 Hello everyone, welcome to join our panel. This panel discussion is the topic is over our intent based networks, requirements, researches, solutions and open source developments. There are four panelists in this panel. I'm Dong Wang from China Telecom and Henry Yu from Huawei, Ke Guang from China Mobile and Ahila from Wipro. In this panel, the first 20 minutes each one of us will give five minutes politicians to show our researches. My topic is research and development of intent based network for next generation networks. And Henry Yu will give his talks about the intent based networking and the closed-loop automation. Ke Guang will give his topic about research on intent network management. And Ahila will give last talk, intent guarantees part for E2E slicing. After the four talks, we will have 10 minutes to discuss about ours. Okay, so first I will give my presentation. The topic is research and development of intent based network for next generation networks. The intent based network is self-driving networks that use the company network control logic and the closed-loop orchestration technicals to automate application intents. Intent based network is an intelligent network which can automatically convert, verify, deploy, configuration and optimize itself to save target network state according to the intent of the operators. And then it can automatically so abnormal in one to ensure the network reliability. What we can see from the figure in the right is a high-level framework of intent based network. First, the user could input the intent and the intent translation where translate it to be the required network parameters. Then it will do the activation works to the physical and the virtual network. Then the network can also monitor the network data and guarantee the user's intents and then feedback to the users. It's a closed-loop of the basic intent based networks. There are lots of academic papers about intent based networks. There are two papers from our partners in the universities. The first paper is from a professor Yang's team. There is a survey on intent-driving networks which publish in actually exercise. There is a survey of the intent based networks. And the two ideas from this paper, they said, in April, in the evening, technicals of IBM are still in the exploring stage. The basic components and the core architectures of intent-based networks are based on different development logic and present different functional characteristics. So it means that the current research are still independent and they are not common standardization words. The second paper is from a professor Peng's team. It's intent-based network for 6G inside and changes. It's all there of the intent-based 6G network architecture. It shows the run network of the 6G network which is based on the intent-based network to support the smart network. There are also many standardization words from most of the standardization communications including ITOT, TSI, 3G, 3G, SA file, TMF, ITF, and CISA. China Telecom and also as companies, our partners, works on all of those standardizations. For example, in ITOT, China Telecom has standardization words on scenario and requirements of intent-based networks for network evolution. The second standardization, which is intent-based network management and orchestration for network sizing in MQ2020, and words and beyond, as I know, is made by China Mobile and China Telecom also covered on these standardizations. In ETSI, there's ENIO 13 is an internet network automatic city is made by China Telecom, and also we cooperate with Hongrui team in Huawei on ETSS and ZSF words. There are three DTP words and the TEMF words, which are read by Huawei and China Mobile. There are also some chapters that have been published in ITF. There's also a standardization read by China Telecom in CISA. From the standardization organizations, all those standardizations are words on the intent-based network, standardization words. However, so far, not all of those standardizations are alignments. As I know that TEMF has tried to alignments the research of autonomous networks, including intent-based network, around all of those standardization words, and also the open source network, open source community, including ONAP. At the end, I will show our words of the open source development in ONAP about intent-based network. This work has been done by all the four partners in ONAP. Based on ONAP, we have provided an electric model of intent-based network. The users could input their intents and they can translate those intents in URI. And then those translated requirements will be used to create a new use case by the use case in SO. And the user's intent will be saved in AI. The close loop management of intent is done in DCAE, which includes two main parts. The first one is intent monitor and insights, and also the intent decision and performing based on the AI models. The performance of the requirements are done as policy and the dominant controllers. In the intent-based network, intent verification is also required. However, in current ONAP, we can't find a platform project to support intent verification and also other network verification words, which could be used digital turns, and this work would be done in the future. Okay, this is my talk. Thanks. I want to talk about how close loop automation may support intent-based networking. We will start with a general introduction on the basic concepts of intent-based networking. Let's take a look at this picture from an IRTF draft. And in this picture, it highlights that the intent-based networking consists of two main functional components, intent fulfillment and intent assurance. For both, we need two control loops as shown in this diagram. The outer loop highlighted in red is the end-to-end intent control loop that requires user interaction with the intent system. In addition, there is an inner loop highlighted in the blue color, which is the classic zero-touch close loop automation. And that inner loop is the focus of our talk for this presentation. This is a former specification of all expectations, including time and story, to give a technique system. Expectations that need to be realized without them. Intent decouples the expectations of the specific implementation details, giving them to allow the producer to exploit alternative options and find optimal solutions. Intent also reduces the degree of manual participation in the network and improves the automation level and operation efficiency. In the same time, the network operation process reduces the network operation first and improves the business growth of operators. And for current situation of internet research, the search on intent is steering a preliminary stage and operators and operators and manufacturers have not reached agreement on internet management. And also local of a constant science on intent interface and internet expression model and local definition of internet information model and internet data mode. For China Mobile, our objective is to build a powerful level autonomous network by 2025. We research on the key technologies, reference implementation and industry standards of intent network management to ensure that the operation of internet network meets the expectations of operators. Hello everyone. I'm Mahila from Wipro. So in this presentation, we'll see the working of end to end network slicing and how it can be associated with the intent is networking. So in order to achieve the end to end network slicing, we have considered a different architecture options. So among these options, option four and one are currently supported within ONAP. In option four, so we will have a CSMF, NSMF functions within ONAP and the NSSMF so kept outside of ONAP that is so it can be replaced by any third party component. So if you see the option one, so in this approach, all the management functions like CSMF, NSMF and NSSMF, all the management functions are kept within ONAP. And so in this way, the network functions layer like the core network functions are realized as cloud native functions and Iran and transport infrastructures are realized with the help of simulators. So these two are the currently supported architecture. To discuss the first one is to introduce the progress and the target of IBM research in ours. The second question is about the cooperation of IBM research and development among the academic standards and open source development. And that's the question is about the how do you think about the contribution of open source development for IBM in generation networks. I'm a research engineer in Huawei and for my work and my team's work, we are mainly focused on the research on how closed loop automation can support intent. As we discussed in the introduction that we are currently seeking to you to use the machine learning technology to enable a better data analysis and the policy generation. And I said we are welcome other participants to join us on this research innovation on the ONAP platform. Okay, good. So, how's your work. Okay. For China Mobile at present to be trapped to the standards and the project of 3GPP, 3GPP, ETSI, TMI for CCSC and RITF and computer and on that last dessert. We have research contents and emphasis of various standards organizations. We have internet related project in ETSI, mainly on internet policy. In the follow up we were paid attention to the points that have not been paid attention to and the parts that have not been improved in the current standard research. In the form of a consensus on general internet management related worker by promoting cross as dual corporations such as a general internet expression model and information model. And our target is to reach agreement with the manufacturer on intent management, including internet processing and related reference points in the expression model into the left side management. Finally, promote the formation of a general industry standards to ensure that the operation of the network meets the expectations of the operators. Okay, thank you. How's about your work. Yeah, yeah, I'm part of center of excellence in the pro. So are most of the contributions are towards the open source projects. Yeah, we are keen to work on the entwined network slicing use case. So our first step is to create a stable version of entwined network slicing, then on the on top of it. Yeah, we are further moving on to this IBM support to this work to this use case. So, yeah, as the next step, we will be focusing more on this IBM. So, like, yeah, in order to integrate both this existing one app use case with this IBM. Okay, thank you. Okay, I'm from China to come and then in our team, we force on the research of national generation words. So for our research on internet based network, most of us on the next generation network, we also call it 6G network. And we think that in 5G network actually in recent about the recent five years, we can't really make sure the automation networks, including the internet based network. The current AI for network is the outside model for the current network operation. It's not the native AI, which we may think it will be make sure in national generation network, national generation network. In our worst, the first stage, we force on the recent network, and we do our research, including online to provide the basic platform and model to make sure the basic architecture and model of internet based network. And in the second stage, we also force on the network or 10 years later, based on the AI native network, how we make sure the autonomous network as well as the internet based network. This is our work in China. Okay, thank you. So the second question, I think we come from Keguang. How do you think about the cooperation with our internet based network among the academic solutions open source and so open source development. Okay. First, a major standard organizations have carried out the internet related standard project approval and research, which are in the initial stage and have their own emphasis. To the limitation of the characteristics of each organization, the definition or intent related worker is limited and not a universe. And third is a tensile cooperation among academic standards, organizations and open source is a conductive to promoting the common internet management consensus in the industry. And putting the cooperation between operators and manufacturers accelerating the formulation of internet related to the standards and enhancing the particular ability of standards. Thank you. Okay, thank you Keguang. How do you think about the cooperation. There is a lot of open source. So if you see the current systems close loop is achieved almost. Yeah, but with the help of IBM, I am expecting like it could be more automated like very less manual intervention so in that perspective. Yeah, it is needed. Yeah. Yeah, this will be achieved. Yeah, so yeah. Okay, thank you. What's your idea. Right, so I definitely favor a collaboration between the standard, standard organizations open source, as well as academic field in all three dimensions. We are trying to bring more collaboration between the standards and the open source implementation or development in the intent area. We are using standards from, for example, I ETF, ZSM, and using that those standards and try to align the standards with our open source implementation on the own app platform. And in the future, we like to also bring the atomic research projects into our intent based closed loop research initiative. And hopefully it's also on the, on the open source basis. Okay, thank you. Thank you. So from my opinion, I think that all of the academic standards and open source development plays an important role in the research and development of intent based network. Firstly, the standardization was most focused on the requirements and the application or industries. So it's really important to make IBM standardization, standardization, standardization. As I mentioned in the previous presentation, the current research on intent based network haven't been salient as the companies and also the academic research for sound they invent and framework and ideas. So, I think the first way, and also the first, the first step to do is to make IBM standard standard. The university of the academic research in university plays the important role on most of the AI model, the research and training our AI models. So if our AI is still a step in the third step research, and it will be able to apply in the future. So currently the university plays have their, they do better words on the AI training, which is much better than the words in industry. So I think that in the current research, the university could play a more important way on the AI model for intent based network and also they provide some new ideas for intent based network. And also I think that it's important to provide open source and open source platform for the academic and also industry to research and training the IBM models. So it's important, it's really important to have open source platform like all nine. I think that for based on ours, we will provide open platform for both academic universities and industries to training and research about the intent based network based on all nine. Okay, about the last question, I think we come from so how do you send open source, the contribution or open source developments, including online for the intent based network research in the future. Yeah, we plan to use IBM in hygiene network slicing for the two scenarios like the slice creation or like slice instantiation so that we want to achieve it with IBM. The other scenario is like the closed loop that I have explained in my deck. So in those two areas, yeah, we are like we are interested to work with IBM, like we are happy to collaborate with your team and the other open source contributors to achieve these two aspects with IBM. Okay, thank you. And how about your idea. Right. I think I, I could be contribute to the, to the IBM in terms of the standardization of the interfaces and models, especially the southbound, sorry, the northbound interfaces of the the transport management domain. I think it's quite important for the interface is to be standardized such that the components can interact each other in the end to end in the end to end orchestration. The, the, the models, the, there, as we see in the intent, intent, a basic system, the, both the intent fulfillment as well as intent assurance requires a multi step of the model translation from the most abstract models, or the definitions of the user intent to the least abstract models that is the configuration of the, of the networks, and the automation of such a multi layer model translation is quite important. And I think that we need to put more innovation in that front. That it will be our next step of our research focus. Okay, thank you, Henry. How's your opinion. Okay, in my opinion, open source is a good way of experimental exploration and process organizational collaborations through development practice, we can clarify the limitations and the depth of current standards, and then reverse the primitive improvement and formulation that are relevant to the standards. And we are also studying the open source implementation scammer based on nap. We are researching the introduction of a generator into the process and processing system in nap to realize the several into the functions, according to our standards, research. Okay, thank you. Thank you. So, as I mentioned above that open source development is really important for the internet network. Certainly, we can see that the IBM first it will be true in the future networks, it will be used in the future network. However, I think that we still need maybe three to five years to make more to do more words to on the IBM. So I think, based on kind of work, the open source development, including online is the best way for all the companies and academics to test and development to the new ideas of about IBM. So I think that hopefully, hopefully, based on our words in online, we can make IBM in advance for the applications in the last several years. And I think, so, thank you all for.