 Mr. McBree wish please great. Thank you very much for coming The presentation today really focuses on What we think is a very key use case for? telco NFV and by extension It is important to have a robust open stack implementation as The virtual infrastructure manager to enable this Application I mean we think about where NFV is today and Really over the last two three years. We've seen a lot of Testing of virtualization in a lab. We've seen a lot of proof of concepts, but we haven't seen a lot of operationalization We haven't seen a lot of commercial Implementations so what we're going to discuss today is how to get from the lab environment To a commercially ready NFV solution and in particular around virtualized customer premises equipment so If you look at what the research is telling us It is saying that virtualized CPE is a top-use case in NFV for the service provider community and I think this makes actually a lot of sense, right? If you think about the top-tier operators today their main challenge of course is driving revenue growth and As a result of that they take a look at all of the different markets they serve and They wonder how they can reduce their cost of sales how they can reduce the risk associated with bringing new services to market to generate higher revenue growth and Virtualized CPE kind of lies at the heart of the answer All right, because if you think about for example the small and medium-sized business market Small and medium-sized businesses have largely been neglected by the largest operators in the world Right because the cost of sales to generate revenue from the small and medium-sized business is so high Right, so a lot of our top-tier operator customers really focus in on generating revenue from The enterprise market the very largest businesses Right and as a result the small and medium-sized business market I think and the middle market the medium-sized business market has largely been neglected with advanced features such as IT based services and applications managed services You know various kinds of enhanced connectivity models They have not had as many of those kinds of services delivered to them right and what virtualized CPE does is Help address that need by reducing the cost of sales for the service provider And it opens up a much broader market you know for them to a very large segment of Their customer base Okay, so that's just a little bit of color commentary around what we think is driving the need for Virtualized CPE and it's not just in the business market We're going to talk today about how VCPE is Driving growth in the residential market as well So virtualized CPE enables many new business models So I would argue that you know VCPE is not just about reducing capital costs I think service providers know how to apply pressure To vendors to reduce the overall cost of equipment And I would argue that NFE is not so much about reducing the cost of infrastructure as it is streamlining operations and Creating an environment to drive revenue growth through new and compelling applications that are delivered from the cloud Okay So one of those use cases is cloud VPN. We actually just had a press release with Telefonica on cloud VPN It is a very promising use case And what it enables an enterprise to do what enables a business customer to do is log in through a self-service portal and Change service parameters in real time Connectivity you can bring up connectivity on a particular site a remote site or a branch office or a headquarters site You can change the attributes of that connectivity. You can change the SLAs you can apply different Policies and filters to the traffic itself. You can enhance the connectivity But even more important is the ability to use that self-service portal to enable new applications So if we have for example network functions like session border control firewall intrusion detection all different kinds of security functions Application acceleration when optimization be able to turn those on in real time You know creates a revenue stream for our customers I think this is also behind what's driving the demand for particular use cases in VCPE like cloud VPN on The residential side we see parental control and policy management being very important We see per device and per user-based policies That are enabled through VCPE Okay, you know by taking the configuration away from the device at the premises and moving it into the cloud simplifying the connectivity to the end user and enabling much more robust policy-based Personalization in the cloud, you know, we have the ability for example to have very granular and personalized Services to the end user. So for example, you might have a parental control service where a Child's device is trying to access a certain kind of website or a certain kind of content the virtualized border broadband network gateway instantiates that connectivity Brings that traffic directly to a parental control server You know where a policy is applied and for a particular period of time that child cannot get access to a kind of content or website And if it's a shared device if it's not just one child's device if it's a shared device You know, then the broadband network gateway can direct that traffic directly to a web authentication portal Right where the user has to authenticate him or herself Right, so we have this ability to use self-service portals so much more robust self-service capability to give end users the ability to Personalize services in real time smart home and smart office is another really key application We've done a number of proof-of-concepts in this area, and we're working with partners for example like Intel to Bring an Internet of Things gateway into the VCPE environment You know, so we have the ability You know to aggregate traffic simplify management and optimize the resources of an Internet of Things Gateway in a VCPE environment so now, you know a service provider for example could start to offer a smart home service where sensors and Different devices within the home are connected and managed all through you know a single pane of glass a single portal So another very promising revenue opportunity And then for enhanced troubleshooting You know if you have all of this data coming in through the cloud I mean one of the things that I think is fundamental about NFV for the service provider market is that you know If it's delivered right if it's optimized correctly or engineered correctly you have an embedded analytics function With the VCPE platform So and that analytics, you know can provide customers with certain trending and certain historical data analysis of particular trends of you know, maybe what's happening in the network that's gone wrong or what's You know causing customers to shift from one service to another or maybe to end their terms and You know a lot of that kinds of data a lot of that troubleshooting data can be used you know to reduce churn You know and to anticipate Certain trends Among the customer base and then finally I would say another key application in VCPE is The transition from IPv4 to IPv6, you know a lot of the CPE that's installed today is not IPv6 ready But it need not be ready if we have the ability for example to introduce IPv6 in the cloud Right, so you know we have the ability to have a carrier grade DHCP function in the cloud as part of the VCPE Schematic and you know we have the ability to deliver an IPv6 service or transition from IPv4 to IPv6 You know so all of these these are just you know five use cases There are many more but these five use cases I think are indicative of you know The promise of VCPE and I should actually just back up and let you know that if we look at our pipeline today We've done more than 40 POCs in the service provider market You know an overwhelming majority of the interest that we have right now is in the VCPE application We do have of course a number of different trials in Virtualized Evolve Packet Core in our colleague from Dokomo Just talked about our participation in their trials that are happening right now but You know overall it's really the VCPE that's driving a lot of the interest among service providers today So if there's so much interest and there's so much promise why is commercialization held up I mean if you actually look at where the Where the trials have been and where the deployments have been there's a big gap right there's been a tremendous number of POCs but there have not been a large number of Real commercial deployments and those commercial deployments that have happened have been subscale I would argue you know in particular domains or in particular areas of the network So why is that? So I would put to you there are a few different reasons first of all the overall business case has not really been proven Right. I think we've spent a lot of time With CIOs and CTOs proving technological concepts and we haven't spent nearly as much time with The CMO the chief marketing officer We haven't spent nearly enough time with the chief financial officer to prove out the business case right and You know that business case is about much more as I said earlier than just reducing capex You know if you actually look at a service provider's cost structure You know you compare capex to op-ex You know it's if you're moving the needle in capex you're not exactly proving the overall business case You have to reduce the cost of operations and you have to increase the top line And I think that's where a lot of the business case that have been Presented to the CFO have essentially fallen down so far There's also a lot of confusion about orchestration So if you think about this term Orchestration you know a lot of people think about OpenStack as an orchestration product, right? But in the way of NFP in terms of NFP OpenStack doesn't do all of the orchestration that's needed. Of course it does orchestration of resources in the cloud in the data center but We have according to the Etsy framework, you know We have the NFP orchestration and even on top of NFP orchestration, which is really just resource orchestration from You know an element or network management type capacity into the infrastructure layer. What we really need is a horizontal Service orchestration layer that cuts across physical and virtual network boundaries We'll talk a little bit more about what that means, but there's a lot of confusion about orchestration and you know just how much Service providers need to invest in it and to what extent it should be open source and you know What's practical with respect to? You know open source approaches or not I would also say that another challenge is OSS and BSS readiness so If we're going to actually commercialize something we need to bill for those services We need to assure those services and a lot of the existing Rating and charging and assurance and policy management is not yet real-time in nature It's not yet ready to function in an automated way for an NFP based service So you know there's a lot of evolution that needs to happen in the billing environment and the operational environment apart from just that service orchestration layer that Is absolutely essential to be making NFP a commercial reality and therefore You know driving some of the use of open stack right for NFP and then finally Another great challenge is the integration to brownfield environments a lot of what we've seen so far in terms of commercial deployments have really been in You know greenfield environments You know in patches of networks that are not necessarily encumbered by By legacy infrastructure by existing physical infrastructure, so What we know though about service providers is that they're going to operate Virtual networks alongside Physical networks for at least the next 15 20 years you think about circuit to packet and that migration I mean that took place over 15 20 even 25 years We still have circuit networks out there today Of course all of the traffic is packet most of the infrastructure is now packet optimized. However, we still have some circuit based Switching so that was one architectural shift now. We're moving to another architectural shift with respect to physical to virtual and I think what we know is that we're going to operate in a hybrid environment And what that means is that services are going to need to be provision and assured across multiple types of domains and that's difficult Right, it's far more difficult far more challenging Than turning up a virtual service in a greenfield environment So this is why commercialization and our estimation remains elusive. So What do we need to do? I mean broadly speaking we think the answer is you know in this term called cloudification or Bringing these functions and services You know from a focus on virtualization to a focus on the cloud So as I said in the very beginning we know how to virtualize functions. We know how to take an appliance and virtualize it on an x dot 86 or virtualize it on a general-purpose piece of hardware Right, that's kind of the easy part if I would we spent the last three years Kind of proving out virtualization in a lab. What we need to do now is orchestrate those functions tie an orchestrated function together with the existing OSS so that we can provision services across physical and virtual boundaries We need a much more robust self-service capability that incorporates third-party virtual network functions that incorporates a far wider range of service granularity and we need of course Much more Capable partner management function if you think about what the promise of open networking brings It is that we can bring together a large ecosystem of third-party vendors in an open environment to create compelling new services and If we do that then we need to have a way of figuring out how the money flows right we need a way of taking what's happening in the legal department and Enforcing that in the network environment Right. We need partner management. We need multi-party settlement that takes care of These contractual arrangements and takes care of service level agreements among partners. So for example a VNF partner to an orchestration vendor We need that to happen in real time and we needed to scale Right. We needed to scale in a fundamentally new way. I mean in the virtualized world We're going to have a far larger ecosystem of third-party vendors at play so it begs The requirement of having a much more robust partner management functionality So that's really what it Cloudifying means that's really what to us making NFV cloud ready is all about it's bringing those kinds of functions orchestration Evolved OSS and BSS and partner management and self-service together To bring new services to market To automate the operations behind it so The cloud ready VCP e-solution has the attributes that I just mentioned and quite a few more first of all, I think we do need to embrace this concept of DevOps right we need an easy service design and modification capability with an intuitive Modeling environment right we don't have that today Right we have a highly manual and workflow based waterfall based types of development behind service creation And that fundamentally needs to change if we're going to move faster We need multi-channel self-service today, you know most self-service is available through a PC We need to change that whole dynamic from You know static connections to mobile connections and more dynamic connectivity and different types of devices We need end-to-end VNF onboarding Think about onboarding. This is a critical functionality We need service descriptors Different kinds of catalogs unified together so that we can onboard functions with the kinds of rating and charging and the kinds of Service life cycle management that is going to result in more compelling offers for our customers And that embedded analytics functionality that I talked about earlier You know one of the things that we need in order to Cloudify VCPE in order to automate the functions that we're delivering to customers is a closed loop among service fulfillment and Assurance and service fulfillment and assurance has to be policy based and analytics driven what I mean is so that we can should be able to take our Real-time monitoring run analytics on that data that we see in real-time Have that feed into a policy engine that Guides the orchestration process Right so that we can make changes to our customer services in real-time. We can optimize the customer experience We can't do that if we don't have a real-time visibility into the network We can't do that if we don't have real-time visibility into customer preferences and How those customer preferences ought to change given different? Changes in the network and what they're paying for and what they require under certain conditions So you know closing that loop between service fulfillment and assurance really does take an embedded analytics capability and policy management Incorporated into orchestration end-to-end and then SDN based service chaining What we need there of course is you know dynamic creation of internal communications between servers VMs and VNF elements right, so service chaining is fundamental to creating new and compelling offers for example for the For the SMB market that I was talking about before And we need to be able to direct the customer's traffic to whatever applications they're paying for be it security when optimization application acceleration intrusion to whatever it might be and that only happens with SDN control working in coordination with an overall orchestration capability Right, so VCP E of course is about virtualizing the customer premises equipment But in order to operationalize it in order to make it commercial, you know, these are the things that we also need behind the scenes so Operationalization to us requires expertise across many domains I think a lot of the Attention in the industry a lot of the conversation has really been at the CTO level Around this green area Right here. I don't know if I'm working this correctly, but that green box the smaller green box That's the NFVO. That's the NFV orchestration function right, but Orchestration in and of itself is really just an automated way or a real-time way to configure and control Infrastructure resources What we really need to operationalize is in that purple column, you know the embedded analytics and policy management and in the orange section, you know all of the OSS and BSS functions that aren't normally thought of in orchestration, but are absolutely essential to Operationalizing VC PE these are things like real-time rating and charging and the partner management functionality that I just discussed and there's more There's more but just focusing on those for the interest of time But the point really is is that operationalization is much more than for example That resource orchestration layer You know at NEC and netcracker We've put together an end-to-end virtualized CPE solution that incorporates not just the VNFs the virtualized broadband network gateway and the CGNAT and DHCP functions and AAA and all of those things that you need just to You know create connections and to deliver services, but you know all of the higher layer Operational capabilities that we've been discussing service and NFV orchestration and the evolved OSS and BSS So I'll just kind of wrap up with a couple of different use cases that we have Implemented at NEC and netcracker that kind of embody what I've been discussing today about what it takes to operationalize So in this particular use case This is with a large tier one operator in Europe. This is a residential VCP e application where what they wanted to do was Enable these value-added services on top of the VC P e infrastructure So in that in that cloud up into the left you see things like home security parental control Multimedia storage and IP TV using multicast DLNA I mean these are services that are all available through a customer self-service Portal now and we designed a self-service portal for our customer to enable their end users to self-select these functions in real time and To configure per device and per user policies Right and at the same time, you know, they have maybe four or five different CPEs that you know are certified in their network and at least one of them You know it is going to need to support an IPv6 migration right so as I was saying before, you know, we have the ability with DHCP and our carrier grade net product to move Traffic or migrate traffic from the IPv4 to the IPv6 IP backbone So this particular CPE POC is Complete it was highly successful I think you know the next step for this customer is pre-production trial Right and that's scheduled to take place really over the next, you know, three to six months We also did a cloud VPN solution with a large tier one operator in North America and The idea here was that we needed to very rapidly onboard third-party network functions, you know from you know, the large Router vendors like Cisco and Juniper that were providing virtual routers Virtual firewalls and a variety of different functions embedded in the provider edge Right and the idea was you know to take a lot of those embedded functions Move them into a cloud environment simplify and cost reduce the provider edge and just optimize it for IP routing as opposed to you know service specific or application specific line cards that frankly not only cost a lot of money but take up a lot of space in those chassis for You know for that are really best optimized for router ports And move those functions into a cloud create, you know, this cloud marketplace concept for them The idea is to create a cloud marketplace concept where virtual network functions coexist along with Software as a service and other other cloud-based applications that can be offered up and delivered with network functions To help our customers generate more revenue from a cloud VPN service Yeah, so This way of course the operator doesn't need to manually configure every CPE device You know with statically provisioned parameters, right Everything it can be done or much more can be done through customer self-service So that the customer themselves can not only control costs, but make sure that each location needs the functions and the Capabilities required for their VPN service So The customer You know really wanted to focus on you know ordered activation for private cloud offerings You know establishing and enhancing the IP VPN connectivity, you know giving the end user the ability to change connectivity parameters and QoS parameters you know quickly and easily and Incorporating SDN control, you know to enable service chaining Which of course is a requirement if you know some of those customers out there for some sites are going to enable multiple Functions that need to be chained together So that's the presentation. I'll just leave you with a few key takeaways I think you know we think about you know, this is a massive conference and OpenStack has so much diversity and It applies to so much in the in the cloud environment and you know Telco NFV is just one of those areas where OpenStack applies But you know, there is a big question of whether or not, you know NFV will become operationalized and commercialized anytime soon and you know in my estimation it might take a couple more years to really get to Where operators want to be? But you know, we have to look to the applications that can help bring us over The goal line so to speak and virtualize CPE is certainly a top-use case We've really spent the last, you know, three four years testing virtualize CPE in a lab it's really now time to focus on the operational aspects of it and You know to do that, you know operators need that end-to-end service layer orchestration right not just the resource orchestration but total control of the service lifecycle across physical and virtual domains and a modernized BSS in particular Automating the rating and charging engine making it real-time Scaling up the partner management functionality and and there's so much more there that also needs to be done You know we at NEC netcracker with netcrackers heritage and the BSS and OSS and NEC's heritage in the network and the IT Infrastructure, you know, we're bringing together what we believe are the right combination of assets to help our customers move into an operational framework and You know, we're working with a number of partners Red Hat is one We're working with a number of partners to make open-stack carrier grade In order for NFV to be able to scale in order for us to operationalize at scale, you know, we need Carrier grade attributes and carrier grade enhancements to open-stack and You know, we have a long history of working in that area You know, we're an early pioneer in developing for example DPDK alongside Intel and getting that into You know our integration with open-stack with Red Hat for example and we've made Significant strides in enhancing open-stack for the carrier environment and we'll continue to do that. We're continuing to We're continuing our commitment there So I'll leave you with that and you know, if any of you have any questions I'm happy to answer them. Okay Thank you so much for coming. I appreciate it