdm),wan gatewaysconnect centralofficeswithotherngcosandcos,central andregionaldatacenters,aswellasothercarriers andthewiderinternet. Computeandstorage Thegeographicclosenessofthengco tousers providesastrongincentivetohousecertainfunctions andservicesthatbenefitfromthisproximityinthe ngco.Computeandstorageresourcesexistinthe ngco torunvirtualizednetworkfunctionssuch asvbng andvp/s-gw,aswellasmoretraditional servicessuchasvod,withlocalcaching. Thegeneralpurposenatureofcomputeresources deployedinthengcoiskey,asanynetworkfunctionor servicecanbeinstantiatedonthem,supporting Figure 2 The ngco in the operator’s network Figure 3 Disaggregation of routing functions Fabric Fabric Access GW WAN/DC GW vP/S-GW NGCO Internet Hyperscale/ OTT data center vBNG CMS NMS SDN OPEN DAYLIGHT OpenStack vPE Control Fabric Subscribers WAN Apps TOR TOR FCAPS management Fabric underlay control Network overlay control Gateway Access functions WAN functions TOR vP/S-GW vPE optional vBNG optional Line cards Spine NB-IoT 〉〉 roaming 〉〉 〉〉 〉〉
  14. 14. 26 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 27 SECOND WAVE CONSOLIDATION ✱✱ SECOND WAVE CONSOLIDATION thebreakawayfromtraditionalhardwareand softwarecoupling. Theamountofcomputeandstoragelocatedin agivenngco willdependonitssizeandoperator preferencesforcentralizationversusdecentralized functiondeployment.Offeringcloudservices,for example,requiresadditionalcomputeandstorage, whichinturnincreasesthesizeofthengco. Overlayservices Ifthengco implementsoverlayservicesusingan underlayswitchingfabric,anoverlayencapsulation techniqueisrequired.Thistechnologycanalsobe usedtoprovidetenantisolationtooperator-internal stakeholdersandsubscriberisolationforngcos withcloudservices. Commonencapsulationtechnologiesinclude vxlan,nvgre, andmpls vpns,andcanbe implementedvirtuallyinvSwitchesorinhardware onleafandpossiblygatewaynodesifhigher performanceisrequired. Regardlessofthelocationandtypeofoverlay technologyused,configurationwillbeautomatedby anoverlaycontrollercoupledthroughnorthbound APIstotheautomaticprovisioningofanytenant- relatedfunctions.Forexample,theoverlaycontroller couldbeodl-basedcoupledtoOpenStack throughNeutronapis.Thesameapiscanbeused byadditionalapplicationssuchastheoss/bss. Theoverlaycontrollercommunicateswithvirtual networkswitchesorbaremetaldevices(gateways andleafswitches,forexample)preferablythrough opensouthboundinterfacessuchasOpenFlow, xmpp,ornetconf. Virtualizednetworkfunctions Intoday’scos,traditionalnetworkfunctionsand workloads,suchascachesandwebservers,run onverticallyintegratedplatforms.Inthengco, theseelementswillberunasvirtualizednetwork functionsoncots hardware. nfv technologymakesiteasiertocreateand scaleseparatelogicalnodesandfunctions,andif necessary,theseelementscanbeisolatedforuse byaspecifictenant.Thisistheconceptofnetwork slicing.Networkslicesareindividuallydesignedto meetaspecificsetofperformancerequirements tailoredtotheapplicationrunningontheslice. Thevirtualinfrastructureofasliceisisolatedfrom otherslicestoensurethatallslicesofthenetwork runefficientlyandperformancetargetsaremet. Thenfv approachprovidestheflexibilityneeded toprovisionnetworkresourcesondemand,andto tailorslicestospecificusecases,enablingoperators todelivernetworkingasaservice.Thebeautyof networkslicesliesintheirabilitytobeoptimizedto suittheapplication.Inotherwords,high-availability servicescanrunonslicesoptimizedforresilienceto hardwareandsoftwarefailures,whereasanm2m signaling-intensiveapplication,forexample,canrun onalow-latency,low-bandwidthslice. Automation Inthengco,allkeyoperationalcomponentsare automated.Thisremovestheneedformanual configuration,whichispronetoerror,costly,and time-consuming. Thefabricmanageroverseestheautomated partsofthengco,configuringandmanagingthe underlyingfabricswitches,andsupervisingthe performanceofthefabric.Thefabricmanager continuallyandautomaticallymonitorsthephysical fabricnode-and-linktopology,itvalidatesthephysical cabling,andconfiguresleaf-and-spineswitcheswith associatedprotocolsandpolicies.Thefabricmanager mayusedevops toolssuchasCheforPuppetfor initialconfigurationandsoftwaremanagementtasks (lldp configuration,managementaddressing,and os componentupgrades),afterwhichprogrammatic interfacessuchasnetconf/yang canbeused toconfigurenetworkprotocols,qos policies,and statisticsontheinterfaces.Forcentralizedsdn- basedcases,thefabricmanagercanuseOpenFlow toconfigurethenecessaryforwardingentriesinthe underlayswitches. Serviceorchestration Serviceorchestrationautomaticallyinstantiates applicationsandconfiguresnetworkservices accordingtoservice-levelspecifications. Automationofthesetaskscandramaticallyreduce thetimetoinstantiateoraddnewdevicesorservices tothenetwork,whichincreasesnetworkagility, makingreal-timeserviceprovisioningpossible. Migration Formostoperators,themigrationofnetwork architecturefromthecurrentco deploymenttoone basedonfewerngcoswillbegradual.Whilesome ngcoswillbebuiltasgreenfielddeployments,for themostpart,existingcoswillevolve,requiring thecoexistenceofdecoupledsdn/nfv equipment, togetherwithtraditional,tightlycoupledhardware andsoftware.Duringthemigration/coexistence period,managementandorchestrationcomponents needtobeabletosupporttheheterogeneous (coupled/decoupled)environment;by,forexample, abstractingthedifferencesbetweenthetwo architectures,andusingcommonnorthbound THE GEOGRAPHIC CLOSENESS OF THE NGCO TO USERS PROVIDES A STRONG INCENTIVE TO HOUSE CERTAIN FUNCTIONS AND SERVICES THAT BENEFIT FROM THIS PROXIMITY IN THE NGCO Figure 4 etsi nfv reference architectural framework OSS/BSS Event manager PNF Hardware NFV orchestrator (NFVO) VNF-specific VNF manager Ericsson VNF NFVI NFV service catalog NFV instances NFVI resources VNF Virtual infrastructure manager (VIM) Os-Nfvo VeEn-Vnfm VeEn-Vnfm Nf-Vi Vi-Vnfm Nfvo-Vi Nfvo- Vnfmcatalog
  15. 15. #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 2928 SECOND WAVE CONSOLIDATION ✱✱ SECOND WAVE CONSOLIDATION interfacestoothersystems,suchasend-to-end serviceorchestrationandoss/bss. Throughouttheperiodofcoexistence,network functionswillbephysicallyandvirtuallyinstantiated withcapacityandsubscriberspooledacrossboth, andasFigure4shows,orchestrationsystemswill berequiredtosupportbothtraditionaland decoupledarchitectures. Conclusions Networkarchitectureisundergoingamassive transformationintermsofincreasedlevels ofautomationandprogrammability.This transformationhasbeenenabledbyanumberof technologies,butprimarilybythedisaggregationof softwareandhardware.Thetransformationisbeing drivenbynewbusinessopportunities,expected gainsinoperationalefficiency,andtheneedfor rapidtimetomarketforservices.Astheunderlying technologies–virtualizationandsdn –become moremature,therateoftransformationwillrise. Thenextgenerationcentraloffice,orngco,has beendesignedtotakeadvantageofthegainsbrought aboutbyadecouplednetworkarchitecture.The benefitsforoperatorscomeintheformofnetwork intelligence,flexibility,andeaseofscalability,allof whichbringopexandcapexbenefits. Thengco isbasicallyaminidatacenterthat providesconvergedfixedandmobileservices. Comparedwithatraditionalco,thengco canserve alargersubscriberbaseacrossawidergeographic area.Thengco hasbeenbroughtaboutthrough: 〉〉 reduced co density, as a result of greater distances achievable by fixed access technologies 〉〉 the introduction of sdn/nfv technologies 〉〉 advancements in hardware technologies in terms of low-cost, high-throughput switches 〉〉 infrastructure automation and service orchestration Architecturally,deployingthengco asa minidatacenterintroducesagreaterlevelof intelligenceintothenetworkinadistributed fashion,asapplicationsarereplicated,orshifted, fromcentralizeddatacentersouttongcos. Computeresourcesinthengco canbeused forrunningapplicationssuchasrichmediaand rendering,orlatency-sensitivegamingapps.With thesecapabilities,thengco willbecomepartofa distributed,intelligentcloudresource. Thengco bringswithitanumberofsavings, requiringlessspace,power,andcoolingthanthe sumoftheindividualtraditionalcostheyreplace. On-sitestaffingrequirementsshouldbereduced, asprovisioningandmanyaspectsofmaintenance arecontrolledremotelyandautomated.Overall,the ngco willresultinfewercentralofficesorincreased accesscoverageandserviceconsolidation,with reducedneedfornewrealestateasequipment continuestocompact. Nail Kavak ◆ joined Ericsson in 2000, and is currently working as principle architect for the system and technology group in Development Unit ip. He has in-depth experience in the design and deployment of ip/mpls and optical networks for carrier networks. Most recently, he has managed a number of network transformation projects for Tier 1 operators in the dc Networking space. He holds an m.sc. in computer science and engineering from Linköping University, Sweden, and a technical licentiate from the kth Royal Institute of Technology in Stockholm. https://www.linkedin.com/ in/nail-kavak-8ba9481 Andrew Wilkinson ◆ is an expert in ip networking at Ericsson’s Development Unit ip. He holds an m.sc. in telecommunications from the University of London. He joined Ericsson in 2011 having previously worked for mobile network operators in Europe and North America. https://www.linkedin. com/in/andrew-wilkinson- 0b377712 John Larkins ◆ is a senior director of technology at Ericsson’s ip Design Unit in San Jose, California, where he is responsible for technology evolution, including network and systems architecture solutions ranging from asic requirements definition to product implementation architectures and collaboration with network operators on future target network architectures. https://www.linkedin.com/ in/larkins Sunil Patil ◆ is a principal engineer in ip networking at Ericsson’s Development Unit ip. He joined Ericsson in 2000, where he has worked on architecture, design, and development of multiple ip routing products. His current focus is on driving technology innovation in the areas of sdn, orchestration, ngco, and data center networking for laas, paas, and caas. He holds an m.sc. in computer networking from North Carolina University, the us, and an m.b.a. from Duke University. https://www.linkedin.com/ in/sunilbpatil Bob Frazier ◆ is an expert in ip system architecture at Ericsson’s Business Unit Cloud & ip. He holds a ph.d. in electrical engineering from Duke University in North Carolina, the us. He joined Ericsson in 2007 and has worked in ietf, ieee, and Broadband Forum standardization. His current interests are ip software architecture and data center networking. https://www.linkedin.com/ in/bob-frazier-a961572 theauthors References: 1. Bell Labs Technical Journal, 1953, A Study of Non-Blocking Switching Networks, Charles Clos, abstract, available at: http://ow.ly/YGZ2F
  16. 16. 30 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 31 STANDARDIZING NARROWBAND ✱TECHNOLOGY TRENDS ✱✱ TECHNOLOGY TRENDS Five trends Keeping up with the relentless pace of change in the ict industry is a daily challenge for modern tech companies. The key to long-term success lies in the ability to understand change almost before it occurs and seize the opportunity to shape evolving technologies. Tech companies often gain competitive advantage by causing market disruption through their ability to understand and act on technology trends. Like waves in the ocean, it’s much easier to ride these trends if you can see them coming and read them right. (But of course, true technology leadership happens when you start making your own waves.) As I see it, there are five key technology trends that will stimulate innovation within the ict industry in the coming year, creating new value streams for consumers, industries and society. All five pivot around a technology-enabled business ecosystem made possible through a universal, horizontal and multipurpose communications platform. shapinginnovationinICT #1SPREADING INTELLIGENCE THROUGHOUT THE CLOUD Distributed machine intelligence moves into the cloud #2SELF-MANAGING DEVICES Intuition, self-learning, and increasingly autonomous devices #3COMMUNICATION BEYOND SIGHT AND SOUND Human interaction augmented by tactile internet #4FUNDAMENTAL TECHNOLOGIES RESHAPING WHAT NETWORKS CAN DO New materials and manufacturing techniques enhance networking capabilities #5WEAVING SECURITY AND PRIVACY INTO THE IOT FABRIC Automation makes security controls real-time and proactive BY ULF EWALDSSON, CTO 3130
  17. 17. 32 ✱ TECHNOLOGY TRENDS TECHNOLOGY TRENDS ✱ 33#02, 2016 ✱ ERICSSON TECHNOLOGY REVIEW c o n n e c t e d smart machines, such as robots and autonomous vehicles, are fundamental to the evolving Networked Society. Enhanced cloud architecture that can distribute and share machine intelligence will enable smart connected machines to work on an increasingly higher level. ■ Supportedbyadvancementsinartificial intelligence(ai)–particularlyintheareas ofbigdataanalytics,machinelearning andknowledgemanagement–rapid progresshasbeenmadeintermsofwhat smartmachinescando.Developmentsin connectivityandcloudtechnologiesare makingitpossibletodistributeandshare machineintelligencemoreeasily,atalower cost,andonamuchwiderscale thanbefore. Whenconnectedtothecloud, smartmachineswillbeabletousethe powerfulcomputational,storageand communicationresourcesofstate-of- the-artdatacenters.Today’sintelligent softwareroboticssystemsarecapable ofsupportingrepetitiveadministrative taskswithcurrentdevelopmentpushing towardadvisorytasks.Cloudification shiftsthecapabilitiesofthesesystems intoanewspherethatincludescomplex problem-solvinganddecision-makingona mass-marketscale. Connect,store,compute…andshare Shiftingsystemsintothecloudenables communitiesofcollaboratingrobots, machines,sensorsandhumanstoprocess andshareinformation.Eachnewinsight collectedwithinacommunitycanbe sharedinstantly,whichincreasesthe effectivenessofcollaborativetasks,and improvesperformancethroughoutthe system,withacommonawarenessof systemstatesharedbyallparticipants,as wellasasharedknowledgebase. Adistributedmachineintelligence architectureofferslowerimplementation costs.Sharingabackboneofalmost unlimitedcomputationalpowermakes itpossibletobuildlightweight,low-cost robotsandsmartmachinesthatrequirea lowlevelofcontrolandaminimumamount ofsensorsandactuators.Application- specificrequirementsrelatedto responsivenessandspeedwilldetermine whetheralocalorglobalcloudismost suitable,andhowmuchintelligencecan bedistributed. Smartandmobilecapabilities virtuallyeverywhere Intelligentcloudswillcreatenewvalue chainsinmanyindustrysegments,but someoftheforerunnersincludemining, agriculture,forestryandhealthcare. Newopportunitieswillopenupforall organizationsandpeopleinvolvedinthe supplychainfromthemanufacturerto thecustomer.Consideranautomated agricultureapplication.Theapplication remotelycontrolsfarmmachinestocarry outvariousfarmingtasks.Toharvest maturecrops,forexample,thesystemwill controlthenecessarymachinestocut, gatherandtransportthem.Eachindividual machinewilltakelocaldecisionstoensure securecompletionofitssettasks,working inconjunctionwithallthemachines involvedintheharvesting.Weather reportsgatheredfromanotherdistributed cloudapplicationareusedbythesystem tocarryoutharvestinginanoptimalway. Contactwiththefarmeroccursonlywhen participatingmachinescannotresolve issuesthemselves. Theharvestingexamplehighlights justoneofthemanycomingapplications thatwillrelyonmultipleinformation sources,cloud,anddistributedmachine intelligence.Toensurescalabilityand widespreaduptakeofsuchapplications, thechallengeliesintherapiddevelopment andproliferationofuniversallyaccessible mobilecapabilities.5g willprovidea resilient,high-availability,low-latency networkthatoffersapplicationswith integratedcomputingandstorage resourcesthatareideallyplacedtomeet latencyrequirements.5g iswellmatched toindustrialroboticsapplicationsbecause, likeotherradiotechnologies,itremoves theneedforcablingandminimizes infrastructureadaptions,butitalsooffers identitymanagement,optimumplacement ofresources,andencryptionforsecurity andprivacy. #1 ERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 Spreadingintelligence throughoutthecloud
  18. 18. 3534 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 TECHNOLOGY TRENDS ✱✱ TECHNOLOGY TRENDS c o m b i n i n g s e n s o r y d ata with AI techniques enables the data from massive numbers of sensors to be merged and processed to create a higher-level view of a system. ■ Connectedsmartdeviceswillchange ourlivesinmanyways.Theserangefrom simpleservicesthatopenyourgarage doorasyourcarapproaches,forexample, toradicallynewbusinessopportunities involvingservicesyettobeinventedand marketsyettobediscovered.Combined withintelligenthandlingofdata,smart devicescanboosttheproductivity andprofitabilityofanybusiness.But toenablethedeploymentofbillionsof smartdevices,thecostofmanagingand monitoringthemneedstobelow.Evolving softwareandcommunicationstechnology areshiftingtowardthecreationof autonomousandself-managingdevices. The Internet of Things (iot) means automation and intelligence in everything that is connected. This implies that a collective intuitive behavior among a wide range of devices for a wide range of applications is possible in the future. The connectivity allows objects to be sensed and actuated remotely, creating a bridge between the physical and digital world. It’sthecombinationthattriggers theeffect Beyond the physical devices embedded with processors, software, sensors, actuators, and connectivity, it is the combination of sensory data and ai that enables more effective and accurate interactions. It is by merging data from a multitude of sensors that a superior baseline for intelligent processing is created. These are the common denominators that push IoT development further. Fromaconnectivityperspective,two distinctanddifferentusecasesemerge. Oneextremeisthemassivemachine- typecommunication(massivemtc) thatcansupportmillionsofconnected devicessuchasenergymetersandlogistics tracking.Here,wearelookingatdevice batterylifetimesbeyond10yearsandcost reductionintheorderof80percentaswell as20db bettercoveragecomparedwith presentstate-of-the-artsolutions. Theotherextremeisthecritical machine-typecommunication(critical mtc),whichentailsreal-timecontrol andautomationofdynamicprocessesin variousfieldssuchasvehicle-to-vehicle, vehicle-to-infrastructure,high-speed motion,andprocesscontrol.Critical parameterstoenabletheperformance requiredarenetworklatencybelow milliseconds,ultra-high“fivenines” (99.999percent)reliability.Thefuture networkarchitectureneedstocaterfor bothmtc scenarios. Keytechnologyadvancements The2016EricssonMobilityReport (https://www.ericsson.com/res/docs/2016/ ericsson-mobility-report-2016.pdf) predictsthattherewillbe28billion connecteddevicesby2021.Onthedevice side,thekeytechnologydriveristhe evolutionofsensors,actuators,processors, memories,andbatteries.Beyond conventionalelectronics,wewillsee implementationsofnanoscaletechnologies basedonthin-film,graphene,andquantum sensors.Wecanexpectanysizeandshape ofdeviceinthefuture. Anotheremergingkeytechnologyis thatofanadvancedsoftwaretoolbox leveragingadvancedanalytics,machine learning,andknowledgemanagement withprocessingcapabilitiesofreal- timestreamingdata.Intelligentcontrol logicisanotherinterestingarea.There isanincreasingneedforstandardized platformsandsoftwareprotocols.These willinevitablydrivemarketconsolidation, withmassivecostsavingsandproductivity gainsasaresult. Effectiveconnectivityandidentity managementarefundamentaltothe futurenetwork.Theseimplyautomated deployments,aggregatedsubscription managementaswellasembedded provisioningandcontrolthroughthe wholelifespanofthedevice. Whatdoesthismeanforthefuture roleofnetworks? iot devicesenableustomonitorsensors andautomatealotofprocesses.Theadded intelligenceneededisafeaturethatwill mainlybeembeddedinthenetworkitself. Foriot technologytoliveuptoits promiseandbeappliedonamassivescale throughoutsociety,itmustbebuiltona secure,global,telecom-gradenetwork thatisbasedoncommonstandards.This willalsoensureahealthycompetitiveand innovativeecosystem. Intermsof5g,suchanunderlying networkinfrastructureisalreadyinplace –readytoshowhowwellitisscalingand howitscost-efficiencypropertiessupport iot applications.5g offersbothsuper- highbandwidthwithultra-lowlatency andextremebatterylifefordevices. Bycombiningcloudintelligencewitha powerfulbutenergy-efficientwireless connection,evenverysimpleand inexpensivedevicescanbemadesmart andgenerategreatbusinessvalue. THE CONNECTIVITY ALLOWS OBJECTS TO BE SENSED AND ACTUATED REMOTELY, CREATING A BRIDGE BETWEEN THE PHYSICAL AND DIGITAL WORLD Self-managingdevices #2
  19. 19. 3736 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 ✱ TECHNOLOGY TRENDS TECHNOLOGY TRENDS ✱ c o m m u n i c at i o n will evolve in a highly remarkable way over the coming years, as interaction between human beings and machines evolves to include additional experiences and senses. The internet you can feel is on the horizon. ■ Today,2d videoisthemostadvanced formofcommunicationpeopleuseto connectwitheachother.Inthefuture, peoplewillbeabletoparticipateindistant businessmeetingsorattendafamily gatheringbysendinganaugmented3d selfie.Iamsuremanypeoplearelooking forwardtothedayitwillbepossibleto attendeventssuchasMobileWorld Congress,thefifa WorldCup,orthe SuperBowlvirtually. Emergingtechnologiesinthefields ofthetactileinternet,virtualrealityand augmentedreality–supportedby5g networkevolution–areshowingsignsthat theabilitytoexperienceaneventvirtually isnolongersciencefiction,butafeasible reality,andindicateagiantstepforward ininnovation. Thetactileinternetisfoundedonthe visionaryprinciplethatallofourhuman sensescanbeembeddedinhuman- machineinteraction.Usinghaptics (interactioninvolvingtouch),remote experiencescanbeanearreal-time representationofreality.Toaccomplish suchrealisticremoteexperiences,however, theloopconnectingthedisciplines ofrobotics,ai,andcommunications needstobeclosedandnear-zerolatency requirementswillneedtobemet. Virtualandaugmentedreality(vr andar)areexpectedtobecomeintegral technologiesoftheNetworkedSociety, potentiallydisruptingtheconsumer electronicsmarket. Pushingtheboundariesof traditionalphysics Toclosetherobotics,ai,and communicationsloopquickly,Ericsson hasstartedacollaborationonthetactile internetwithKing’sCollegeLondon.As theresearchteamputsit,“Weneedto beatthelimitsofthetraditionallawsof physics,aseventhespeedoflightisnot fastenoughtoenablethesekinds ofapplications.” Inthiscontext,tactilecommunication enableshapticinteractionbetween controlandmachinewithvisualfeedback. Technicalsystemswillneedtosupport audiovisualinteraction,andenable remoteroboticsystemstobecontrolled withanunnoticeabletimelag.End-to- end,componentsotherthanthephysical distanceseparatingcontrolfrommachine addtothetotalsystemdelay.Forinstance, videocodingandrenderingrequirea substantialamountofcomputational power,andsothesecomponentsincrease overallsystemdelay. Thistypeofnext-generation communicationwillcontributetothe resolutionofcomplexchallengesthatarise inmanysectorssuchaseducation,health care,personalsafety,smartcity,traffic managementandenergyconsumption. Somebusiness-relatedexamplesinclude virtualstores,interactive3d designlabs, training,interactiveentertainment,and enterprisecommunication.Presently,the gamingindustryistheprimaryincubator forar andvr. Notjustrawspeed – some intelligencetoo Human-to-humanandhuman-to- machinecommunicationswillputhigh demandsonfuturenetworks.Solutions supportinghighcapacityandextremely lowlatencyincombinationwithhigh availability,reliability,andsecuritywill definethecharacteristicsofthenetwork. Inmassivevideodistribution,forexample, theneedforcapacityiscreatedbycertain applicationneedsforhighresolution, highdynamicrange,andhighframerate, whichinturnnecessitatelinkspeedsin gigabitspersecond.Butit’snotjustabout rawspeed.Ourresearchinthisarea has,forinstance,investigatedtheideaof dividingtheamountoftransmitteddata intopriorityhierarchieswithdifferenttime requirements,transmittingonly datathathasbeenmodifiedand anticipatingchanges. Communicationbeyond sightandsound #3
  20. 20. TECHNOLOGY TRENDS ✱✱ TECHNOLOGY TRENDS 39#02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 201638 theprocessingunitandothervital componentsinacommunicationnetwork. Photonicswilladdpropertiessuchas lowpropagationloss,highdata-transfer density,andexcellentsignalintegrity. Bridgingthegapbetweenopticaland electroniccomponents,siliconphotonics willshrinkeverythingincludingthe footprint,powerconsumption,andcost ofhigh-speednetworkapplications. Furthermore,siliconphotonicswillallow forgreaterdisaggregationoffunctions, whichopensupformoreefficienthardware architectures,whileenablingmore aggregateddatatraffic. Qubits–smallbutpowerful Slightlyfurtherintothefuture,quantum computingpromisestobringaboutan exponentialincreaseincomputational power.Quantumcomputingisa technologythatbuildsonthequantum propertiesofelementaryparticles (qubits).Qubitscanbeentangledwith eachotherandcantakeonintermediate valuescomparedwithordinarybits, whichcanonlybeeither1or0.This way,aquantumcomputercanincrease parallelismandradicallyreducethe computingeffortsneededtoaddress certaintypesofproblems.Researchers havealreadysucceededincreatingqubits withinasemiconductor,andthefirstfully operationalquantumcomputer wasdisplayedattheendof2015.One ofthemainchallengesistokeepthe quantumstateunperturbed,which requiresextremelylowtemperatures andverygoodinsulationfromthe surroundingenvironment. Bymatchingtheexponentialexpansion ofthedigitaluniversewithcomputational powerthatalsogrowsexponentially, weareconfidentthatwewillbeableto continuetostayontopoffuturedemands forcommunication. Fundamental technologiesreshaping whatnetworkscando #4 NEW MATERIALS IN COMBINATION WITH INNOVATIVE MANUFACTURING TECHNOLOGIES PROMISE TO RADICALLY ENHANCE NETWORK CAPABILITIES t h e l aw s o f p h y s i c s are the only real restriction on the development of communication networks. Ericsson is firmly committed to pursuing innovations that challenge present system limitations to help us reach beyond what is possible today. ■ Whilebecomingincreasinglyversatile, thenetwork’sfundamentalbuilding blocksarealsobecomingmuchsmaller, mimickingthewaylivingthingshave evolved.Thenetworkofthefuturewill beakintothedigitalembodimentof anintuitiveorganismthatisableto handlevastamountsofconsciously intelligentautomatedresources.New materialsincombinationwithinnovative manufacturingtechnologiespromiseto radicallyenhancenetworkcapabilities. Whichtechnologieshavethegreatest potentialtospurnetworkevolutionin thenearfuture? Inthesemiconductorarea,awiderange ofnewmaterialsandmanufacturing technologieswillsoonbecome mainstream.Newpackagingand integrationtechnologiesoffersubstantially increasedbandwidthinadditionto powerreductions. Thesemiconductorindustryisalso atthecuspofleveragingnewmemory technologiesthatwillbeabletotakeon differentrolesinthesystemmemory hierarchy,aswellasofferingsubstantial improvementsinsysteminputand outputperformance. Thesemiconductorindustryadvances throughcontinuousscalingoftraditional cmos.Majorplayersareworkingonthe 10nmnode,andindustryroadmapsinclude 7nmand5nmmanufacturingtechnologies. Advanced2.5d/3d integrationtechniques fornon-monolithicintegrationhave thecapabilitytoofferawholesystem functionintegratedonasinglechip. Thesesolutionsarebothcostandenergy efficient.Theintroductionofmulticore centralprocessingunitsolutionsatpower consumptionequaltoorlowerthantheir predecessorsisapredominanttrend. Othertrendsincludethedevelopmentof varioustypesofarchitecturesaimedat significantlyacceleratingprocessingspeed, suchasmassiveparallelcomputing. Electronsandlightblending innewways Advancesinsiliconphotonicsallow foropticalintegrationdirectlyinto
  21. 21. 40 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 41 TECHNOLOGY TRENDS ✱✱ TECHNOLOGY TRENDS i n a w o r l d where everyone’s personal and financial information is available online, cyber security and privacy are very serious issues for consumers, corporations and governments alike. And the rapid rise of wearables, smart meters, and connected homes and vehicles makes security and privacy more vital than ever. ■ The complexity and heterogeneous nature of future networks and connected devices will require security and privacy controls to be made an intrinsic part of every device, network, cloud and application. However, controls are only valuable if they can be managed in a fast and coordinated manner across all layers – preferably in an automated fashion, steered by policies and analytical insights rather than by the choices of an individual. Automated security and privacy management that is pervasive yet observable and auditable are the core characteristics that can enable the future Networked Society. Weavingintelligenceonthreelevels Threelayersoftechnologymakeitpossible toweavesecurityandprivacyprotection intoeverylayerofict:actualsecurity controls,securityanalytics,andanadaptive securityposture. Overthenextdecade,keysecurity controlswillincludedatasovereigntyand novelidentitymanagementcontrolsthat aretailoredtopeopleanddevices,aswellas encryptiontechnologies.Someencryption technologiesareintheearlyphasesof developmentbutwillbegintoappearonthe marketinthenextthreetofiveyears,asthe underlyingtechnologiesmature.Newroot- of-trusttechnologiesthatareapplicableto bothphysicalandvirtualenvironmentsalso showgreatpromise,andsignificanteffortwill beputintomakingthemareality. Novelsecurityanalyticstechnologies cannowprovideinsightsthatmakeit possibletocreatepredictivesecurity systemsasopposedtoreactiveones. Thesetechnologiescouldbeusedtocreate disruptivedatamanagementsolutionsin thenearfuture,butforthistohappen,we needtohavecontext-awaresecurityfeeds andsecurityanalyticsalgorithmsthat correlatethesefeeds,oftenacross multipledomains. Thethirdtechnologylayer,theadaptive securityposture,isachievedthrough automation,basedonsecurityanalytics insightsandpolicy-basedautomated orchestrationofsecuritycontrols. Itwillallbebuiltontrustednetworks Nosingleindustryplayerwillbeable toaddressallofthesechallengesonits own.Industry-widecollaboration,joint development,andstandardization–including vendors,serviceproviders,andusers–will beessentialinordertorealizethevisionof asecureNetworkedSocietythatprotects businessassetsandeveryone’sprivacy. Traditionally,networkserviceproviders rankamongthemosttrustedindustry players.Withthisinmind,Ibelievethat networkserviceprovidersandtheirnetworks willbethefoundationuponwhichthe trustforeverythingelse–devices,clouds, communicationsandapplications–isbuilt.At Ericsson,ourfocusisonenablingnetworksto playthiskeyroleacrossmultipleindustries. Weavingsecurityand privacyintotheIOTfabric #5 4140 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016
  22. 22. 42 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 43 CLOSING THE GAPS ✱✱ CLOSING THE GAPS EDVARD DRAKE IBTISSAM EL KHAYAT RAPHAËL QUINET EINAR WENNMYR JACKY WU ■Inthestackofcloudservicemodels,shownin Figure 1,paas fitsinbetweensoftwareasaservice (saas)(whichtargetsuserswithlicensedsoftware offerings)andinfrastructureasaservice(iaas) (whichaddressesthemanagementandsharingof hardwareresources). paas workswithvariouscloudmodels:public, private,orhybrid.Thehybridmodelcan,for example,beusedbyenterprisesandtelecom serviceproviderstooptimallycombinethedifferent handlingneedsofsensitiveandnon-sensitive workloads,wherethecommonmanagement interfaceenablessometobedeployedonaprivate cloudandothersonapubliccloud–asshown inFigure 2.Latency-sensitiveworkloads,for example,ortasksthatrequiresecurityorcontrolfor proprietarydatacanbedeployedonpremisesina privatecloud,whilenon-sensitiveworkloadscanbe deployedinapubliccloud,maximizingagilityand optimizingcosts. Dependingonthelevelofautomationand integrationprovided,paas solutionscanbe furtherdividedintotwocategories:structured andunstructured.Unstructuredplatforms leveragebasiccontainertechnologiesorpublic paas offeringsandareusuallymanagedor monitoredwithhomegrowntools. Technology- centriccompaniestendtofavorsuchunstructured platforms,astheyfacilitatedevelopmentand maintenanceofsolutionscustomizedtomeet businessneeds. Structuredplatforms,ontheotherhand,come withbuilt-infeaturessuchasorchestration, monitoring,governance,loadbalancing,andhigh availability.Thesecharacteristicsmakestructured platformssuitableforenterprisesortelecomservice providers,andarethereasonbehindEricsson’s focusonstructuredpaas. Thebenefitsbroughtbypaas Whatbenefitspaas canoffervaryfrombusiness tobusinessandfromoneapplicationtothenext, dependingonwhetherithasbeenspecifically designedforpaas orwhetheritsimplyrunsina paas environment.Thepaas approachiswellsuited toapplicationdevelopersandvendors,butitcan alsobeofgreatvaluetootheruserssuchassystem integratorsandserviceoperators. Someoftheconceptsusedinpaas,suchas multipleapplicationinstancesandcomponent- basedarchitecture,areestablishedapproaches inthetelcodomain.Tokeepthecomplexityof componentsatamanageablelevel,thetelcodomain hasalong-standingtraditionofmodulardesign. However,designingapplicationsspecificallyfor paas increasesthenumberofbenefitsforthe differentusergroups. Benefitsforapplicationdevelopers paas enablesdeveloperstofocusonthebusiness logicoftheirapplications,asitfreesthemfromthe concernsassociatedwithsettingupthenecessary foundationfordeployment,testing,adaptation, androllout.Indoingso,paas enablesinnovation accelerationandrapidtimetomarket. Independent of business, ways of working, or even technology adoption, the pressure on modern industries to shorten time to market through rapid development cycles is constant. The concepts of platform as a service (paas) and microservices – which have been gaining traction in the it world – are deeply rooted in this need to cut development times. And the benefits are equally important in the telco domain. But there are gaps that need to be closed before paas is suitable for telco. Most of the challenges relate to the need for additional features that telco applications typically require. Once PaaS is telco approved, new applications will need to follow a number of design patterns, so that the full advantages of the platform-as-a-service approach can be realized. p a a s is a cloud service model that allows developers to build, run, and manage applications in a way that best suits their business needs, and most significantly, in a way that is independent of the underlying hardware or software infrastructure. Typically, paas enables developers to deploy code on top of a software stack that includes a runtime environment for one or several programming languages, an operating system, and basic services to build upon. paas provides the building blocks for automated testing, continuous deployment, as well as supporting the devops approach, and as such simplifies the development process and reduces time to market. STRUCTURED PLATFORMS, COME WITH BUILT-IN FEATURES SUCH AS ORCHESTRATION, MONITORING, GOVERNANCE, LOAD BALANCING, AND HIGH AVAILABILITY telco-grade PAVING THE WAY TO Terms and abbreviations laas – infrastructure as a service | mmtel – multimedia telephony | paas – platform as a service | saas – software as a service | sctp – Stream Control Transmission Protocol | udp – User Datagram Protocol | vnf – Virtualized Network Function PaaS
  23. 23. 44 #02, 2016 ✱ ERICSSON TECHNOLOGY REVIEWERICSSON TECHNOLOGY REVIEW ✱ #02, 2016 45 CLOSING THE GAPS ✱✱ CLOSING THE GAPS Benefitsforserviceoperators Apaas-designedapplicationcanscalequickly andeasilywithflexibleworkloaddeployment, whichleadstooptimaluseofhardwareresources. However,careshouldbetakenwhendealing withapplicationsdesignedwithlargenumbersof lightweightcomponentsthatneedtocommunicate witheachother,toensurethatworkload deploymentsdonotnegativelyimpactperformance. Ingeneral,securityassuranceandgovernance bothbenefitwhenapplicationsrunonacommon frameworkthatprovidescollectiveapplication managementandsupportsintra-service communication.Forexample,theplatform approachremovestheneedtomanagemassesofad hocsecuritysolutionsandtherulesgoverninghow theyapplytoapplications. Howdomicroservicescontribute? Thesoftwareindustryiscurrentlyexperiencinga riseintheuseofmicroservicesandmicroservices architecture.Andwhilepaas andmicroservicesare twoseparateconcepts,viewingpaas incombination withmicroservicesandotherconceptslike containersanddevops,cansubstantiallyincrease theleverageofeachofthem. Microservicesisanarchitecturalpatternand anapproachtodevelopment.Essentially,this approachbuildsapplicationsfrom(ordeconstructs existingapplicationsinto)smallparts–eachwitha singleandwell-definedpurpose.Tocommunicate, theparts(ormicroservices)uselanguage-and technology-agnosticnetworkprotocols,andeach partcanbedeveloped,maintained,deployed, executed,upgraded,andscaledindependently. Technologychoicesarespecifictothemicroservice andeachmicroserviceshouldbeownedbyasmall teamofdeveloperstominimizetheoverheadof intra-teamcommunication. Overall,theabilitytodeveloppartsinan independentwayenablesrapidprogress,allowing developmenttokeeppacewithmarketdemands, andfacilitatesscalingofdevelopment. Decouplingandindependencybetween microservicesisfundamentaltoamicroservices architecture.Independencesupportsscalingover multipleteamsbecauseitenablesmanysmallteams toworkinparallel,withclearresponsibilities,a largedegreeoffreedom,andminimalinteraction. Decouplingalsoenablesthedifferentpartsofthe systemtoevolveattheirownpace. Avoidingdependenciesenablestechnology choicestobemadeonaper-microservicebasis.As newtechnologiesbecomeavailable,theycanbe implementedappropriatelywithouttheneedfor asynchronizedcross-microserviceupgrade.Asa result,eachmicroservicecanevolveattherightpace inawaythatismostappropriateforaparticular service:anefficientsystemthatlendsitselftothe creationofever-improvingservices. Whiletheadvantagesofamicroservices architectureareapparent,inpractice,thisapproach posesanumberofsignificantchallenges.Tostart with,thewell-knownfallaciesofdistributed computing[1]shouldbeavoided.Toperforma giventask,anumberofmicroservicesareinvoked sequentially,eachofwhichcontributesignificantly tooveralllatency,makingitmoredifficulttopredict thetheoveralllatencyofaservice.So,assuming,for example,thatbandwidthisinfinite,orthatlatencyis zerocanresultincostlyredesignwork.Challenges includetheoverallcomplexity,bothindevelopment andinruntime,ofalarge,highlydistributedsystems. Theabilitytotestasystemisequallychallenging, particularlywhenitcomestoadditionalcomplex failurescenarios. Oneway–andmaybetheonlyway–toovercome thechallengessurroundinglatencyistoacceptthat somepartsofthesystemneedtobedesignedwith THE ABILITY TO DEVELOP APPLICATION PARTS IN AN INDEPENDENT WAY ENABLES RAPID PROGRESS, ALLOWING DEVELOPMENT TO KEEP PACE WITH MARKET DEMANDS, AND FACILITATES SCALING software as a service (SaaS) platform as a service (PaaS) infrastructure as a service (IaaS) consume build and run on host on Users Developers and testers System administrators Figure 1 Cloud service models (from the point of view of the service consumer) Applicationsdesignedtoruninapaas environmentarelikelytobelesscomplexand consumelessresourcesthantheirtraditionally- programmedcounterparts,astheydonotneed tore-implementtheservicesthatareprovidedby theplatform.Asaresult,apaas applicationtakes lesstimetostartupthanapplicationsdeployedon afullsoftwarestack.Thesimplifiednatureofpaas applicationsbringsbenefitsintermsofscalability, especiallyforthosethatarestateless. Designinganapplicationforpaas withloosely- coupledinternalandexternalinterfacesmakesit easiertomanagelifecyclesforthecomponentsof anapplicationandfortheservicestheyuseinan independentmanner.Deployingcomponentsthat arelooselycouplednotonlysimplifiesanupgrade, italsoreducesthecomplexityofvalidatingan upgrade.Combinedwiththefreedomtochoosethe programminglanguageandruntimeenvironment bestsuitedtothetaskathand,loose-coupling enablescomponentstobereplacedatanytimewith adifferentimplementation–eveninadifferent language–whichinturnsupportsthegradual introductionofnewtechnologies. Thepaas frameworkprovidescommonways toexposeandbindtoservices,whichsimplifies thedeploymentofnewservices.Servicegateways andbrokerscanalsoexposeexternalservices,so theycanbeusedbyapplicationsrunninginsideor outsidethepaas environment. Theeaseofintegrationofnewservicesbrought aboutbypaas contributestofasterinnovation,which isoneofthemodel’sprimarybenefits. Benefitsforsystemintegrators Someofthebenefitsthatapplytodevelopersalso applytosystemintegrators.Loosely-coupled servicesandindependentlifecycles,forexample, cansimplifythetestingandupgradeofcomponents, asthesetaskscanbecarriedoutseparately.Andthe commonbindingandserviceexposureframework facilitatestheintegrationofnewservices.

×