Research on Intelligent Logic Design and Application of Campus MMTC Scene Based on 5G Slicing Technology

Lili Tong,Chen Zhang,Ronghuai Huang

1 Faculty of Education Beijing Normal University,Beijing 100875,China

2 Wuhan University of Technology,Wuhan 430070,China

Abstract:A consensus has been reached that 5G network slicing technology is becoming increasingly important in EMBB,MMTC and URLLC.However,this technology still has a long way to go in education industry.By selecting four types of demands in campus MMTC scenarios as specific application references,this research explores the architectural relationship between network slicing layer,intelligent terminal layer and application capability layer.Besides,emphasis has been put on the influences that virtual logic network design technology,isolation technology and standard technology has on application experiences.Under such goal,a two-dimensional route has been proposed to deploy network slicing technology based on the concept of full life cycle and the synergistic effect that AI algorithms has on slicing technology.This research is expected to nurture slicing technology,not only its technology chain,but more importantly,facilitate its value chain and large-scale vertical industry applications.

Keywords:5G slicing technology;MMTC;AI

China experienced four stages in the global mobile communication technology reform:1G blank,2G following,3G breakthrough and 4G simultaneous development.Now China has catch up and strive to pioneer 5G.The first 4 generations focused on the needs of telephone and Internet access while 5G was born for scenario applications.According to scholars,ITU(International Telecommunication Union)includes three 5G scenarios[1]:enhanced mobile broadband(EMBB),massive machine type communication(MMTC)as well as ultra-reliable low-latency communication(URLLC).Typical application scenarios of EMBB include high-speed rail transit wireless communications,such as densely populated high-speed rails and frequent switching requirements across signal towers.Typical scenarios of MMTC include smart cities and shared bicycles that use sensors and data collection as a pathway.Typical URLLC Application scenarios include applications that are extremely sensitive to delay and reliability such as autonomous driving,telemedicine etc.[2].

5G network slicing enjoys great advantages in high throughput,connection density,spectrum efficiency and traffic capacity etc.Such character facilitates network operators’flexibility,adjustability and user experience,therefore playing a greater role in MMTC scenarios.This thesis focuses on campus MMTC scene logic design and application research based on 5G slicing technology.By discovering key technology paths and building simulation application space with smaller granularity,this research expects more applications that slicing technology contributes to education industry.

I.EVOLUTION OF 5G SLICING TECHNOLOGY

China contributed fundamental support for 5G technology including standardization,network slicing structure,user data fusion architecture,functional softwarization,C/U separation,edge computing etc.[3].This technology provides isolated network environments for different application scenarios.Composed of “wireless sub-slices,bearing sub-slices,and core network sub-slices”,it is able to customize functions and features for application scenarios according to their requirements.In short,this technology cut one physical network into multiple virtual end-to-end networks.Networks,equipment,access,transmission and core network are logically independent between each end-to-end network[4].On one hand,malfunction in one network won’t bother the other; on the other hand,different functional characteristics meet different needs and that’s exactly how 5G outstands(Figure 1).

Figure 1.Schematic diagram of 5G network slice networking architecture.

Figure 2.Campus MMTC scene architecture design based on 5G slicing technology and AI.

Figure 3.Schematic diagram of some parameters.

Figure 4.Schematic diagram of deployment and advance route.

5G network slicing technology in China experienced five main stages and they are as follows:

•February 2013:China established the IMT-2020(5G)research team,with domestic businesses and colleges included,to carry out 5G research and development.

•2016:China initiated 5G technology research and experiment.

•2018:5G pilot projects are carried out in multiple zones.May 17th,known as the World Telecommunications and Information Society Day,the Ministry of Industry and Information Technology and SASAC jointly issued the “Implementation Opinions on the 2018 Special Action on Deepening the Advancement of Network Speed and Fee Reduction and Accelerating the Cultivation of New Drivers of Economic Development”.This Opinion suggested fast-speed development in 5G technology industry and integrated development of 5G and vertical industries,so as to prepare for 5G’s future large-scale networking and application.June 2018 witnessed the achievement of network slicing’s 1st-stage standard.

•2019:5G is preparing for marketization in 2019.

•2020:5G is fully commercialized and China,in accordance with scheduled timetable,pioneered in the development of 5G technology[5].

In the following three years from 2018,China completed 1st stage standard of 5G network slicing technology.Now on its way to scaled application,5G faces three following fundamental challenges except for the global COVID-19 pandemic.Firstly,the high cost of standalone(SA)dissipated motivation to conduct technological innovation.Secondly,the MMTC(largescale machine communication)still face insufficient virtual logic network design and value system after technology reform is yet being building.Thirdly,under fierce global competition on communication technology,China is supposed to integrate 5G and AI as soon as possible in order to enhance network performance and intelligence[6].

II.ANALYSIS ON CAMPUS MMTC SCENARIO DEMAND

1G to 4G technology are born from industrial research institutions,so they are mainly designed to meet industrial demands.However,with 5G technology’s concept of IoE,5G network slicing technology is supposed to serve multi-scenario integration and application customization.In education industry,there are rich ecological environment with multi-scenario communication application demands and there are typical application scenarios at the campus level,such as high-definition remote interactive teaching,AR/VR immersive teaching,holographic classroom,remote supervision,campus high-definition video security,and flat-panel classrooms.To advance 5G technology innovation and further application in education industry,China should form a collaborative research mechanism and integrate the actual trend of intelligent education into 5G’s development.In this way,5G network slicing technology will be applied into education industry and enjoy lower development costs as soon as possible,and,together with AI,assist the construction of intelligent campus scenarios.

Among several campus communication scenarios,this research focuses on the specific analysis on large-scale machine communication MMTC demands through sensors and data collection.As early as 2013,experts in education equipment system industry put forward the idea of building an Android tablet network classroom for middle school teaching.They also highlighted that hyperlinked knowledge structure in tablet network classrooms contribute to personalized learning.If every teacher and student is equipped with tablet computer,they can conduct intelligent homework correction and academic analysis more efficiently[7].

Interactive smart tablet computers in higher educational institutions represent another type of demands for device communication:interactive smart tablet classrooms with electronic products such as TVs,computers,electronic whiteboards,projectors,and speakers as a collaborative environment[8].It integrates many functions and applications such as courseware display,QR code scanning,interactive teaching software,synchronized screen recording and recording,subject tools,lesson preparation function,resource search,classroom hotspot sharing,and courseware on the cloud and this won wide recognition from teachers and students.With the in-depth application of the Internet of Things technology,smart campus has higher expectations to data collection.Researchers propose dual-active data collection systems in multiple areas such as campus smart access control management,smart course management,smart library management etc.With this two data center operating at the same time,one works as a backup for the other.They can also transfer applications from one to the other with ease based on cloud computing,virtualization,and high-speed network technology[9].

In massive connection scenarios,smart terminals are numerous and widely distributed,therefore device density is pretty high in few zones.5G slicing technology,together with edge computing,provides connectivity and intelligence to campus terminal equipment,realizing full-scene,omni-directional,and all-

weather monitoring and management.Campuses are therefore ensured to be green,safe and reliable[10].Based on this idea,this research is designed under 4 demand bases:contactless intelligent access control,cross-screen sharing system,Flat-panel network classroom and intelligent physical testing platform.The scenario application is described as follows:

1.Contactless intelligent access control

The contactless intelligent access control manages the passage of people and vehicles on campus and the access control system supports response to alarm and video monitoring at the same time.

2.Cross-screen sharing system

The multi-screen sharing system performs interconnection management and content sharing for multiple devices distributed in various office environments:teaching with management equipment,teaching with research conference computers so as to promote the efficiency of teaching and administration.

3.Flat-panel network classroom

The flat-panel network classroom serves subject teaching.Most of its hardware are IPAD,and the software is composed of teacher and student terminals.By facilitating the efficiency of teachers’reading and correction work,this system provides personalized guidance in real time.Besides,teachers can conduct content interaction with student terminals as well.It also facilitates content

interaction between student terminals.

4.Intelligent physical testing platform

The platform mainly carries the teaching content of physical education,such as intelligent rope skipper.Students can use different rope skipping methods after entering student number.The counting content is transmitted back to the physical education teaching and research group in real time for all kinds of analysis and making followup teaching plans.

The use of 5G technology’s high bandwidth,low latency,large connection technical characteristics,slicing technology,and edge computing technology can promote the construction of smart campus scenarios.contactless intelligent access control can use cameras and sensors to monitor and analyze campus safety behaviors,and provide safety assurance services for students;the multi-screen sharing system realizes the intercommunication and sharing of resources,and enables teachers from multiple locations to teach simultaneously to promote education fairness;The flatpanel network classroom can enable teachers to understand the learning status of students in real time,accurately teach,and achieve face-to-face personalized teaching; the intelligent physics testing platform can test students’physics knowledge mastery,and personalize exercises according to students’ answering conditions to help students check Missing vacancies,the above four scenarios can realize intelligent teaching environment,individualized talent training,diversified teaching evaluation,and scientific school management.

III.ARCHITECTURE DESIGN AND KEY TECHNOLOGY ANALYSIS OF CAMPUS MMTC SCENE BASED ON 5G SLICING AND AI FUSION

Compared with traditional QoS,network slicing technology can provide differentiated service quality based on different services and can also provide differentiated comprehensive user experience for different users for the same service.The network can allocate logically or physically isolated network resources for these services with different communication needs and provide a complete end-to-end virtual network for a given user[11].At the same time,combined with the development trend of AI,integration of enabling campus MMTC scene.Different AI capabilities are introduced into the network element layer of the infrastructure and different layers of the slice management system,and a unified AI-based intelligent slice management architecture is proposed[12].Data acquisition layer:The data acquisition layer is mainly all kinds of intelligent terminals for human and machine acquisition,transmission and preliminary analysis.With the help of 5G communication network and AI algorithm strategies,intelligent terminals can gather process and result data such as teachers’shared resources,students’behavioral preferences and learning styles for teaching lines.The campus management line collects the scene data such as book borrowing,robot food delivery,school security and all-weather monitoring data,etc.,so as to realize the interconnection and communication between people,people and things,and things and things,so that the smart campus can achieve intelligent perception,intelligent management,intelligent service and other purposes[13].

Network slicing layer:The core of network slicing lies in the virtualization of network functions.The hardware environment is deployed uniformly through the separation of software and hardware,and the software is undertaken by different network functions NF.The most basic composition of a network slice consists of three parts:wireless sub-slice(to complete access),carrier sub-slice(end-to-end connection)and core sub-slice(information processing,routing,etc.).Edge computing in Figure 2 is a node with caching and computing processing capacity,which is closely connected with mobile devices,sensors and users to reduce data transmission delay[14].

The three major applications brought by 5G,EMBB,MMTC and URLLC,can realize flexible service assembly through secondary sub-slicing.The four typical applications in the MMTC scene selected in this paper,“contactless intelligent entrance guard,cross-screen sharing system,flat-panel network classroom,and intelligent body measurement platform”,are more in-depth third-level sectioning applications under MMTC section.In addition,as more and more intelligent terminals in the data collection layer will be integrated with the application of artificial intelligence algorithms,the core network information processing and edge computing in the network slicing layer also need to be promoted together with AI as soon as possible,so make intelligent data and intelligent network complex as a whole,network switching and data analysis more in line with the application needs of the intelligent era.

Information platform layer:the information platform layer is a preliminary capability formed on the basis of low-level data collection and network interconnection,including AI big data platform,system log database,systematic student digital portrait,intelligent security monitoring platform,etc.AI big data platform and system log database are normalized process data formed with the acquisition capability of intelligent terminals and the support of 5G+AI,which serve the basic analysis needs of various businesses of the school.Systematic student digital portrait and intelligent security monitoring platform model individual users through data mining technology and learning analysis technology,and through comprehensive analysis of process data aggregation result data,intelligent security monitoring platform covers data extraction,operation and maintenance management of people,assets and network[15].

Intelligent application layer:Intelligent application layer is the scene space based on 5G slicing technology and AI intelligent modules.The intelligent attendance check,network teaching and research,tablet teaching,intelligent rope skipping and other applications corresponding to scenarios such as contactless intelligent access control,cross-screen sharing system,network classroom,intelligent body measuring platform and so on cited in this paper are all components of them.Scene in-depth analysis and virtual logic design are important foundations for the application of slicing technology.In the future,more comprehensive scene design and logic networking,including teaching,education,learning,evaluation,management and public service,will also be included[16].

The AI-based intelligent slice management architecture can make 5G networks more intelligent.At the data collection layer,smart terminals collect and preprocess the user’s voice,video,text and other information through 5G communication networks and AI algorithm strategies; at the network slicing layer,information processing and edge computing are combined with AI to make data analysis more accurate and scientific;at the information platform layer,comprehensively analyze user data through AI big data platform,system log database,etc.,and build machine learning models; at the intelligent application layer,according to the learning model,the corresponding business scenarios are output for users use.It can be seen that the integration of 5G slicing technology and AI can provide users with more accurate services.

On the basis of architecture design,the key technologies and parameters that slice design relies on are the core of ensuring user experience,mainly including virtual logic network design technology,isolation technology,standard technology,etc.[17].

Virtual logic network design is more about the indepth understanding of business scenarios,which is different from the past when 5G enters the scene application.It requires the entry of more demand side of vertical industries and transforms the practical activities into the logic design idea of network slice division.

Slice isolation scheme mainly realizes the resource isolation and guarantee of Network slices in NR(New Radio)/RAN(Radio Access Network).According to the time delay,reliability and isolation requirements of services,it can be divided into slice-level QoS guarantee,port dynamic reservation,static reservation,etc.[18].

Standard technologies focus on terminal acquisition and transmission.S-NSSAI(Single Network Slice Selection Assistance Information)is used to select and use slice service identification Information and is also the connecting surface through the end-to-end slice process.The URSAP(UE Route Selection Policy)is the core rule for slicing configuration and management of terminals.Traffic Descriptor(TD)is a key attribute to realize the diversification and customization of slicing services.TD provides different business granularity,and the appropriate TD should be flexibly selected according to diversified business requirements(pointline-tube)[19].

With the support of virtual logic network design technology,isolation technology,standard technology,etc.,end-to-end parameters of 5G network slice will be the most intuitive factor affecting customer experience(some parameters are shown in Figure 3),which should be paid attention to and fully tested in the design[20].

The use of a network slice composed of three parts:wireless sub-slice,bearer sub-slice,and core network sub-slice can provide different communication services and network capabilities for different types of 5G scenarios.Network slicing has the following characteristics:(1)Multiple network slices can be run on a unified network infrastructure,reducing the cost of network construction;(2)Each slice is logically isolated without affecting each other;(3)It can be based on the scenario Customized on demand,supporting dynamic scheduling of slice resources based on live network real-time monitoring data;(4)End-to-end,which fully meets the characteristics of differentiated network requirements.

IV.SUGGESTIONS ON DEPLOYMENT DESIGN AND PROMOTION

The campus MMTC scene architecture design drawing based on 5G slicing technology and AI gives the main composition of four levels from bottom to top.Among them,the intelligent devices of the data acquisition layer are deployed in each physical environment of the campus according to the requirements of application scenarios,while the information platform layer and the intelligent application layer are the external output of service capabilities.Therefore,the network slice layer becomes the most important deployment link after the architecture blueprint is formed.

Based on a large number of practical activities in the field of telecom operators and various supply chain entities[21],this paper suggests that the deployment of network slice layer should be carried out in a twodimensional way:the vertical dimension is guided by the development idea of project management,following the four processes of “target-key node taskdocument”;The horizontal dimension takes the generation process of slice opening as the order,and follows six processes of“planning stage,demand stage,design stage,ordering stage,operation and maintenance stage and optimization stage”,as shown in Figure 4.

V.CONCLUSION

In this paper,based on 5G slice technology campus MMTC scenario design and application research on intelligent logic as the theme,based on the principle of 5G slice technology in-depth scenarios[22],into the AI collaborative application consideration,from the campus MMTC scene architecture design,the important section of parameter set,deployment,and promote the route three levels,the 5G slice technology application in the campus MMTC scene from the idea into practice.

The main innovative contents of this paper are as follows:

1.Deep integration of 5G slicing technology with MMTC application scenarios on campus.

2.The important effect of virtual logic network design technology,isolation technology and standard technology on the application of slicing technology is proposed.

3.Put forward the specific deployment route in the two-dimensional way of slicing the whole life cycle management process and project management process.

The next research direction will follow the typical applications of contactless smart access control,multiscreen sharing system,flat-panel network classroom,intelligent physical testing platform mentioned in the article,and carry out actual selection,network virtual logic design and environment in the construction of smart campus.The links of deployment,testing and optimization are carried out in detail[23],and the technology chain of 5G slicing technology in the education industry is discovered.At the same time,all relevant schools and companies need to actively explore,accumulate experience,apply technology to practice,and continuously explore its deep value; realize the transformation from the technology chain to the value chain; refine the collaborative research and development mechanism for the integration and promotion of 5G slicing technology and AI education applications,and find a way to solve the three fundamental problems of the large-scale application of 5G network slicing technology.

ACKNOWLEDGEMENT

This work was supported by a grant from the 2020 joint annual project of the Ministry of Education of China and China Mobile:“Research on Educational Experiments with AI”(No.MCM20200404)