 Hi, I'm Rani Hybe, the CEO of Networking Edge and Access at Linus Foundation. Today I'm going to talk to you about how the metaverse is going to change the way we design and build our networks. So first of all, let's talk about definitions and what the metaverse actually is. So while there are multiple definitions in existence, it's clear that when we say metaverse, we usually mean a convergence between the virtual and the physical world in some sort of way. And when we think about use cases, there are multiple different use cases. We were of course all familiar with the immersive entertaining entertainment 3D experiences, but the metaverse comes in many different shapes and forms, and they can be categorized into various groups. So taking for example the industrial group of use cases, you can imagine XR enabled workforces where people get remote assistant using VR or AR glasses to perform industrial tasks on the production floor. Things like performing predictive maintenance and remote control of industrial equipment. And when we look at the enterprise use cases, there are things like immersive team collaboration and doing training and simulation with the either AR or VR, and of course enhanced consumer engagement. And then of course there are the consumer related use cases like immersive gaming, but also things like shopping and travel models that are implemented in the virtual world. So when we think about the effects of metaverse on networks, we need to consider all these different types of use cases. This is the metaverse actually mean for network and edge computing. When we look at that technology stack that is required for implementing the metaverse, there are many different layers ranging from the content layer but going all the way down to the infrastructure, where we have highlighted here, the computational meaning how compute power and especially edge computing will be required to implement the metaverse, and also the communication and network and how will they need to adapt for metaverse use cases and when we talk about networks. It's many times these are mobile networks so we need to figure out how 5G and in the near future 6G networks will have to adapt and change to support the metaverse use cases. So the metaverse presents several challenges for the network. So if you think about, for example, the round trip latency that is required to support live single participant interactions, then that has to be according to research no more than 10 milliseconds, which means if you take into account consideration physical consideration like the speed of light that the user, the distance between user and data center where processing is performed cannot exceed something around 600 miles. It also means that because of this constraint, some of the processing or a lot of the processing actually will have to be carried out at edge locations and not on a central cloud. It means that data and computing hungry algorithms will have to run on resources that are constrained in edge locations either constrained by power real estate network bandwidth and so on. So that poses a new type of challenge. And when we think about AR and VR and also hectic experience, traffic has stringent constraints around rate, the reliability and latency. So, in order to prevent disruptions for the end user, we need to be provide a very high quality of reliability and very low latency for this, these experiences. And then when you talk about real time synchronization of the virtual physical words, what happens a lot of times is that requires symmetrical bandwidth capacity, meaning that, unlike most of the existing networking applications which require much more bandwidth in the download channel, but not so much in the upload channel. Those real time synchronizations between the physical and virtual world actually required a lot of bandwidth, going from the end user to the network or what is called the upload, the network and that for the first time poses a need for enhanced symmetrical capacity of the network. And when we look at those many different requirements, they can actually be grouped into certain categories so some groups of requirements stem from the immersive 3D streaming use cases. Some sets of requirements are applicable for multi sensory communication, meaning the visual, audible and hectic types of communication. Of course, there is, there are requirements that are derived from the real time aspects of the metaverse interactions. And there's other sets of requirements that come from this need for seamless physical virtual synchronization, and also the multi dimension collaboration. So if you look at different use cases or services as this table shows here, we can see that those requirements for reliability, latency, data rate, and connection density, meaning how many devices per unit of area. They can be different from use case to use case. So if you look at virtual reality, for example, it can tolerate, maybe something like seven to 15 milliseconds of latency and the expected data rate is around 250 megabits per second, but for tactile interaction. So the latency does need to not exceed one millisecond, although the data rate is slightly lower and maybe limited to something like one megabit. If you go all the way to experiences like hologram education or hologram real time communication, we can see that the data rate or the bandwidth goes to staggering amounts of bandwidth in the area of terabits per second. So how does network and edge computing can be designed and evolve to address those, those use cases. So in a desirable metaverse, which will be decentralized. It may be much like the internet standards made it possible to connect previously isolated communities. The open metaverse will need similar standard to allow metaverse personas and virtual property to persist across all the metaverse platforms. What actually is that we will see a paradigm of cloud edge and end device, meaning that computation or processing for metaverse use cases will be carried out, not in a single location, but over a continuum, starting from the user and device, going through edge locations and all the way to the centralized cloud. For example, local computation on edge devices can be used for least resource consuming tasks. For example, computation required by a physics engine to determine movement and position of an avatar. Then edge servers can be leveraged to perform some of the foreground rendering, which requires less graphical data, but also requires very low latency. And then less time is sensitive processing can still be carried out at central cloud location. Now how is this enabled. What it means is that we will need to have tighter integration between the apps and the network. So due to the stringent performance depends of the metaverse. We cannot just simply throw more bandwidth at the problem and hope for the best. So when we talk about video streaming, for example, in this case, if you just provide the end device or application enough bandwidth, the application doesn't need to interact much with the network and it can stream and render the video stream to the end user. So when we talk about metaverse applications, then just throwing more bandwidth is not enough and we need tighter integration between the application and the network. What does it mean it means that application first of all will need a way to get information from the network so they should be able to ask the network, how much bandwidth is available to me or to the device that I'm running on. Or where is the nearest edge location that can serve me that has the right resources and is close enough to me to provide the necessary latency and performance. In addition, apps will need a way to control the services that they get from the network, doing things like requesting quality on demand that is guaranteed by some service level agreement. And also application need a way to request assignment of edge resources for the applications that they're running. So this on demand secure and controlled exposure of the capabilities through API will pave the way for transforming the operator network into a more of a service enablement platform. This will facilitate the application to network integration, and that will be key to delivering the metaverse era use cases. One project I wanted to mention is the Kamara open source initiatives that deals exactly with this exposure of API's for applications. Modern networks whether it's for G or 5G and in the future 6G are very rich in capabilities and many of these capabilities are exposed through API's. However, these API's tend to be rather complex and not designed for consumption by application or application developers. So there is a need to abstract from the network API's to service API's that can be consumed by applications and to simplify the telco complexity and make these API's easy to consume by application developers in a manner that they are familiar with. There's also a need to provide these API's in a secure and in a manner that follows data privacy and regulatory requirements. And if all that is done correctly, then this will facilitate this desired application to network integration that is so important for many use cases but especially for metaverse use cases. So the Kamara open source project focuses on exactly that on creating that exposure gateway or exposure layer for network services that exist in the telco network and exposing them to the consumers which in this case are the metaverse developers. So what are some of the use cases that are enabled through those API's so our first example is remote surgery with XR. In this example, a remote surgeon is interacting with a 3D model of an organ that is created by an imaging device that is located far away. They can using this XR technology they can manipulate and see this 3D image of the organ being located many miles away from the location of the patient and they can instruct the local team on how to perform the necessary surgical procedures. So for this to work properly, we need to make sure that there is low latency and there are no network disruption so there is a quality of service guaranteed for the duration of this remote surgical procedure. So the quality on demand API's that are available through the Kamara project are enabling this reservation of the necessary network resources making sure that this procedure can be carried out smoothly and without interruption. Interruption. So recently telephonic and Microsoft and apoclar demonstrated how this use case can be performed and carried out. Another use case is holographic calls or real holographic real time communication. So again in this case there is a need to provide, first of all, low latency and high bandwidth but there's also a need to place processing resources close to where the end user is. And that means using edge location but using the right edge location that is closest to the end user and that has all the available resources. So once again using the Kamara API is Deutsche telecom orange and Mitsuko demonstrated how this can be done reliably and how the holographic applications can use the API to reserve the quality and bandwidth that is required for this application and you can see how it dramatically improves the quality of this holographic communication. Another use case is remote maintenance. This comes from the more industrial side of things and not everything is 3d entertainment. So in this case, there's a need to perform maintenance for energy production equipment think of power generating turbine that needs to be maintained by a technician on the field but he needs to hear she needs to be assisted by someone in a remote location. So with the aid of HoloLens in this case as an XR device, those technicians can get all the remote assistance they need but in order to make that work reliably and provide the best experience to the operator there's a need to make sure that network latency and jitter are kept to a minimum. And again, this is done once again by the remote maintenance API, remote maintenance application using the quality and demand API to preserve that necessary low latency and low jitter communication channel for this application. These are some of the families of API's that are available by the Kamara project and over time there will be even more. You can see the quality and demand and edge resource location API's but you can also see several of the other existing families of API's that could be leveraged for metaverse use cases. For example, the carrier building and number of education that are maybe used for creating scenarios of e commerce in the metaverse, while still tapping into physical resources of the of the user in the real is their mobile building count. So we, you can imagine that with these type of API's you can create virtual e commerce metaverse applications that are actually tied to the physical world and the physical monetary resources of the end users so there is a large set of API's that are available and they keep growing and addressing the current and future needs of metaverse applications. Another open source initiatives that is addressing the metaverse networking and edges cases is the a cranial open source project. So a cranial is an open source project that provides blueprints for architectures of edge devices that are addressing use cases. As you can see here in this diagram, the screen project deals again with that continuum of remote edge close to the edge device or end user and more central locations all the way up to a central cloud, and the use cases that are supported by the cranial blueprints are are all over this spectrum of locations and use cases. Of course, metaverse use cases are becoming a big part of what the cranial project is dealing with and what the cranial project is starting to provide solutions for. Here's one example of one of the blueprints that is provided by the cranial project. In this case, this is an edge computing infrastructure designed to support AR or VR processing on edge locations. And as you can see, it starts with a platform for running the actual VR application, but it goes all the way down and provides all the necessary pieces of infrastructure operating systems connectivity that are required to support these applications that can have high demand for networking or processing power. Such a blueprint takes into consideration things like hardware acceleration and how to most efficiently expose it to those applications and provide them with the optimal execution environment for for this type of applications. So we talked about API's and we mentioned edge computing but of course everything has to work together and just recently the communities of the Kamara open source project and the acreno open source community work together and announced that they're joining forces to boost the API integration of edge computing, which with metaverse use cases of course being front and center for these type of integrations. So we not only have specific open source communities that are addressing the demands of the metaverse but we're seeing a cross industry and cross project collaboration that is designed to address the needs of these use cases and provide the necessary solutions. So to conclude, it's important to remember that the metaverse will come in different shapes and sizes for multiple domains multiple industries multiple use cases so we need to keep all of them in mind when we talk about the metaverse. It's also clear that network resources will have to be managed efficiently, if you want to address the demand of the metaverse apps, and that network API's will be crucial to the success of the metaverse and to successful management of these network resources. And also it's obvious that network edge is where most of the heavy lifting of processing of metaverse applications will happen because the edge is kind of the sweet spot between the end devices that are constrained in resources and the cloud that may have all the resources but maybe too far away from the end user. And of course, we need to remember that edge resources should be available for the dynamic consumption by the metaverse workloads and again those the API is that I mentioned will be a key factor in successful assignment of these resources to the metaverse applications. If you're interested in taking part in shaping the future of network and edge for the metaverse world you are more than welcome to join our open source communities that are defining these exciting technologies. So, all the open source communities are working in an open manner which is open to the public everybody is welcome to participate whether it's through regular meetings and calls whether it's through mailing list and communication channels like Slack and Discord. Just to mention a few of the forums that are discussing the metaverse use cases and networking. We have the metaverse, the open metaverse foundation networking interest group that has regular meetings and an active discord channel. We have the Linux foundation networking technical advisory council that meets regularly and discusses emerging technology challenges and how to address them and metaverse of course is a big chunk of that. You may join this, this forum as well. The Linux foundation edge creator project that I mentioned is also open to participation by anyone so you may follow this link and find all the information about the relevant work groups and mailing list. Finally, the Kamara project is also an open source initiative welcoming anybody who wants to learn and contribute. So, welcome to follow this link and see all the different activities of these community. Thank you very much for listening. I hope you learned something about the metaverse and I also hope to see you very soon collaborating in one of our open source projects. Thank you very much.