WO2022007899A1 - Upf选择方法及装置 - Google Patents

Upf选择方法及装置 Download PDF

Info

Publication number
WO2022007899A1
WO2022007899A1 PCT/CN2021/105253 CN2021105253W WO2022007899A1 WO 2022007899 A1 WO2022007899 A1 WO 2022007899A1 CN 2021105253 W CN2021105253 W CN 2021105253W WO 2022007899 A1 WO2022007899 A1 WO 2022007899A1
Authority
WO
WIPO (PCT)
Prior art keywords
mec
computing resource
computing
network element
mecs
Prior art date
Application number
PCT/CN2021/105253
Other languages
English (en)
French (fr)
Inventor
李云飞
余庆华
杨庆昌
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2022007899A1 publication Critical patent/WO2022007899A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46Multiprogramming arrangements
    • G06F9/50Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46Multiprogramming arrangements
    • G06F9/50Allocation of resources, e.g. of the central processing unit [CPU]
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network

Definitions

  • the present application relates to the field of communication technologies, and in particular, to a method and apparatus for selecting a UPF.
  • Multi-Access Edge Computing can deploy applications, content, and some business processing and resource scheduling functions of the core network to the network edge near the access side, thereby improving mobile network speed and reducing latency. and improve connection reliability.
  • UPF User Plane Function
  • ME APP Mobile Edge Application
  • the present application provides a UPF selection method and device, which are used to avoid the problem of service processing failure caused by insufficient computing resources at the same site.
  • an embodiment of the present application provides a method for selecting a UPF, including:
  • the first network element obtains the computing resource information of at least one service processing device of the MEC, and the first network element can select the first network element according to the obtained computing resource information of the service processing device of the MEC. MEC, so that the first network element determines the user plane function UPF corresponding to the first MEC.
  • the first MEC is selected according to the computing resource information of the service processing devices of each MEC, and the first network element determines the UPF corresponding to the first MEC, so that the UPF can forward the service data message to the same site.
  • APP wherein the selected first MEC has enough computing resources for the APP to process at least one service, so that when the APP of the same site processes the service, the service processing failure caused by insufficient computing resources of the same site can be effectively avoided.
  • the computing resource information includes at least one of the following information: computing resource type, total computing resource quantity, availability of computing resource corresponding to computing resource type number of resources.
  • the method further includes:
  • the first network element obtains computing resource requirements of the terminal device, wherein the computing resource requirements include at least one of the following: a second computing resource type, the required quantity of each second computing resource type .
  • the first network element can determine the type of computing resources required by the service currently requested by the terminal device and the number of required types of computing resources by acquiring the computing resource requirements of the terminal device, so as to determine the type of computing resources required by the service currently requested by the terminal device.
  • the selection of UPF is carried out according to the needs of the equipment.
  • the first network element selects the first MEC according to the computing resource information of the service processing apparatuses of the MECs, including:
  • the MEC with the largest number of available resources of the second computing power resource type is selected as the first MEC.
  • the implementation logic is simple, and the computing resources of the selected first MEC are effectively guaranteed to meet the computing resources required by the APP to process services.
  • the first network element selects the first MEC according to the computing resource information of the service processing apparatuses of the MECs, including:
  • For each of the second computing power resource types obtain the MECs in which the number of available resources of the second computing power resource type meets the preset threshold information in each of the MECs, so as to obtain the corresponding second computing power resource type of at least one second MEC;
  • the first MEC can be selected quickly and efficiently, and the preset threshold information can effectively ensure that the number of computing resources of the selected first MEC is sufficient APP for business processing.
  • the first network element selects the first MEC according to the computing resource information of the service processing apparatuses of the MECs, including:
  • the number of the second computing power resource type is 1, according to the ratio of the available resources corresponding to the second computing power resource type of each of the MECs in the remaining MECs, obtain the corresponding MECs. weight, and select the first MEC according to the weight corresponding to each of the MECs; or
  • the intersection of the remaining MECs of the second computing power resource types is taken, and the second computing power resources of the MECs in the intersection set are The ratio of the number of available resources corresponding to the type is obtained, the weight corresponding to each of the MECs in the intersection set is obtained, and the first MEC is selected according to the weight corresponding to each of the MECs.
  • the first MEC is selected from the remaining MECs, which can effectively ensure that the selected first MEC can meet the computing power resources required by the terminal equipment services.
  • the weight of each MEC is obtained according to the ratio of the number of computing power resources of each remaining MEC, so that a balanced selection of each remaining MEC can be realized.
  • the first network element obtains the computing power resource information of the service processing device of at least one MEC, including:
  • the first network element receives the computing resource information of the service processing device of the MEC reported by the MEC according to a preset period.
  • the computing power resource information is reported according to a preset period through the MEC, so that no additional signaling instruction is required.
  • the first network element acquires the computing resource information of the service processing device of at least one MEC, including:
  • the first network element For each of the MECs, the first network element sends a query request to the MEC;
  • the computing power resource information of the service processing device of the MEC can be obtained through a query request, so that the required computing power resource information of the service processing device of the MEC can be obtained according to actual needs.
  • the first network element is an application function AF
  • the first network element obtains the computing resource requirements of the terminal device, including:
  • the AF receives a first request message from the SMF, where the first request message includes a computing resource requirement of the terminal device.
  • the first network element is an AF
  • the AF receives a first request message from the SMF, and the first request message includes a first candidate UPF list;
  • the first network element selects the first MEC according to the computing resource information of the service processing devices of each of the MECs, including:
  • the AF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the first candidate UPF list;
  • the AF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing apparatuses of the MECs corresponding to the UPFs.
  • the first network element is an AF
  • the first network element obtains the computing resource requirements of the terminal device, including:
  • the AF receives a second request message from a terminal device, wherein the second request message includes a computing resource requirement of the terminal device,
  • the second request message is a hypertext transfer protocol domain name system HTTPDNS request, or the second request message is a domain name system DNS query request.
  • the first network element is an AF
  • the AF receives a second request message from a terminal device, where the second request message includes location information of the terminal device;
  • the first network element selects the first MEC according to the computing resource information of the service processing devices of each of the MECs, including:
  • the AF determines, according to the location information of the terminal device, a MEC candidate list corresponding to the location information of the terminal device;
  • the AF selects a first MEC in the MEC candidate list according to the computing resource information of the service processing apparatuses of each MEC in the MEC candidate list.
  • the first network element is a policy control function PCF, wherein the data network access identifier DNAI corresponding to each MEC is configured in the PCF.
  • the first network element obtains the computing resource requirements of the terminal device, including:
  • the PCF receives a third request message from the session management function SMF, wherein the third request message includes a computing resource requirement of the terminal device.
  • the first network element is a PCF
  • the first network element selects the first MEC according to the computing resource information of the service processing devices of each of the MECs, including:
  • the PCF determines, according to the location information of the terminal device, a DNAI candidate list corresponding to the location information of the terminal device;
  • the PCF selects the first MEC according to the computing resource information of the service processing device of the MEC corresponding to each of the DNAIs in the DNAI candidate list;
  • the method also includes:
  • a first DNAI is selected in the DNAI candidate list, and the first DNAI is sent to the SMF.
  • the method further includes:
  • the PCF sends the first DNAI to the SMF, so that the SMF determines a first UPF based on the first DNAI.
  • the first network element is an SMF
  • the first network element obtains the computing resource information of the service processing device of at least one MEC, including:
  • the SMF and the AF synchronize the computing resource information of the service processing device of the at least one MEC obtained by the AF;
  • the SMF and the PCF synchronize the computing resource information of the service processing device of the at least one MEC acquired by the PCF.
  • the first network element is an SMF
  • the first network element selects the first MEC according to the computing resource information of the service processing devices of each of the MECs, including:
  • the SMF determines a second candidate UPF list according to the access location of the terminal device and the capabilities of each UPF;
  • the SMF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the second candidate UPF list;
  • the SMF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing apparatuses of the MECs corresponding to the UPFs.
  • the computing resource requirement is determined according to a PDU session establishment request sent by the terminal device.
  • the PDU session establishment request carries the computing resource requirement of the terminal device
  • the protocol data unit PDU session establishment request carries service characteristics, and there is a first correspondence between the service characteristics and the computing resource requirements of the terminal device; or
  • the PDU session establishment request carries single network slice selection assistance information S-NSSAI, and there is a second correspondence between the S-NSSAI and the computing resource requirements of the terminal device.
  • the computing resource requirement is determined according to a second request message sent by the terminal device.
  • the second request message carries the computing resource requirement of the terminal device.
  • the second request message carries service characteristics, and there is a third correspondence between the service characteristics and the computing resource requirements of the terminal device.
  • the selection of the UPF is implemented based on different first network elements, so that the flexibility and applicability of the selection of the UPF can be effectively improved.
  • an embodiment of the present application provides a UPF selection device, including:
  • an acquisition module used for the first network element to acquire computing resource information of at least one service processing device of the multi-access edge computing MEC;
  • a processing module used for the first network element to select the first MEC according to the obtained computing resource information of the service processing device of the MEC;
  • the processing module is further configured to: determine, by the first network element, a user plane function UPF corresponding to the first MEC.
  • the computing resource information includes at least one of the following information: computing resource type, total computing resource quantity, availability of computing resource corresponding to computing resource type number of resources.
  • the obtaining module is further used for:
  • the first network element obtains computing resource requirements of the terminal device, wherein the computing resource requirements include at least one of the following: a second computing resource type, the required quantity of each second computing resource type .
  • the processing module is specifically used for:
  • the MEC with the largest number of available resources of the second computing power resource type is selected as the first MEC.
  • the processing module is specifically used for:
  • For each of the second computing power resource types obtain the MECs in which the number of available resources of the second computing power resource type meets the preset threshold information in each of the MECs, so as to obtain the corresponding second computing power resource type of at least one second MEC;
  • the processing module is specifically used for:
  • the number of the second computing power resource type is 1, according to the ratio of the available resources corresponding to the second computing power resource type of each of the MECs in the remaining MECs, obtain the corresponding MECs. weight, and select the first MEC according to the weight corresponding to each of the MECs; or
  • the intersection of the remaining MECs of the second computing power resource types is taken, and the second computing power resources of the MECs in the intersection set are The ratio of the number of available resources corresponding to the type is obtained, the weight corresponding to each of the MECs in the intersection set is obtained, and the first MEC is selected according to the weight corresponding to each of the MECs.
  • the obtaining module is specifically used for:
  • the first network element receives the computing resource information of the service processing device of the MEC reported by the MEC according to a preset period.
  • the obtaining module is specifically used for:
  • the first network element For each of the MECs, the first network element sends a query request to the MEC;
  • the first network element is an application function AF
  • the acquisition module is specifically used for:
  • the AF receives a first request message from the SMF, where the first request message includes a computing resource requirement of the terminal device.
  • the first network element is an AF
  • the AF receives a first request message from the SMF, and the first request message includes a first candidate UPF list;
  • the processing module is specifically used for:
  • the AF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the first candidate UPF list;
  • the AF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing apparatuses of the MECs corresponding to the UPFs.
  • the first network element is an AF
  • the acquisition module is specifically used for:
  • the AF receives a second request message from a terminal device, wherein the second request message includes a computing resource requirement of the terminal device,
  • the second request message is a hypertext transfer protocol domain name system HTTPDNS request, or the second request message is a domain name system DNS query request.
  • the first network element is an AF
  • the AF receives a second request message from a terminal device, where the second request message includes location information of the terminal device;
  • the processing module is specifically used for:
  • the AF determines, according to the location information of the terminal device, a MEC candidate list corresponding to the location information of the terminal device;
  • the AF selects a first MEC in the MEC candidate list according to the computing resource information of the service processing apparatuses of each MEC in the MEC candidate list.
  • the first network element is a policy control function PCF, wherein the data network access identifier DNAI corresponding to each MEC is configured in the PCF.
  • the obtaining module is specifically used for:
  • the PCF receives a third request message from the session management function SMF, wherein the third request message includes a computing resource requirement of the terminal device.
  • the first network element is a PCF
  • the processing module is specifically used for:
  • the PCF determines, according to the location information of the terminal device, a DNAI candidate list corresponding to the location information of the terminal device;
  • the PCF selects the first MEC according to the computing resource information of the service processing device of the MEC corresponding to each of the DNAIs in the DNAI candidate list;
  • the processing module is also used for:
  • a first DNAI is selected in the DNAI candidate list, and the first DNAI is sent to the SMF.
  • the processing module is further used for:
  • the PCF sends the first DNAI to the SMF, so that the SMF determines a first UPF based on the first DNAI.
  • the first network element is an SMF
  • the acquisition module is specifically used for:
  • the SMF and the AF synchronize the computing resource information of the service processing device of the at least one MEC obtained by the AF;
  • the SMF and the PCF synchronize the computing resource information of the service processing device of the at least one MEC acquired by the PCF.
  • the first network element is an SMF
  • the processing module is specifically used for:
  • the SMF determines a second candidate UPF list according to the access location of the terminal device and the capabilities of each UPF;
  • the SMF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the second candidate UPF list;
  • the SMF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing apparatuses of the MECs corresponding to the UPFs.
  • the computing resource requirement is determined according to a PDU session establishment request sent by the terminal device.
  • the PDU session establishment request carries the computing resource requirement of the terminal device
  • the protocol data unit PDU session establishment request carries service characteristics, and there is a first correspondence between the service characteristics and the computing resource requirements of the terminal device; or
  • the PDU session establishment request carries single network slice selection assistance information S-NSSAI, and there is a second correspondence between the S-NSSAI and the computing resource requirements of the terminal device.
  • the computing resource requirement is determined according to a second request message sent by the terminal device.
  • the second request message carries the computing resource requirement of the terminal device.
  • the second request message carries service characteristics, and there is a third correspondence between the service characteristics and the computing resource requirements of the terminal device.
  • an embodiment of the present application provides a UPF selection device, which is characterized by comprising a memory and a processor, the memory stores computer program instructions, and the processor executes the computer program instructions to execute the above first aspect and The method of any of the various possible embodiments of the first aspect.
  • an embodiment of the present application provides a computer storage medium, which is characterized in that it includes computer instructions, and when the computer instructions are executed by a processor, the above first aspect and various possible implementations of the first aspect are implemented. any of the methods.
  • Embodiments of the present application provide a method and apparatus for selecting a UPF.
  • the method includes: a first network element acquires computing resource information of at least one service processing apparatus of a multi-access edge computing MEC.
  • the first network element selects the first MEC according to the obtained computing resource information of the service processing device of the MEC.
  • the first network element determines the user plane function UPF corresponding to the first MEC.
  • the first MEC is selected according to the computing resource information of the service processing devices of each MEC, and the first network element determines the UPF corresponding to the first MEC, so that the UPF can forward the service data packet to the APP of the same site, wherein the selected The first MEC has enough computing resources for the APP to process at least one business, so that when the APP of the same site processes the business, the business processing failure caused by insufficient computing resources of the same site can be effectively avoided.
  • FIG. 1 is a schematic diagram of a network architecture to which this application is applicable;
  • Figure 2 is a schematic diagram of a 5G network architecture
  • FIG. 3 is a schematic diagram of a 5G network architecture based on a service-oriented interface in a non-roaming scenario
  • FIG. 4 is a schematic diagram of the architecture of the MEC system provided by the embodiment of the present application.
  • FIG. 5 is a schematic diagram of the architecture of UPF same-site forwarding provided by the embodiment of the present application.
  • Fig. 6 is the process that UE requests PDU session establishment in the prior art
  • FIG. 8 is a schematic diagram 1 of grade division of the MEC provided by the embodiment of the present application.
  • FIG. 9A is a second schematic diagram of grade division of the MEC provided by the embodiment of the present application.
  • FIG. 9B is a schematic diagram three of the grade division of the MEC provided by the embodiment of the present application.
  • FIG. 10 is a schematic flowchart of selecting a UPF by an AF during a PDU session establishment process provided by an embodiment of the present application;
  • FIG. 11A is a schematic flowchart of an AF selecting a UPF after a PDU session establishment process is completed provided by an embodiment of the present application;
  • FIG. 11B is another schematic flowchart of the AF selecting a UPF after the PDU session establishment process is completed, provided by an embodiment of the present application;
  • FIG. 12 is a schematic flowchart of a PCF selecting a UPF according to an embodiment of the present application.
  • FIG. 13 is a schematic flowchart of selecting a UPF by an SMF according to an embodiment of the present application.
  • FIG. 14 is a schematic diagram of an implementation manner of obtaining computing resource information of a service processing device of an MEC in a reporting manner provided by an embodiment of the present application;
  • 15 is a schematic diagram of an implementation manner of acquiring computing resource information of a service processing device of an MEC in a query response manner provided by an embodiment of the present application;
  • TLV 16 is a schematic diagram of a TLV provided by an embodiment of the present application.
  • FIG. 17 is a schematic diagram of a nested TVL provided by an embodiment of the present application.
  • FIG. 19 is a schematic structural diagram of a UPF selection apparatus provided by an embodiment of the application.
  • FIG. 20 is a schematic diagram of a hardware structure of an in-vehicle computing device provided by an embodiment of the present application.
  • Terminal device can be various handheld devices, vehicle-mounted devices, wearable devices, computers, and network devices with communication functions.
  • the handheld device may be a smartphone.
  • the in-vehicle device may be an in-vehicle navigation system.
  • the wearable device can be a smart bracelet.
  • the computer may be a personal digital assistant (PDA) computer, a tablet computer, and a laptop computer.
  • the network devices may be residential gateways (RGs) and switches.
  • Network equipment It can be the base station (Base Transceiver Station, BTS) in the Global System for Mobile Communications (GSM) system or the Code Division Multiple Access (Code Division Multiple Access, CDMA) system, or it can be in the WCDMA system
  • the base station (NodeB, NB) it can also be an evolved base station (Evolutional Node B, eNB or eNodeB) in the LTE system, or a wireless controller in a cloud radio access network (Cloud Radio Access Network, CRAN), or
  • the network device can be a mobile switching center, a relay station, an access point, a vehicle-mounted device, a wearable device, a hub, a switch, a bridge, a router, a network-side device in a 5G network, or a future evolved public land mobile network (Public Land Mobile Network). Network, PLMN) network equipment and so on.
  • PLMN Public Land Mobile Network
  • S-NSSAI Single Network Slice Selection Assistance Information
  • S-NSSAI refers to identifying a network slice, where S-NSSAI includes slice/service type (Slice/ServiceType, SST) and slice difference (Slice Differentiator, SD) two parts:
  • SST refers to the expected network slicing behavior in terms of features and services.
  • the standard value range of SST is 1, 2, and 3.
  • the value 1 represents eMBB
  • 2 represents URLLC
  • 3 represents MIoT (Massive Internet of Things).
  • SD is optional information used to supplement SST to distinguish multiple network slices of the same slice/service type. Among them, SD can be used to identify the specific business to be performed by the user.
  • NSSAI Network Slice Selection Assistance Information
  • 5G networks include Requested NSSAI, Allowed NSSAI, and Configured NSSAI.
  • the Requested NSSAI is the NSSAI that the UE expects to use, and the UE provides it to the network side during the registration process.
  • Allowed NSSAI is provided by the serving PLMN to the UE during registration and other processes, indicating the S-NSSAI value that the UE can use in the current registration area of the serving PLMN, and the UE saves it locally.
  • Configured NSSAI is an NSSAI applicable to one or more PLMNs.
  • AMF sends it to the UE in messages such as registration acceptance or configuration update command, and the UE saves it locally.
  • DNAI The data network access identifier (DN Access Identifier, DNAI) can identify the deployment location of the APP.
  • the method of the present application can be applied in the 5th generation mobile networks (5th generation mobile networks or 5th generation wireless systems, 5G) system, and can also be applied in the evolved packet system (Evolved Packet System, EPS), first with reference to FIG. 1 - Figure 3 illustrates the system and architecture of the present application.
  • 5G 5th generation mobile networks or 5th generation wireless systems, 5G
  • EPS evolved Packet System
  • FIG. 1 is a schematic diagram of a network architecture to which this application applies.
  • the network architecture includes a user equipment (User Equipment, UE), an access network (access network, AN), a core network, and a data network (Data network).
  • UE User Equipment
  • AN access network
  • AN access network
  • Data network Data network
  • Network, DN data network
  • the access network device is mainly used to realize functions such as wireless physical layer functions, resource scheduling and wireless resource management, wireless access control and mobility management
  • the core network equipment may include management equipment and gateway equipment
  • the management equipment is mainly used for terminals Device registration, security authentication, mobility management, and location management of devices, etc.
  • Gateway devices are mainly used to establish a channel with terminal devices, and forward data packets between terminal devices and external data networks on the channel;
  • data networks can include network equipment (such as servers, routers, etc.), the data network is mainly used to provide a variety of data business services for terminal equipment.
  • FIG. 1 is only an exemplary architecture diagram. In addition to the functional units shown in FIG. 1 , the network architecture may further include other functional units, which are not limited in this embodiment of the present invention.
  • the network architecture may be a 5G network architecture
  • FIG. 2 is a schematic diagram of the 5G network architecture.
  • the 5G system is also called a new wireless communication system, a new access technology (New Radio, NR) or a next-generation mobile communication system.
  • New Radio NR
  • the access network in the 5G system can be a radio access network (R)AN, and the (R)AN device in the 5G system can be composed of multiple 5G-(R)AN nodes.
  • the R)AN node may include: a non-3GPP access network such as an access point (access point, AP) of a WiFi network, a next-generation base station (which may be collectively referred to as a new-generation radio access network node (NG-RAN node), wherein, Next-generation base stations include new air interface base stations (NR nodeB, gNB), new generation evolved base station (NG-eNB), central unit (central unit, CU) and distributed unit (distributed unit, DU) separate form of gNB, etc.), Send and receive point (transmission receive point, TRP), transmission point (transmission point, TP) or other nodes.
  • a non-3GPP access network such as an access point (access point, AP) of a WiFi network
  • NG-RAN node next-generation radio access network
  • the 5G core network (5G core/new generation core, 5GC/NGC) includes access and mobility management function (Access and Mobility Management Function, AMF) network elements, session management function (Session Management Function, SMF) ) network element, User Plane Function (UPF) network element, Authentication Server Function (AUSF) network element, Policy Control Function (PCF) network element, Application Function (Application Function, AF) network element, unified data management function (unified data management, UDM) network element, network slice selection function (Network Slice Selection Function, NSSF) network element and other functional units.
  • AMF Access and Mobility Management Function
  • SMF Session Management Function
  • UPF User Plane Function
  • AUSF Authentication Server Function
  • PCF Policy Control Function
  • AF Application Function
  • UDM Network Slice Selection Function
  • NSSF Network Slice Selection Function
  • the AMF network element is mainly responsible for services such as mobility management and access management.
  • the SMF network element is mainly responsible for session management, UE address management and allocation, dynamic host configuration protocol function, and user plane function selection and control.
  • UPF is mainly responsible for external connection to the data network (DN) and data packet routing and forwarding on the user plane, packet filtering, and performing quality of service (QoS) control related functions.
  • AUSF is mainly responsible for the authentication function of terminal equipment.
  • the PCF network element is mainly responsible for providing a unified policy framework for network behavior management, providing policy rules for control plane functions, and acquiring registration information related to policy decisions. It should be noted that these functional units can work independently, or can be combined to implement certain control functions, such as access control and mobility management functions such as access authentication, security encryption, location registration, etc. Session management functions such as establishment, release, and modification of plane transmission paths.
  • the functional units in 5GC can communicate through the next generation network (NG) interface.
  • the UE can transmit control plane messages with the AMF network element through the NG interface 1 (N1 for short), and the RAN equipment can communicate through the NG interface.
  • Interface 3 (N3 for short) establishes a user plane data transmission channel with the UPF
  • AN/RAN equipment can establish a control plane signaling connection with the AMF network element through NG interface 2 (N2 for short)
  • the UPF can communicate with the AMF through NG interface 4 (N4 for short).
  • FIG. 2 is only an exemplary architecture diagram, in addition to the functional units shown in FIG. 2 , the network architecture may further include other functional units.
  • the network architecture shown in FIG. 2 is a reference point-based network architecture, and the network architecture is a network architecture in a non-roaming scenario.
  • the method of the present application can also be applied in a roaming scenario, and the network architecture is not limited to a reference point-based network. architecture, or a network architecture based on service-oriented interfaces.
  • Figure 3 is a schematic diagram of a 5G network architecture based on a servitized interface in a non-roaming scenario. As shown in Figure 3, in a scenario based on a servitized interface, the core network of the 5G system also includes NEF and NRF network elements.
  • some network elements in the 5G core network are connected through a bus, as shown in Figure 3, AUSF network elements, AMF network elements, SMF network elements, AF network elements, UDM, PCF network elements, network elements Storage function (Network Repository Function, NRF) network element, network exposure function (Network Exposure Function, NEF) network element and NSSF network element are interconnected through a bus, and the network elements are interconnected through the bus, using a service interface, such as AUSF network element is connected to the bus through Nausf interface, AMF network element is connected to the bus through Namf interface, SMF network element is connected to the bus through Nsmf interface, AF network element is connected to the bus through NAF network element interface, UDM network element is connected to the bus through Nudm interface The interface is connected to the bus, the PCF network element is connected to the bus through the NPCF network element interface, the NRF is connected to the bus through the Nnrf interface, the NEF is connected to the bus through
  • 5G includes three major application scenarios: Enhanced Mobile Broadband (eMBB), Massive Machine Type of Communication (mMTC) and Ultra-reliable and Low Latency Communications (uRLLC) ).
  • eMBB Enhanced Mobile Broadband
  • mMTC Massive Machine Type of Communication
  • uRLLC Ultra-reliable and Low Latency Communications
  • eMBB focuses on services with extremely high bandwidth requirements, such as high-definition video, virtual reality (VR) and augmented reality (AR), etc.
  • mMTC focuses on services with high connection density requirements, such as smart cities , smart agriculture, smart home, etc.
  • uRLLC focuses on services that are extremely sensitive to latency, such as autonomous driving, industrial control, telemedicine, etc.
  • Multi-Access Edge Computing is an open platform that integrates network, computing, storage, and application core capabilities at the edge of the network near the source of people, things or data. MEC can provide edge intelligent services nearby. Meet the key requirements of industry digitalization in terms of agile connection, real-time business, data optimization, and application intelligence.
  • 5G protocol modules can be flexibly invoked according to business needs, providing technical standards for building edge networks, so that MEC can be flexibly deployed in wireless access cloud, edge cloud or aggregation on demand and in different scenarios.
  • cloud can provide the following values: through local offloading of high-bandwidth services such as 4K/8K, VR/AR, etc., reduce the occupation of the core network and backbone transmission network, and effectively improve the utilization rate of the operator's network; In the future, operators’ networks will effectively support future delay-sensitive services (Internet of Vehicles, remote control, etc.) and services that require large computing and high processing capabilities (video surveillance and analysis, etc.).
  • FIG. 4 is a schematic diagram of the architecture of the MEC system provided by the embodiment of the present application, as shown in FIG. 4 :
  • network equipment is deployed in the wireless access network, and the network equipment can interact with the terminal equipment.
  • the MEC is deployed between the radio access network and the core network, wherein the MEC includes a UPF and a mobile edge application (Mobile Edge Application, ME APP).
  • the UPF is responsible for distributing and diverting the traffic of the edge network to the MEC service system.
  • ME APP can be a third-party application developed by a third-party developer, such as VR/AR application, etc.
  • the computing resources of the MEC server may include but are not limited to at least one of the following: Neural-network Processing Unit (NPU) resources, Central Processing Unit (Central Processing Unit, CPU) ) resources, graphics processor (Graphics Processing Unit, GPU) resources.
  • NPU Neural-network Processing Unit
  • CPU Central Processing Unit
  • GPU Graphics Processing Unit
  • the processing of artificial intelligence (Artificial Intelligence, AI) services should use NPU resources
  • the processing of computing services should use CPU resources
  • the processing of image rendering services should use GPU resources.
  • the entity of the MEC introduced in FIG. 4 may be an MEC server, where the MEC server is a server on which the MEC platform is deployed and managed by the MEC platform. Also, the MEC server can be connected to cloud data centers and other networks, such as corporate networks. Therefore, the MEC server utilizes the wireless access network to provide services and cloud computing functions for the terminal nearby.
  • the UPF is responsible for forwarding the service messages of the terminal equipment.
  • the UPF can forward the service messages of the terminal equipment to the APPs of other sites, that is to say, it is forwarded across MEC sites, for example, as shown in Figure 4
  • the selected UPF at this time is, for example, the UPF of MEC1, wherein the UPF of MEC1 forwards the service message to the ME APP of MEC2, and forwards it across MEC sites, but , the forwarding across MEC sites will increase the delay and consume network bandwidth.
  • FIG. 5 is a schematic structural diagram of the same-site forwarding of the UPF according to the embodiment of the present application.
  • the terminal device 1 sends a service message to the network device 1, and the UPF selected at this time is, for example, the UPF of MEC1, wherein the UPF of MEC1 forwards the service message to the ME APP of MEC1, and performs the same operation as the MEC site. Forward.
  • the UPF is responsible for forwarding the business message to the corresponding MEC APP, so the selection of the UPF is particularly important.
  • PDU Protocol Data Unit
  • Step 1 The UE sends a PDU session establishment request to the AMF.
  • the PDU session establishment request includes S-NSSAI(s) (network slice selection assistance information), DNN (data network name, Data Network Name), PDU session ID, Request Type, Old PDU Session ID.
  • Step 2 The AMF selects the SMF for the initial session, and prepares for further routing of the SM (Session Management, Session Management) information.
  • SM Session Management, Session Management
  • the AMF sends an SM request message to the SMF.
  • the SM request message may include subscriber permanent ID, DNN, S-NSSAI, PDU session ID, AMF ID, N1SM information, user location information and access technology type.
  • the N1SM information may include a PDU Session ID and a PDU Session Setup Request message.
  • Step 4 The SMF sends a subscriber data request message to the UDM.
  • the subscriber data request message may include subscriber permanent ID and DNN.
  • UDM can send subscription data response data to SMF.
  • Subscription data may include information on authenticated request types, authenticated SSC modes, and basic QoS profiles.
  • Step 5 The SMF sends an SM response message to the AMF.
  • Step 6 authentication/authorization of the PDU session.
  • Step 7a SMF selects PCF.
  • the SMF may initiate the establishment of the PDU-CAN session in the direction of the PCF in order to obtain the basic PCC rules relative to the PDU session. If the request type during procedure 3 indicates "existing PDU session", the PCF may initiate a PDU-CAN session modification.
  • Step 9 The SMF can start the PDU-CAN session.
  • Step 10a when the request type indicates "initial request" and step 5 is not performed, the SMF starts the N4 session establishment process using the selected UPF. Otherwise, the SMF can initiate the N4 session modification procedure using the selected UPF.
  • Step 10b if the SMF uses the selected UPF to start the N4 session establishment process, the UPF returns an N4 session establishment response message to the SMF; if the SMF uses the selected UPF to start the N4 session modification process, the UPF returns an N4 session modification response message to the SMF .
  • Step 11 the SMF sends an SM response message to the AMF.
  • the SM response message may include the reason, N2SM information and N1SM information.
  • the N1SM information includes a PDU Session Accept message that will be provided from the AMF to the UE.
  • Step 12 the AMF sends an N2PDU session request message to the RAN.
  • the N2PDU Session Request message may include N2SM information and NAS messages.
  • the NAS message may include a PDU Session ID and a PDU Session Setup Accept message.
  • the AMF may send a NAS message including a PDU Session ID and a PDU Session Setup Accept message.
  • the AMF adds the N2SM information from the SMF to the N2PDU Session Request message to send the N2PDU Session Request message to the RAN.
  • the RAN may designate a handshake with the UE associated with the information received from the SMF.
  • Step 14 the RAN sends an N2PDU session response message to the AMF.
  • the N2PDU Session Response message may include PDU Session ID, Reason and N2SM information.
  • the N2SM information may include PDU Session ID, (AN) tunnel information and a list of allowed/denied QoS profiles.
  • Step 15 The AMF may send an SM request message to the SMF.
  • the SM request message may include N2SM information.
  • the AMF may transmit the N2SM information received from the RAN to the SMF.
  • Step 16a the SMF sends an N4 session modification request to the UPF.
  • Step 16b the UPF sends an N4 session modification response to the SMF.
  • Step 17 The SMF may send an SM response message to the AMF. If the above process is terminated, the AMF may transmit the relevant events to the SMF. A related event occurs when the RAN tunnel information is changed or when a handover is performed to reset the AMF.
  • Step 18 The SMF sends a session management context status notification Nsmf_PDUSession_SM ContextStatusNotify(Release) to the AMF.
  • Step 19 the SMF sends a message to the UE through the UPF.
  • the SMF creates an IPv6 router advertisement to send the IPv6 router advertisement through N4 and UPF.
  • Step 20 When a PDU session establishment request is caused due to the handover between 3GPP access and non-3GPP access, that is, if the request type is set to "existing PDU session", the SMF passes the source access (3GPP access) access or non-3GPP access) to release the user plane.
  • the SMF de-registers the PDU session.
  • the SMF reports the above to the AMF.
  • the current UPF selection is the UPF selected by the SMF during the PDU session establishment process. Refer to step 8 above.
  • the SMF when the SMF selects the UPF, it usually selects the UPF according to the user's access location, the dynamic load of the UPF itself, and the capability of the UPF itself.
  • the SMF can select the UPF according to the following parameters and information Selection or re-selection of UPF:
  • -UPF is in SMF.
  • the appropriate UPF can be selected by matching the functions and characteristics required by the UE.
  • IPv4 IPv6, IPv4v6, Ethertype or unstructured
  • static IP address/prefix if applicable
  • SSC Service Continuity
  • DNAI DN Access Identifier
  • PCC Policy Control and Charging
  • S-NSSAI Single Network Slice Selection Assistance Information
  • the UE uses an access technology.
  • - AN provided identity (eg CellID, TAI), available UPF and DNAI(s);
  • This information can be obtained by SMF using N4;
  • N3 user plane termination of the AN serving the UE This can be inferred from the identity provided by the AN (eg CellID, TAI);
  • the current SMF usually considers the user's access location, the dynamic load of the UPF itself, and the capability of the UPF itself when selecting the UPF.
  • the UPF selected based on this method does not consider the number of computing resources, so When using an APP on the same site to process business, it is very likely that the business processing fails due to insufficient computing resources on the same site.
  • the present application proposes the following technical idea: since the APP of the same site of the UPF is to be used to process services, when selecting the UPF, the number of computing resources should be considered, so as to ensure that the selected APP of the same site of the UPF is selected There are enough computing resources to process the business.
  • FIG. 7 is a flowchart of the UPF selection method provided by one of the embodiments of the present application.
  • the method includes:
  • a first network element acquires computing resource information of at least one service processing device of a multi-access edge computing MEC.
  • the first network element is a network element for selecting a UPF
  • the first network element may be, for example, an AF, a PCF, or an SMF, where the AF is an AF used for intelligent scheduling , or, the first network element may also be other network elements that implement similar functions.
  • This embodiment does not limit the specific implementation of the first network element, as long as the first network element can obtain the computing power of the service processing devices of each MEC resource information, and the UPF can be selected according to the computing resource information of the service processing devices of each MEC.
  • the first network element in this embodiment may also be called a selection entity, and its specific implementation may be selected according to actual requirements.
  • the service processing device can be various servers, the server can provide various physical computing resources for the APP to process services, and the physical computing resources can include GPU, CPU, NPU, etc.
  • the servers can provide one or more physical computing resources. Example resource.
  • the first network element may obtain the computing resource information of the service processing device of at least one MEC.
  • the computing resource information of the service processing device of the MEC It may include at least one of the following: computing resource type, total resource quantity of each computing resource type, and available resource quantity of each computing resource type.
  • the type of computing power resource may be the CPU, NPU, and GPU described above, or may also be any possible type of computing power resource. It can be understood that the types of computing power resources included in each MEC may be different.
  • MEC1 and MEC2 there are currently MEC1 and MEC2. It is assumed that the computing resource types of MEC1 may include CPU and NPU, and the computing resource types of MEC2 may include NPU and GPU.
  • the computing resource information of the service processing devices of each MEC depends on the Implementation is not limited in this embodiment.
  • the computing resource information of the MEC1 may include: CPU, the total resource quantity of the CPU, the available resource quantity of the CPU, the NPU, the total resource quantity of the NPU, and the available resource quantity of the NPU.
  • the MEC may report the computing power resource information
  • the first network element may receive the computing power resource information reported by each MEC according to a preset period.
  • the first network element may send a query request for each MEC, and the first network element may obtain computing resource information returned by the MEC according to the query request.
  • the first network element selects the first MEC according to the obtained computing resource information of the service processing device of the MEC.
  • the first network element determines the user plane function UPF corresponding to the first MEC.
  • the first network element in order to avoid the situation that the computing resources of the MEC are insufficient when the APP of the same site of the selected UPF is used to process services, the first network element can obtain the computing resource information of the service processing device of the MEC according to the obtained computing resources.
  • the first MEC is selected so that the computing power resources of the selected first MEC are sufficient for the APP to process a service at least once, so as to avoid a situation in which the APP of the same site is used to process the service and the service processing fails.
  • the MEC with the largest number of available resources may be selected from the computing resource information of the service processing devices of each MEC as the first MEC, or the computing resources of the service processing devices of each MEC may be selected according to the computing resources of each MEC.
  • the ratio of the number of available resources in the information, the first MEC is selected according to the proportion in each MEC in turn. This embodiment does not limit the specific implementation of the selection of the first MEC, as long as the selected first MEC can satisfy the APP processing at least once business.
  • the UPF is responsible for forwarding the service data packets of the terminal device to the corresponding APP.
  • the first network element can determine the UPF corresponding to the first MEC, so that the UPF can forward the service data packets to the same site.
  • the APP of the same site has suitable computing resources for processing business, so that the business can be completed in the MEC site, effectively avoiding business processing failure caused by computing resources on the same site.
  • the UPF selection method provided by the embodiment of the present application includes: a first network element acquires computing resource information of at least one service processing device of a multi-access edge computing MEC. The first network element selects the first MEC according to the obtained computing resource information of the service processing device of the MEC. The first network element determines the user plane function UPF corresponding to the first MEC.
  • the first MEC is selected according to the computing resource information of the service processing devices of each MEC, and the first network element determines the UPF corresponding to the first MEC, so that the UPF can forward the service data packet to the APP of the same site, wherein the selected The first MEC has enough computing resources for the APP to process at least one business, so that when the APP of the same site processes the business, the business processing failure caused by insufficient computing resources of the same site can be effectively avoided.
  • the computing resource requirement of the terminal device can also be obtained, so as to jointly select the first MEC according to the computing resource requirement and the computing resource information of the service processing devices of each MEC.
  • the computing resource requirements of the terminal device may include at least one of the following: a second computing resource type, and the required quantity of each second computing resource type.
  • the computing resource requirement of the terminal device is used to indicate the computing resource type and computing resource quantity required by the service currently requested by the terminal device
  • the second computing resource type refers to the computing power required by the service currently requested by the terminal device.
  • the type of resource In the actual implementation process, the second computing resource type may include only one computing resource type, or may also include multiple computing resource types, which depends on the computing power required by the service currently requested by the terminal device resource.
  • the required quantity of each second computing power resource type refers to the quantity of the second computing power resource type required by the service currently requested by the terminal device.
  • the second computing power resource type may be NPU, and the required quantity of the second computing power resource type may be, for example, 2TFLOPS.
  • FLOPS Floating-point operations per second
  • an MEC whose available resource quantity of the second computing power resource type meets the preset threshold information may be obtained from each MEC, so as to obtain the second computing power resource at least one second MEC corresponding to the type, and the first MEC is selected among the second MECs.
  • a round selection is performed in at least one second MEC corresponding to the second computing power resource type to select the first MEC;
  • the intersection of at least one second MEC corresponding to each second computing power resource type is taken, and a round selection is performed in the intersection set to select the first MEC.
  • FIG. 8 is a schematic diagram 1 of the grade division of the MEC provided by the embodiment of the application
  • FIG. 9A is the The second schematic diagram of the grade division of the MEC provided by the application embodiment
  • FIG. 9B is the third schematic diagram of the grade division of the MEC provided by the application embodiment.
  • the corresponding MEC can be obtained according to the available resources of the CPU of each MEC and the preset threshold information. 's grade.
  • the preset threshold information in this embodiment includes a first preset threshold and a second preset threshold.
  • MECs with an amount of available resources greater than the first preset threshold may be classified into the first grade, then in the first grade The MEC meets the first preset threshold, and the MECs whose number of available resources is greater than the second preset threshold and less than or equal to the first preset threshold can be divided into the second grade, then the MEC in the first grade satisfies the second preset threshold, and the number of available resources less than or equal to the second preset threshold can be divided into third grades, so that the grades corresponding to each MEC can be obtained.
  • the first grade can be a grade with sufficient computing resources to process new services
  • the second grade can be a grade with a small amount of resources to process new services
  • the third grade can be a grade that cannot process new services.
  • a round selection can be performed among the MECs corresponding to the first grade to select the first MEC, taking FIG. 8 as an example, for example, when processing the current one.
  • MEC1 can be selected
  • MEC2 can be selected
  • MEC3 can be selected
  • MEC1 can be selected, and the selection is performed in turn.
  • the round selection can be performed in the second grade, that is, in MEC1, MEC2, MEC3, and MEC4. round-robin.
  • the first preset threshold and the second preset threshold may be set based on empirical values.
  • the set second preset threshold may be, for example, the minimum number of resources required to complete one CPU service , then by dividing the MEC with the number of available resources less than the second preset threshold into the third grade, it is possible to avoid selecting the MEC with insufficient CPU resources, and it can effectively ensure the CPU of the MEC selected in the first grade or the second grade.
  • the number of available resources can at least complete one service of a terminal device, which can effectively ensure that the selected first MEC has sufficient computing resources for the APP to process services, and effectively avoid insufficient computing resources on the same site, resulting in business processing failure. condition.
  • the above description takes CPU resources as an example.
  • the above operations can be performed for any type of second computing power resources.
  • the second computing power resource type required by the business of the current terminal device is NPU resources.
  • at least one second MEC can be obtained according to the number of available resources of the NPU resources of each MEC and the preset threshold information corresponding to the NPU, and the first MEC is selected in the second MEC, and its implementation is the same as that of the CPU described above.
  • the implementation methods of each computing resource type are similar, and will not be repeated here.
  • FIG. 9B is a schematic diagram of an implementation of determining the first MEC provided by this embodiment of the present application.
  • each MEC can be obtained according to the number of available resources of the CPU of each MEC and the preset threshold information.
  • the grade corresponding to the MEC, and the grade corresponding to each MEC can be obtained according to the number of available resources of the NPU of each MEC and the preset threshold information, and the intersection of the MECs is taken in the corresponding grade to determine the first MEC.
  • the corresponding MECs in the first grade include MEC1, MEC2, and MEC3.
  • the corresponding MECs in the first grade include MEC1 and MEC2, then the MECs corresponding to the CPU resources and the corresponding NPUs can be selected. The intersection of the MECs, obtains MEC1 and MEC2, and then performs round selection among MEC1 and MEC2 to realize the selection of the first MEC.
  • MECs can be selected in the second grade and an intersection set, and a round selection is performed in the intersection set to select the first MEC.
  • the first MEC can be selected quickly and efficiently, and the classification of grades can effectively ensure that the selected first MEC has enough computing resources for the APP to perform business processing.
  • the MEC with the largest number of available resources of the second computing power resource type is selected as the first MEC.
  • the current implementation is only applicable to the implementation in which the number of the second computing power resource type is one.
  • the CPU resource is also used as an example for description.
  • the CPU resources of each MEC can be For example, it can be combined with the number of available resources of the CPU resources of each MEC in Figure 8.
  • each MEC can be sorted according to the number of available resources of the CPU: MEC2, MEC1, MEC3, MEC4, MEC5, then select The MEC2 with the largest amount of available resources is the first MEC.
  • selecting the MEC with the largest number of available resources as the first MEC can ensure that the available resources of the CPU of the first MEC are sufficient for the APP to process CPU services. In extreme cases, if the number of available resources is the largest. When the MEC still cannot provide enough computing resources for the APP to process the business, it means that any current MEC cannot meet the computing resources required by the APP to process the business.
  • the implementation logic is simple, and the computing resources of the selected first MEC can effectively meet the computing resources required by the APP to process services.
  • the MEC whose available resource quantity corresponding to the second computing power resource type is less than the required quantity of the second computing power resource type is excluded to obtain the remaining MEC;
  • the weight corresponding to each MEC is obtained according to the ratio of the available resources corresponding to the second computing power resource type of each MEC in the remaining MECs, and the selection is made according to the corresponding weight of each MEC. first MEC; or
  • the number of the second computing power resource type is greater than 1, then take the intersection of the remaining MECs of each second computing power resource type, and obtain according to the ratio of the number of available resources corresponding to the second computing power resource type of each MEC in the intersection set to obtain weight corresponding to each MEC in the intersection set, and select the first MEC according to the weight corresponding to each MEC.
  • the implementation of the number of the second computing resource type is 1.
  • the second computing resource type including CPU assuming that the required number of CPU resources is 2TFLOPS in the computing resource requirements, the CPU can be The MEC with the available resource quantity less than 2TFLOPS is excluded first, because this part of the MEC cannot meet the CPU resources required by the current service of the terminal device, and the remaining MEC corresponding to the CPU resources is obtained at this time.
  • the weights of MEC1 and MEC2 can be obtained according to the ratio of the available resources of the CPUs of MEC1 and MEC2.
  • the ratio of MEC1 and MEC2 can be obtained as 2:1.
  • the weight of MEC1 can be obtained as 2, and the weight of MEC2 can be obtained as 1.
  • the first MEC is selected according to the weights of MEC1 and MEC2. For example, there are currently three services. When the first MEC is selected according to the weight, two MEC1s and one MEC2 may be selected.
  • the first service selects MEC1 as the first MEC
  • the second service selects MEC2 as the first MEC
  • the third service selects MEC1 as the first MEC
  • the first service may select MEC1 as the first MEC
  • the second service selects MEC1 as the first MEC
  • the third service selects MEC2 as the first MEC, that is to say, the number of times of the selected MECs satisfies the above ratio.
  • the implementation manner in which the number of the second computing power resource type is greater than one is described.
  • the second computing power resource type including CPU and NPU as an example, assuming that in the computing power resource requirements, the required number of CPU resources is 2TFLOPS, then you can Exclude MECs whose available resources of CPU resources are less than 2TFLOPS first to obtain the remaining MECs corresponding to CPU resources. Assuming that the required number of NPU resources is 4TFLOPS in the computing resource requirements, the number of available resources of NPU resources can be set to be less than 4TFLOPS. The MEC is first excluded to obtain the remaining MEC corresponding to the NPU resource.
  • the intersection of the two remaining MECs can be taken to obtain MEC1 and MEC2.
  • the respective weights of MEC1 and MEC2 can be obtained according to the number of NPU resources and CPU resources in MEC1 and MEC2.
  • the number of available resources of CPU and NPU of MEC1 is 1000
  • the number of available resources of CPU and NPU of MEC2 is both 500
  • the ratio of MEC1 and MEC2 can be obtained as 2:1.
  • the weight of MEC1 can be obtained as 2, and the weight of MEC2 can be obtained as 1.
  • the first MEC is selected according to the weights of the MEC1 and the MEC2, the implementation manner of which is similar to that described above, and will not be repeated here.
  • the first MEC is selected from the remaining MECs, which can effectively ensure that the selected first MEC can meet the computing power resources required by the services of the terminal equipment.
  • the weight of each MEC is obtained according to the ratio of the number of computing power resources of each remaining MEC, so that a balanced selection of each remaining MEC can be achieved.
  • the foregoing embodiment introduces a possible implementation manner of selecting the first MEC according to the computing power resource information of the service processing device of at least one MEC. Based on the foregoing description, it can be determined that the first network element in this embodiment has many Various possible implementation manners of the first network element are introduced in detail below.
  • FIG. 10 is a schematic flowchart of an AF selecting a UPF during a PDU session establishment process provided by an embodiment of the present application.
  • the AF may select the UPF during the establishment of the PDU session.
  • the intelligent scheduling AF interacts with each MEC to obtain computing resource information of the service processing device of each MEC.
  • the situation illustrated in FIG. 10 is that MEC1 and MEC2 currently exist, and the intelligent scheduling AF can obtain the computing resource information of MEC1 and MEC2.
  • the implementation is similar.
  • the AF obtains the computing resource information of the service processing devices of each MEC
  • the AF may be that, for each MEC, the AF receives the computing resource information of the service processing devices of the MEC reported by the MEC according to a preset period;
  • the AF can also send query information to the MEC for each MEC, where the query information is used to instruct the MEC to return computing resource information, and the AF can receive the computing resource of the MEC's service processing device returned by the MEC according to the query request.
  • the implementation manner of acquiring the computing resource information of the service processing devices of each MEC can be selected according to actual needs.
  • the UE sends a PDU session establishment request to the AMF, where the PDU session establishment request includes DNN, S-NSSAI, and computing resource requirements.
  • the computing resource requirement is directly carried in the PDU session establishment request.
  • the AMF sends an SM request message to the SMF, and the SMF returns an SM response message to the AMF.
  • the SMF sends the SM policy association establishment to the PCF.
  • the PCF sends an SM policy association establishment response to the SMF.
  • the implementation manner of 103-105 is the same as the implementation manner in the above-mentioned PDU session establishment process, which will not be repeated here.
  • the SMF requests the AF to select the UPF according to the computing resource information of the service processing devices of each MEC.
  • the SMF when the SMF needs to select a UPF, it can determine the first candidate UPF list according to the user's access location and UPF capability.
  • the SMF may send a first request message to the AF, where the first request message is used to request the AF to select the UPF according to the computing resource information of the service processing devices of each MEC.
  • the first request message may include, for example, the first UPF candidate list, and may also include the computing resource requirements of the terminal device.
  • the message interaction between the SMF and the intelligent scheduling AF may be forwarded through the NEF, or may be processed by the NEF. SMF and intelligent scheduling AF directly interact with each other.
  • the implementation manner in which the SMF determines that the UPF needs to be selected according to the computing resource information may be that the SMF determines that the PDU session establishment request carries the computing resource requirement, or may pre-designate a certain computing resource type. , you need to select the UPF according to the computing power resource information.
  • the AF selects a UPF according to the computing resource information of the service processing devices of each MEC, and the AF returns the selected UPF to the SMF.
  • the AF can obtain the computing resource information of the service processing devices of the MEC corresponding to the UPFs in the first candidate UPF list according to the computing resource information of the service processing devices of the MECs, and the AF can obtain the computing resource information of the service processing devices of the MEC corresponding to the UPFs in the first candidate UPF list, and the According to the computing resource information of the service processing device of the MEC, the first MEC is selected from the MECs corresponding to each UPF.
  • the first candidate UPF list includes UPF1, UPF2, and UPF4, the computing resource information of MEC1 corresponding to UPF1, the computing resource information of MEC2 corresponding to UPF2, and the computing resource information of MEC4 corresponding to UPF4 can be obtained.
  • the first MEC is selected among the three MECs, and the implementation method of selecting the first MEC among multiple MECs is described in the above-mentioned embodiment. It has been introduced in , and will not be repeated here.
  • the AF may select the UPF corresponding to the first MEC, and return the selected UPF to the SMF, and the SMF may determine the UPF according to the information returned by the AF.
  • the SMF sends an N4 session establishment request to the UPF, and the UPF sends an N4 session establishment response to the SMF.
  • the SMF sends the N1SM information and the N2SM information to the AMF.
  • the AMF returns a NAS message to the UE.
  • the UPF by directly carrying the computing resource requirement in the PDU session establishment request, the UPF can be selected in combination with the computing resource requirement and the MEC computing resource information.
  • the service feature may also be carried in the PDU session establishment request, where there is a first correspondence between the service feature and the computing resource requirements of the terminal device, and the service feature can be obtained by obtaining the service feature. Determine the corresponding computing resource requirements.
  • the service feature may be the rendered image resolution and frame rate, and it is assumed that a first correspondence between service features and computing resource requirements is configured in the AF, and the resolution and frame rate of the rendered image may be determined according to the first correspondence. Computing resource requirements.
  • the embodiment of the present application can effectively improve the flexibility of obtaining the computing resource requirements by carrying the service features in the PDU session establishment request, and determining the computing resource requirements corresponding to the service features according to the first correspondence configured in the AF.
  • the S-NSSAI may also be carried in the PDU session establishment request, where there is a second correspondence between the S-NSSAI and the computing resource requirements of the terminal device, then the S-NSSAI is obtained. , the corresponding computing resource requirements can be determined.
  • the S-NSSAI can be used to identify the services to be performed by the user.
  • the computing resource requirements can be obtained through the S-NSSAI, so the S-NSSAI and computing resource requirements can be configured. the second correspondence.
  • the computing resource requirement corresponding to the S-NSSAI is determined according to the S-NSSAI in the PDU session establishment request, so that no additional information needs to be carried.
  • the APP processing service with the MEC site can be implemented based on the selected UPF.
  • the UPF forwards the message to the nearest APP when forwarding the message.
  • the APP of the same site as the UPF can be selected.
  • the APP can be selected through the HTTPDNS process, HTTPDNS
  • the IP of the APP of the site can also be returned to the terminal device, then the service data message sent by the terminal device can include the IP of the APP of the site, and the UPF can forward the service data message to the service data message according to the IP in the service data message. corresponding APP.
  • the AF is used as the first network element, and the UPF is selected according to the computing resource information of the service processing devices of each MEC, which can effectively avoid that the computing resources of the same site are insufficient when the APP of the same site processes services. The resulting business processing fails.
  • the above embodiment describes that in the process of establishing a PDU session, the AF is used as the first network element to select the UPF.
  • the establishment of the PDU session can also be completed. After that, the selection of the UPF is performed.
  • the following describes the implementation of selecting the UPF as the first network element with the AF as the first network element after the establishment of the PDU session is completed.
  • the AF can receive the second request message from the terminal device to Obtain computing power resource requirements, location information of the terminal device, etc., in a possible implementation manner, the second request message may be an HTTP DNS request, or the second request message may also be a DNS query request, the following is combined with FIG. 11A and FIG. 11B The two implementations are introduced separately.
  • FIG. 11A is a schematic flowchart of the AF selecting a UPF after the PDU session establishment process is completed, provided by an embodiment of the present application.
  • the intelligent scheduling AF interacts with each MEC to obtain the computing resource information of the service processing device of each MEC.
  • 11A1 is similar to the implementation of 101 above, and details are not described herein again.
  • the UE sends an HTTPDNS request to the HTTPDNS server.
  • the terminal device has established a PDU session and is conducting Internet (internet) services.
  • the terminal device can send an HTTPDNS request to the HTTPDNS server, wherein the HTTPDNS request includes the computing resource requirements of the terminal device and the terminal device. Location information of the device.
  • the HTTPDNS server forwards the HTTPDNS request to the AF.
  • the HTTPDNS server can determine that the current HTTPDNS request requires ME APP according to the HTTPDNS request, and the HTTPDNS server can forward the HTTPDNS request to the AF.
  • the AF determines the MEC candidate list, and selects the first MEC in the MEC candidate list.
  • the intelligent scheduling AF determines the MEC candidate list for deploying the APP nearby according to the location information of the terminal device.
  • the AF can determine the MEC candidate list corresponding to the location information of the terminal device according to the location information of the terminal device. For example, for Each area is preset with a corresponding MEC candidate list. If the area where the location information of the terminal device is located is Shenzhen, the MEC candidate list corresponding to Shenzhen can be obtained.
  • AF selects the first MEC in the MEC candidate list according to the computing power resource information of the service processing devices of each MEC in the MEC candidate list, wherein the implementation of selecting the first MEC is the same as the implementation of selecting the first MEC described above. The method is similar and will not be repeated here.
  • AF requests the 5G core network (5G Core Network, 5GC) to adjust the bearer path through the NEF.
  • 5G Core Network 5G Core Network
  • the information carried in the information requesting the 5GC to adjust the bearer path includes: the IP of the terminal device, the DNAI corresponding to the selected first MEC, and the IP of the APP.
  • UPF1 and UPF2 currently exist, because the terminal device in this embodiment has completed the establishment of the PDU session, so in the process of establishing the PDU session, the SMF has selected a UPF, which is UPF2 in FIG. 11 , in which UPF2 can provide Internet access services, which have no special requirements for computing resources, and there are no special requirements for whether to forward nearby, so the selection of UPF2 does not need to consider computing resources.
  • UPF1 in Figure 11 is the UPF corresponding to the first MEC selected considering computing resources. In order to ensure successful service processing and forward service data packets to the APP at the same site, the selected UPF1 needs to be inserted into the bearer path middle.
  • the PCF notifies the SMF to adjust the bearer path.
  • the SMF needs to be inserted into the UPF of the nearest offload according to the DNAI, and the information carried in the SMF includes: DNAI and the filtering information of the offload (the IP corresponding to the APP).
  • the DNAI in this embodiment is the DNAI of the first MEC
  • the SMF can determine the first MEC and the UPF corresponding to the first MEC according to the DNAI, so as to determine the UPF to be inserted.
  • the AF sends an HTTPDNS request response to the HTTPDNS server.
  • the HTTPDNS server forwards the HTTPDNS request response to the UE.
  • the AF can return the APP IP to the terminal device through the HTTPDNS server.
  • Subsequent bearer interactions can be distributed by the inserted UPF1 according to the filtering rules. Packets whose destination IP is the IP of the nearest APP can be processed nearby in the edge area.
  • the message of the terminal device first arrives at UPF1 selected according to the computing resource information, and UPF1 determines whether to forward the message to the APP on the same site or to UPF2.
  • the terminal device can receive the APP IP returned by the HTTPDNS server, and the destination IP of the service data packet sent by the terminal device to the APP can be the APP IP.
  • the destination IP may be the IP of the VR APP, or for services accessing the Internet (such as Taobao.com), the destination IP may be the IP of the VR APP.
  • the IP may be the IP of Taobao. This embodiment does not limit the implementation of the destination IP, which may depend on actual service requirements.
  • the destination IP can be the IP of the VR APP, and the UPF1 can forward the service data packet of the service to the VR APP according to the destination IP.
  • the destination IP can be the IP of Taobao.com
  • UPF1 can use the service data of the service.
  • the message is forwarded to UPF2, which is then forwarded to the internet (such as Taobao.com).
  • the HTTPDNS request carries computing resource requirements.
  • the HTTPDNS request can also carry business features. There is a relationship between the business features and the computing resource requirements of the terminal device. The third correspondence.
  • FIG. 11B is another schematic flowchart of the AF selecting a UPF after the PDU session establishment process is completed, provided by the embodiment of the application.
  • the intelligent scheduling AF interacts with each MEC to obtain the computing resource information of the service processing device of each MEC.
  • the UE sends a DNS query request to the DNS server.
  • the DNS server forwards the DNS query request to the AF.
  • the AF determines the MEC candidate list, and selects the first MEC in the MEC candidate list.
  • AF requests 5G Core Network (5G Core Network, 5GC) to adjust the bearer path through NEF.
  • 5G Core Network 5G Core Network
  • the PCF notifies the SMF to adjust the bearer path.
  • the AF sends a DNS query response to the DNS server.
  • the DNS server forwards the DNS query response to the UE.
  • the implementations of 11B1 to 11B10 are similar to the implementations of 11A1 to 11A10 described above. The difference is that the implementation in FIG. 11A is an HTTP DNS request, while the current implementation in FIG. 11B is a DNS query request. Various possible implementation manners are not repeated here, and for details, reference may be made to the description in FIG. 11A above.
  • the computing resource requirement can be obtained through the HTTP DNS request or DNS query request of the terminal device, and the first MEC is selected based on the AF, and the UPF corresponding to the first MEC is selected, thereby After the terminal device establishes a PDU session, the UPF can also be selected according to the computing resource information, which ensures the comprehensiveness of the application scenario.
  • FIG. 12 is a schematic flowchart of the PCF selecting a UPF according to the embodiment of the application:
  • the PCF interacts with each MEC to acquire computing resource information of the service processing device of each MEC.
  • the UE sends a PDU session establishment request to the AMF, where the PDU session establishment request includes DNN, S-NSSAI, and computing resource requirements.
  • the AMF sends an SM request message to the SMF, and the SMF returns an SM response message to the AMF.
  • the SMF sends a second request message to the PCF.
  • the SMF selects the PCF, applies for policy information to the selected PCF, and specifically sends a second request message to the PCF, where the second request message includes the computing resource requirements of the terminal device.
  • a corresponding relationship between DNAI and MEC is configured in the PCF.
  • one DNAI is allocated to one MEC, that is, DNAI corresponding to each MEC is configured.
  • the second request message may be a newly added message, that is, the second request message indicated by 125 in FIG. 12 .
  • the second request message can also be an existing message, so there is no need to add a special request message, wherein the second request message can be, for example, the message established by the SM policy association in FIG. 12 , That is, for the message indicated by 127 in FIG. 12 , when the second request message is an existing message, the method in this embodiment may not include step 125 in FIG. 12 .
  • the PCF sends a response message of the second request message to the SMF.
  • the response message of the second request message may include, for example, the DNAI corresponding to the first MEC, and the SMF may determine the selected UPF according to the DNAI.
  • the PCF can determine the DNAI candidate list corresponding to the location information of the terminal device according to the DNN/S-NSSAI and the location information of the terminal device;
  • the PCF selects the first MEC according to the computing resource information of the service processing device of the MEC corresponding to each DNAI in the DNAI candidate list.
  • the implementation manner of determining the DANI candidate list is similar to the implementation manner of determining the MEC candidate list described above, and the DANI candidate list corresponding to each area may be preset.
  • the computing resource information of MEC1 corresponding to DNAI1, the computing resource information of MEC2 corresponding to DNAI2, and the computing resource information of MEC4 corresponding to DNAI4 can be obtained, and then PCF can According to the computing power resource information of MEC1, the computing power resource information of MEC2, and the computing power resource information of MEC4, the first MEC is selected among the three MECs, and the implementation method of selecting the first MEC among multiple MECs is described in the above-mentioned embodiment. It has been introduced in , and will not be repeated here.
  • the PCF may determine the DNAI corresponding to the first MEC as the first DNAI, and then the PCF may notify the SMF of the determined first DNAI through the response message of the second request message, wherein the configuration in the SMF If there is a UPF corresponding to each DNAI, the SMF can determine the UPF according to the DNAI, thereby realizing the selection of the UPF.
  • the SMF sends the SM policy association establishment to the PCF.
  • the PCF sends an SM policy association establishment response to the SMF.
  • the SMF sends an N4 session establishment request to the UPF, and the UPF sends an N4 session establishment response to the SMF.
  • the SMF sends the N1SM information and the N2SM information to the AMF.
  • the AMF returns a NAS message to the UE.
  • the PDU session establishment request carries the computing resource requirements.
  • the service feature may also be carried in the PDU session establishment request, or may also be carried in the PDU session establishment request.
  • Various implementation modes of S-NSSAI are the same as the above-mentioned implementation modes of directly carrying computing resource requirements in the PDU session establishment request, which will not be repeated here.
  • the PCF is used as the first network element, and the UPF is selected according to the computing resource information of the service processing devices of each MEC, which can effectively prevent the computing resources of the same site from being insufficient when the APP of the same site processes services. The resulting business processing fails.
  • FIG. 13 is a schematic flowchart of the SMF selection of the UPF provided by the embodiment of the application:
  • the AF acquires computing resource information of the service processing device of at least one MEC.
  • the SMF and the AF synchronize the computing resource information of the service processing apparatus of at least one MEC.
  • the SMF and the AF can synchronize the computing resource information of the service processing device of the at least one MEC acquired by the AF;
  • the SMF and the PCF may synchronize the computing resource information of the service processing device of the at least one MEC acquired by the PCF.
  • the UE sends a PDU session establishment request to the AMF, where the PDU session establishment request includes DNN, S-NSSAI, and computing resource requirements.
  • the computing resource requirement is directly carried in the PDU session establishment request.
  • the AMF sends an SM request message to the SMF, and the SMF returns an SM response message to the AMF.
  • the SMF sends the SM policy association establishment to the PCF.
  • the PCF sends an SM policy association establishment response to the SMF.
  • the SMF may determine the second candidate UPF list according to the access location of the terminal device and the capability of each UPF;
  • the SMF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the second candidate UPF list;
  • the SMF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing apparatuses of the MECs corresponding to the UPFs.
  • the implementation manner of determining the second candidate UPF list is similar to the implementation manner of determining the first candidate UPF list above. Assuming that the second candidate UPF list includes UPF1, UPF2 and UPF4, the computing resource information of MEC1 corresponding to UPF1 can be obtained , the computing resource information of MEC2 corresponding to UPF2, and the computing resource information of MEC4 corresponding to UPF4, then AF can use the computing resource information of MEC1, the computing resource information of MEC2, and the computing resource information of MEC4.
  • the first MEC is selected from among the MECs, and the implementation manner of selecting the first MEC among the multiple MECs has been introduced in the foregoing embodiments, and will not be repeated here.
  • the SMF may select the UPF corresponding to the first MEC, thereby realizing the selection of the UPF.
  • the SMF sends an N4 session establishment request to the UPF, and the UPF sends an N4 session establishment response to the SMF.
  • the SMF sends the N1SM information and the N2SM information to the AMF.
  • the AMF returns a NAS message to the UE.
  • the SMF is used as the first network element, and the UPF is selected according to the computing power resource information of the service processing devices of each MEC, which can effectively avoid that the computing power resources of the same site are insufficient when the APP of the same site processes services. The resulting business processing fails.
  • the AF and the PCF can directly obtain the computing resource information of the service processing devices of each MEC, and the SMF can synchronize with the AF or PCF to obtain the computing resources of the service processing devices of each MEC.
  • the first network element is AF or PCF
  • the implementation method of acquiring the computing resource information of the service processing device of each MEC is introduced.
  • FIG. 14 is a reporting method provided by this embodiment of the application
  • FIG. 15 is a schematic diagram of an implementation manner of acquiring computing resource information of a service processing apparatus of an MEC.
  • FIG. 15 is a schematic diagram of an implementation manner of acquiring computing resource information of a service processing apparatus of an MEC in a query response manner provided by an embodiment of the present application.
  • the current first network element may be an AF or a PCF, where each MEC may report its own computing resource information according to its own preset period, and the first network element may receive The computing resource information reported by each MEC according to the preset period.
  • the MEC when the grade of the MEC changes, the MEC will immediately report the computing resource information to the first network element.
  • the computing power resource information is reported according to a preset period through the MEC, so that no additional signaling instruction is required.
  • the first network element may send a query request for each MEC, where the query request is used to query the computing resource information of the MEC, and then the first network element may receive the MEC according to the query request The returned computing power resource information, so as to obtain the computing power resource information of the service processing devices of each MEC.
  • the computing power resource information of the service processing device of the MEC can be obtained through a query request, so that the required computing power resource information of the service processing device of the MEC can be obtained according to actual needs.
  • the computing resource requirements of the terminal device can be carried in the PDU session establishment request, or in the HTTP DNS request, or in the DNS query request.
  • the following is a possible implementation of the parameters carried in this application. How to explain:
  • text mode can be used.
  • the second computing power resource type includes GPU
  • the required quantity of the second computing power resource information is 2TFLOPS
  • the implementation of the text mode can be as follows:
  • Resource1 Type: GPU; Number: 2TFLOPS
  • FIG. 16 is a schematic diagram of a TLV provided by this embodiment of the application
  • FIG. 17 is a schematic diagram of a nested TVL provided by an embodiment of the present application
  • FIG. 18 is a schematic diagram of a TVL of a computing resource requirement provided by an embodiment of the present application.
  • TLV is a simple and practical custom communication protocol, which adopts the ASN.1 standard. It can be known from the full name of TLV that it includes three fields, as shown in Figure 16, namely: tag field (Tag), length field (Length), and content field (Value).
  • Tag stores the tag information
  • Length stores the length of the defined value
  • Value represents the actual value of the stored field.
  • TLV can be nested, TLTLV, where the nested TVL can be as shown in Figure 17, and the value in the TVL can also be a TVL.
  • the format of the TVL required by computing power resources can be as shown in Figure 18.
  • the tag field therein is Tag (ResourceRequire), where the TAG identifier is a parameter required by computing power resources, For example, it can be agreed that the Tag value is 10000;
  • This parameter carries two sub-parameters, which are the second computing power resource type (Type) and the specific required computing power quantity (Number).
  • subtag(Type) is a sub-parameter: used to indicate the second computing resource type in the computing resource requirement, for example, the tag value of which is stipulated to be 10.
  • TypeValue is the enumeration value of the agreed computing resource type. For example, 1 indicates that the computing resource type is GPU, 2 indicates that the computing resource type is NPU, and so on.
  • subtag(Number) is a sub-parameter: used to indicate the required number of computing resource types in the computing resource requirements, for example, its tag value is 11.
  • NumberValue is the agreed required quantity.
  • FIG. 19 is a schematic structural diagram of a UPF selection apparatus according to an embodiment of the present application.
  • the apparatus 190 includes: an acquisition module 1901 and a processing module 1902 .
  • the obtaining module 1901 is used for the first network element to obtain the computing resource information of at least one service processing device of the multi-access edge computing MEC;
  • a processing module 1902 configured for the first network element to select the first MEC according to the obtained computing resource information of the service processing device of the MEC;
  • the processing module 1902 is further configured to: determine, by the first network element, a user plane function UPF corresponding to the first MEC.
  • the computing resource information includes at least one of the following information: computing resource type, total computing resource quantity, availability of computing resource corresponding to computing resource type number of resources.
  • the obtaining module 1901 is further configured to:
  • the first network element acquires computing resource requirements of the terminal device, wherein the computing resource requirements include at least one of the following: a second computing resource type, the required quantity of each second computing resource type .
  • processing module 1902 is specifically configured to:
  • the MEC with the largest number of available resources of the second computing power resource type is selected as the first MEC.
  • processing module 1902 is specifically configured to:
  • processing module 1902 is specifically configured to:
  • the corresponding MECs are obtained. weight, and select the first MEC according to the weight corresponding to each of the MECs; or
  • the intersection of the remaining MECs of the second computing power resource types is taken, and the second computing power resources of the MECs in the intersection set are The ratio of the number of available resources corresponding to the type is obtained, the weight corresponding to each of the MECs in the intersection set is obtained, and the first MEC is selected according to the weight corresponding to each of the MECs.
  • the obtaining module 1901 is specifically used for:
  • the first network element receives the computing resource information of the service processing device of the MEC reported by the MEC according to a preset period.
  • the obtaining module 1901 is specifically used for:
  • the first network element For each of the MECs, the first network element sends a query request to the MEC;
  • the first network element is an application function AF
  • the acquisition module 1901 is specifically used for:
  • the AF receives a first request message from the SMF, where the first request message includes a computing resource requirement of the terminal device.
  • the first network element is an AF
  • the AF receives a first request message from the SMF, and the first request message includes a first candidate UPF list;
  • the processing module 1902 is specifically used for:
  • the AF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the first candidate UPF list;
  • the AF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing devices of the MECs corresponding to the UPFs.
  • the first network element is an AF
  • the acquisition module 1901 is specifically used for:
  • the AF receives a second request message from a terminal device, wherein the second request message includes a computing resource requirement of the terminal device,
  • the second request message is a hypertext transfer protocol domain name system HTTPDNS request, or the second request message is a domain name system DNS query request.
  • the first network element is an AF
  • the AF receives a second request message from a terminal device, where the second request message includes location information of the terminal device;
  • the processing module 1902 is specifically used for:
  • the AF determines, according to the location information of the terminal device, a MEC candidate list corresponding to the location information of the terminal device;
  • the AF selects a first MEC in the MEC candidate list according to the computing resource information of the service processing apparatuses of each MEC in the MEC candidate list.
  • the first network element is a policy control function PCF, wherein the data network access identifier DNAI corresponding to each MEC is configured in the PCF.
  • the obtaining module 1901 is specifically used for:
  • the PCF receives a third request message from the session management function SMF, wherein the third request message includes a computing resource requirement of the terminal device.
  • the first network element is a PCF
  • the processing module 1902 is specifically used for:
  • the PCF determines, according to the location information of the terminal device, a DNAI candidate list corresponding to the location information of the terminal device;
  • the PCF selects the first MEC according to the computing resource information of the service processing device of the MEC corresponding to each of the DNAIs in the DNAI candidate list;
  • the processing module 1902 is also used for:
  • a first DNAI is selected in the DNAI candidate list, and the first DNAI is sent to the SMF.
  • processing module 1902 is further configured to:
  • the PCF sends the first DNAI to the SMF, so that the SMF determines a first UPF based on the first DNAI.
  • the first network element is an SMF
  • the acquisition module 1901 is specifically used for:
  • the SMF and the AF synchronize the computing resource information of the service processing device of the at least one MEC obtained by the AF;
  • the SMF and the PCF synchronize the computing resource information of the service processing device of the at least one MEC acquired by the PCF.
  • the first network element is an SMF
  • the processing module 1902 is specifically used for:
  • the SMF determines a second candidate UPF list according to the access location of the terminal device and the capabilities of each UPF;
  • the SMF acquires, according to the computing resource information of each service processing device of the MEC, the computing resource information of the service processing device of the MEC corresponding to each UPF in the second candidate UPF list;
  • the SMF selects a first MEC among the MECs corresponding to the UPFs according to the computing resource information of the service processing apparatuses of the MECs corresponding to the UPFs.
  • the computing resource requirement is determined according to a PDU session establishment request sent by the terminal device.
  • the PDU session establishment request carries the computing resource requirement of the terminal device
  • the protocol data unit PDU session establishment request carries service characteristics, and there is a first correspondence between the service characteristics and the computing resource requirements of the terminal device; or
  • the PDU session establishment request carries single network slice selection assistance information S-NSSAI, and there is a second correspondence between the S-NSSAI and the computing resource requirements of the terminal device.
  • the computing resource requirement is determined according to a second request message sent by the terminal device.
  • the second request message carries the computing resource requirement of the terminal device.
  • the second request message carries service characteristics, and there is a third correspondence between the service characteristics and the computing resource requirements of the terminal device.
  • FIG. 20 is a schematic diagram of a hardware structure of an in-vehicle computing device provided by an embodiment of the present application.
  • the in-vehicle computing device 200 of this embodiment includes: a processor 2001 and a memory 2002;
  • the processor 2001 is configured to execute the computer-executed instructions stored in the memory, so as to implement each step performed by the UPF selection method in the above-mentioned embodiment.
  • the processor 2001 is configured to execute the computer-executed instructions stored in the memory, so as to implement each step performed by the UPF selection method in the above-mentioned embodiment.
  • the memory 2002 may be independent or integrated with the processor 2001 .
  • the in-vehicle computing device further includes a bus 2003 for connecting the memory 2002 and the processor 2001 .
  • the above-mentioned processor may be a central processing unit (Central Processing Unit, CPU), or other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC) )Wait.
  • a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
  • Embodiments of the present application further provide a computer storage medium, including computer instructions, when the computer instructions are executed by a processor, the UPF selection method as performed by the above-mentioned vehicle-mounted computing device is implemented.
  • Embodiments of the present application provide a computer program product, which, when the computer program product runs on a processor, implements the UPF selection method performed by the above vehicle-mounted computing device.
  • the aforementioned program can be stored in a readable memory.
  • the steps including the above method embodiments are executed; and the aforementioned memory (storage medium) includes: read-only memory (English: read-only memory, abbreviation: ROM), RAM, flash memory, hard disk, Solid state drive, magnetic tape (English: magnetic tape), floppy disk (English: floppy disk), optical disc (English: optical disc) and any combination thereof.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions
  • the apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.
  • the term “comprising” and its modifications may refer to non-limiting inclusion; the term “or” and its modifications may refer to “and/or”.
  • the terms “first”, “second” and the like in the embodiments of the present application are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence.
  • "a plurality of” refers to two or more.
  • “And/or”, which describes the association relationship of the associated objects means that there can be three kinds of relationships, for example, A and/or B, which can mean that A exists alone, A and B exist at the same time, and B exists alone.
  • the character “/" generally indicates that the associated objects are an "or” relationship.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer And Data Communications (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

一种UPF选择方法及装置,该方法包括:第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息(S701)。第一网元根据获取到的MEC的业务处理装置的算力资源信息选择第一MEC(S702)。第一网元确定第一MEC对应的用户面功能UPF(S703)。通过根据各个MEC的业务处理装置的算力资源信息选择第一MEC,并且第一网元确定第一MEC对应的UPF,从而使得UPF可以将业务数据报文转发给同站点的APP,其中选择的第一MEC有足够的算力资源供APP处理至少一次业务,从而可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。

Description

UPF选择方法及装置
本申请要求于2020年07月10日提交中国专利局、申请号为202010663446.8、申请名称为“UPF选择方法及装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及通信技术领域,尤其涉及一种UPF选择方法及装置。
背景技术
多接入边缘计算(Multi-Acess Edge Computing,MEC)能够将应用、内容和核心网部分业务处理和资源调度的功能一同部署到靠近接入侧的网络边缘,从而提高移动网络速率、降低时延并提升连接可靠性。
目前,在MEC的站点中部署用户面功能(User Plane Function,UPF)、移动边缘应用(Mobile Edge Application,ME APP)时,为满足业务时延要求、降低网络带宽占用,希望APP业务在MEC站点内完成,因此UPF的选择就显得尤为重要,现有技术中在选择UPF时,通常是SMF根据终端设备的接入位置、UPF的动态负载和UPF本身的能力来选择UPF。
然而,基于上述方式选择的UPF,在使用同站点的APP处理业务时,可能出现同站点的算力资源不够的情况,从而导致业务处理失败。
发明内容
本申请提供一种UPF选择方法及装置,用于避免同站点的算力资源不够时,导致的业务处理失败的问题。
第一方面,本申请实施例提供一种UPF选择方法,包括:
第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息,以及所述第一网元可以根据获取到的所述MEC的业务处理装置的算力资源信息选择第一MEC,从而所述第一网元确定所述第一MEC对应的用户面功能UPF。
在上述过程中,通过根据各个MEC的业务处理装置的算力资源信息选择第一MEC,并且第一网元确定第一MEC对应的UPF,从而使得UPF可以将业务数据报文转发给同站点的APP,其中选择的第一MEC有足够的算力资源供APP处理至少一次业务,从而可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。
在一种可能的实施方式中,所述算力资源信息包括如下中的至少一种信息:算力资源类型、算力资源总资源数量,所述算力资源类型的对应的算力资源的可用资源数量。
在一种可能的实施方式中,所述方法还包括:
所述第一网元获取终端设备的算力资源要求,其中,所述算力资源要求包括如下 中的至少一种:第二算力资源类型,各所述第二算力资源类型的要求数量。
其中,第一网元通过获取终端设备的算力资源要求,能够确定终端设备当前请求的业务所需要的算力资源类型是什么,以及需求的算力资源类型的数量是多少,从而能够根据终端设备的需求适应性的进行UPF的选择。
在一种可能的实施方式中,所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
针对所述第二算力资源类型,选择所述第二算力资源类型的可用资源数量最大的MEC为所述第一MEC。
在上述过程中,通过选择可用算力资源数量最大的MEC为第一MEC,其实现逻辑简单,并且有效保证选择的第一MEC的算力资源能够满足APP处理业务所需的算力资源。
在一种可能的实施方式中,所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
针对每一个所述第二算力资源类型,在各个所述MEC中获取所述第二算力资源类型的可用资源数量满足预设阈值信息的MEC,以得到所述第二算力资源类型对应的至少一个第二MEC;
若所述第二算力资源类型的数量为1个,则在所述第二算力资源类型对应的至少一个第二MEC中进行轮选,以选择所述第一MEC;或者
若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型对应的至少一个第二MEC的交集,并在所述交集中进行轮选,以选择所述第一MEC。
在上述过程中,通过确定满足预设阈值信息以进行第一MEC的选择,能够快速高效的选择到第一MEC,并且预设阈值信息能够有效保证选择到的第一MEC的算力资源数量足够APP进行业务处理。
在一种可能的实施方式中,所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
针对每一个所述第二算力资源类型,将所述第二算力资源类型对应的可用资源数量小于所述第二算力资源类型的要求数量的MEC排除,得到剩余MEC;
若所述第二算力资源类型的数量为1个,则根据所述剩余MEC中各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC;或者
若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型的剩余MEC的交集,根据所述交集中的各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到所述交集中的各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC。
在上述过程中,通过首先排除可用资源数量不足的MEC,从而在剩余的MEC中选择第一MEC,能够有效保证选择的第一MEC能够满足终端设备的业务所需要的算力资源,同时,本实施例中根据剩余的各个MEC的算力资源数量的比例得到各个MEC的权重,从而能够实现对剩余的各MEC的均衡选择。
在一种可能的实施方式中,所述第一网元获取至少一个MEC的业务处理装置的 算力资源信息,包括:
针对每一个所述MEC,所述第一网元接收所述MEC按照预设周期上报的所述MEC的业务处理装置的算力资源信息。
其中,通过MEC按照预设周期上报算力资源信息,从而无需额外的信令指示。
在一种可能的实施方式中,所述第一网元获取至少一个MEC的业务处理装置的算力资源信息,包括:
针对每一个所述MEC,所述第一网元向所述MEC发送查询请求;
接收所述MEC根据所述查询请求返回的所述MEC的业务处理装置的算力资源信息。
通过查询请求获取MEC的业务处理装置的算力资源信息,从而可以根据实际需求获取需要的MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述第一网元为应用功能AF;
所述第一网元获取终端设备的算力资源要求,包括:
所述AF接收来自于SMF的第一请求消息,其中,所述第一请求消息包括所述终端设备的算力资源要求。
在一种可能的实施方式中,所述第一网元为AF;
其中,所述AF接收来自于SMF的第一请求消息,所述第一请求消息包括第一候选UPF列表;
所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
所述AF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第一候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述AF根据所述各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
在一种可能的实施方式中,所述第一网元为AF;
所述第一网元获取终端设备的算力资源要求,包括:
所述AF接收来自于终端设备的第二请求消息,其中,所述第二请求消息包括所述终端设备的算力资源要求,
所述第二请求消息为超文本传输协议域名系统HTTPDNS请求,或者所述第二请求消息为域名系统DNS查询请求。
在一种可能的实施方式中,所述第一网元为AF;
其中,所述AF接收来自于终端设备的第二请求消息,所述第二请求消息包括所述终端设备的位置信息;
所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
所述AF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的MEC候选列表;
所述AF根据所述MEC候选列表中的各个所述MEC的业务处理装置的算力资源信息,在所述MEC候选列表中选择第一MEC。
在一种可能的实施方式中,所述第一网元为策略控制功能PCF,其中,所述PCF中配置有各个MEC分别对应的数据网络访问标识符DNAI。
在一种可能的实施方式中,所述第一网元获取终端设备的算力资源要求,包括:
所述PCF接收来自于会话管理功能SMF的第三请求消息,其中,所述第三请求消息包括所述终端设备的算力资源要求。
在一种可能的实施方式中,所述第一网元为PCF;
所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
所述PCF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的DNAI候选列表;
所述PCF根据所述DNAI候选列表中的各个所述DNAI对应的MEC的业务处理装置的算力资源信息,选择第一MEC;
所述方法还包括:
根据所述第一MEC,在所述DNAI候选列表中选择第一DNAI,并将所述第一DNAI发送给SMF。
在一种可能的实施方式中,所述方法还包括:
所述PCF将所述第一DNAI发送给所述SMF,以使得所述SMF根据所述第一DNAI确定第一UPF。
在一种可能的实施方式中,所述第一网元为SMF;
所述第一网元获取至少一个MEC的业务处理装置的算力资源信息,包括:
所述SMF和所述AF同步所述AF获取的所述至少一个MEC的业务处理装置的算力资源信息;或者
所述SMF和所述PCF同步所述PCF获取的所述至少一个MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述第一网元为SMF;
所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
所述SMF根据所述终端设备的接入位置和各UPF的能力,确定第二候选UPF列表;
所述SMF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第二候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述SMF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
在一种可能的实施方式中,所述算力资源要求为根据所述终端设备发送的PDU会话建立请求确定的。
在一种可能的实施方式中,所述PDU会话建立请求中携带有所述终端设备的算力资源要求;或者
所述协议数据单元PDU会话建立请求中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第一对应关系;或者
所述PDU会话建立请求中携带有单网络切片选择辅助信息S-NSSAI,所述S-NSSAI和所述终端设备的算力资源要求之间存在第二对应关系。
在一种可能的实施方式中,所述算力资源要求为根据所述终端设备发送的第二请求消息确定的。
在一种可能的实施方式中,所述第二请求消息中携带有所述终端设备的算力资源要求;或者
所述第二请求消息中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第三对应关系。
在上述过程中,通过基于各个不同的第一网元实现UPF的选择,从而能够有效提升UPF的选择灵活性和适用性。
第二方面,本申请实施例提供一种UPF选择装置,包括:
获取模块,用于第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息;
处理模块,用于所述第一网元根据获取到的所述MEC的业务处理装置的算力资源信息选择第一MEC;
所述处理模块还用于:所述第一网元确定所述第一MEC对应的用户面功能UPF。
在一种可能的实施方式中,所述算力资源信息包括如下中的至少一种信息:算力资源类型、算力资源总资源数量,所述算力资源类型的对应的算力资源的可用资源数量。
在一种可能的实施方式中,所述获取模块还用于:
所述第一网元获取终端设备的算力资源要求,其中,所述算力资源要求包括如下中的至少一种:第二算力资源类型,各所述第二算力资源类型的要求数量。
在一种可能的实施方式中,所述处理模块具体用于:
针对所述第二算力资源类型,选择所述第二算力资源类型的可用资源数量最大的MEC为所述第一MEC。
在一种可能的实施方式中,所述处理模块具体用于:
针对每一个所述第二算力资源类型,在各个所述MEC中获取所述第二算力资源类型的可用资源数量满足预设阈值信息的MEC,以得到所述第二算力资源类型对应的至少一个第二MEC;
若所述第二算力资源类型的数量为1个,则在所述第二算力资源类型对应的至少一个第二MEC中进行轮选,以选择所述第一MEC;或者
若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型对应的至少一个第二MEC的交集,并在所述交集中进行轮选,以选择所述第一MEC。
在一种可能的实施方式中,所述处理模块具体用于:
针对每一个所述第二算力资源类型,将所述第二算力资源类型对应的可用资源数量小于所述第二算力资源类型的要求数量的MEC排除,得到剩余MEC;
若所述第二算力资源类型的数量为1个,则根据所述剩余MEC中各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC;或者
若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型的剩 余MEC的交集,根据所述交集中的各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到所述交集中的各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC。
在一种可能的实施方式中,所述获取模块具体用于:
针对每一个所述MEC,所述第一网元接收所述MEC按照预设周期上报的所述MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述获取模块具体用于:
针对每一个所述MEC,所述第一网元向所述MEC发送查询请求;
接收所述MEC根据所述查询请求返回的所述MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述第一网元为应用功能AF;
所述获取模块具体用于:
所述AF接收来自于SMF的第一请求消息,其中,所述第一请求消息包括所述终端设备的算力资源要求。
在一种可能的实施方式中,所述第一网元为AF;
其中,所述AF接收来自于SMF的第一请求消息,所述第一请求消息包括第一候选UPF列表;
所述处理模块具体用于:
所述AF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第一候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述AF根据所述各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
在一种可能的实施方式中,所述第一网元为AF;
所述获取模块具体用于:
所述AF接收来自于终端设备的第二请求消息,其中,所述第二请求消息包括所述终端设备的算力资源要求,
所述第二请求消息为超文本传输协议域名系统HTTPDNS请求,或者所述第二请求消息为域名系统DNS查询请求。
在一种可能的实施方式中,所述第一网元为AF;
其中,所述AF接收来自于终端设备的第二请求消息,所述第二请求消息包括所述终端设备的位置信息;
所述处理模块具体用于:
所述AF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的MEC候选列表;
所述AF根据所述MEC候选列表中的各个所述MEC的业务处理装置的算力资源信息,在所述MEC候选列表中选择第一MEC。
在一种可能的实施方式中,所述第一网元为策略控制功能PCF,其中,所述PCF中配置有各个MEC分别对应的数据网络访问标识符DNAI。
在一种可能的实施方式中,所述获取模块具体用于:
所述PCF接收来自于会话管理功能SMF的第三请求消息,其中,所述第三请求消息包括所述终端设备的算力资源要求。
在一种可能的实施方式中,所述第一网元为PCF;
所述处理模块具体用于:
所述PCF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的DNAI候选列表;
所述PCF根据所述DNAI候选列表中的各个所述DNAI对应的MEC的业务处理装置的算力资源信息,选择第一MEC;
所述处理模块还用于:
根据所述第一MEC,在所述DNAI候选列表中选择第一DNAI,并将所述第一DNAI发送给SMF。
在一种可能的实施方式中,所述处理模块还用于:
所述PCF将所述第一DNAI发送给所述SMF,以使得所述SMF根据所述第一DNAI确定第一UPF。
在一种可能的实施方式中,所述第一网元为SMF;
所述获取模块具体用于:
所述SMF和所述AF同步所述AF获取的所述至少一个MEC的业务处理装置的算力资源信息;或者
所述SMF和所述PCF同步所述PCF获取的所述至少一个MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述第一网元为SMF;
所述处理模块具体用于:
所述SMF根据所述终端设备的接入位置和各UPF的能力,确定第二候选UPF列表;
所述SMF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第二候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述SMF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
在一种可能的实施方式中,所述算力资源要求为根据所述终端设备发送的PDU会话建立请求确定的。
在一种可能的实施方式中,所述PDU会话建立请求中携带有所述终端设备的算力资源要求;或者
所述协议数据单元PDU会话建立请求中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第一对应关系;或者
所述PDU会话建立请求中携带有单网络切片选择辅助信息S-NSSAI,所述S-NSSAI和所述终端设备的算力资源要求之间存在第二对应关系。
在一种可能的实施方式中,所述算力资源要求为根据所述终端设备发送的第二请求消息确定的。
在一种可能的实施方式中,所述第二请求消息中携带有所述终端设备的算力资源 要求;或者
所述第二请求消息中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第三对应关系。
第三方面,本申请实施例提供一种UPF选择设备,其特征在于,包括存储器和处理器,所述存储器存储计算机程序指令,所述处理器运行所述计算机程序指令以执行如上第一方面以及第一方面各种可能的实施方式中任一的方法。
第四方面,本申请实施例提供一种计算机存储介质,其特征在于,包括计算机指令,当所述计算机指令在被处理器运行时,实现如上第一方面以及第一方面各种可能的实施方式中任一的方法。
本申请实施例提供一种UPF选择方法及装置,该方法包括:第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息。第一网元根据获取到的MEC的业务处理装置的算力资源信息选择第一MEC。第一网元确定第一MEC对应的用户面功能UPF。通过根据各个MEC的业务处理装置的算力资源信息选择第一MEC,并且第一网元确定第一MEC对应的UPF,从而使得UPF可以将业务数据报文转发给同站点的APP,其中选择的第一MEC有足够的算力资源供APP处理至少一次业务,从而可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。
附图说明
图1为本申请适用的一种网络架构的示意图;
图2为5G网络架构的一种示意图;
图3为非漫游场景下基于服务化接口的5G网络架构的示意图;
图4为本申请实施例提供的MEC系统的架构示意图;
图5为为本申请实施例提供的UPF同站点转发的架构示意图;
图6为现有技术中UE请求PDU会话建立的过程;
图7为本申请其中一实施例提供的UPF选择方法的流程图;
图8为本申请实施例提供的MEC的档次划分示意图一;
图9A为本申请实施例提供的MEC的档次划分示意图二;
图9B为本申请实施例提供的MEC的档次划分示意图三;
图10为本申请实施例提供的在PDU会话建立过程中AF选择UPF的流程示意图;
图11A为本申请实施例提供的在PDU会话建立过程完成之后AF选择UPF的一种流程示意图;
图11B为本申请实施例提供的在PDU会话建立过程完成之后AF选择UPF的另一种流程示意图;
图12为本申请实施例提供的PCF选择UPF的流程示意图;
图13为本申请实施例提供的SMF选择UPF的流程示意图;
图14为本申请实施例提供的上报方式获取MEC的业务处理装置的算力资源信息的实现方式示意图;
图15为本申请实施例提供的查询响应方式获取MEC的业务处理装置的算力资源信息的实现方式示意图;
图16为本申请实施例提供的TLV示意图;
图17为本申请实施例提供的嵌套TVL示意图;
图18为本申请实施例提供的算力资源要求的TVL示意图;
图19为本申请一实施例提供的UPF选择装置的结构示意图;
图20为本申请实施例提供的车载计算装置的硬件结构示意图。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
首先对本申请所涉及的相关概念进行介绍:
终端设备:例如可以为各种具有通信功能的手持设备、车载设备、可穿戴设备、计算机、网络设备。例如,手持设备可以是智能手机。车载设备可以是车载导航系统。可穿戴设备可以是智能手环。计算机可以是个人数字助理(personal digital assistant,PDA)电脑、平板型电脑以及膝上型电脑(laptop computer)。网络设备可以是家庭网关(residentialgateway,RG)以及交换机。
网络设备:可以是全球移动通信系统(Global System for Mobile Communications,GSM)系统或码分多址(Code Division Multiple Access,CDMA)系统中的基站(Base Transceiver Station,BTS),也可以是WCDMA系统中的基站(NodeB,NB),还可以是LTE系统中的演进型基站(Evolutional Node B,eNB或eNodeB),或者是云无线接入网络(Cloud Radio Access Network,CRAN)中的无线控制器,或者该网络设备可以为移动交换中心、中继站、接入点、车载设备、可穿戴设备、集线器、交换机、网桥、路由器、5G网络中的网络侧设备或者未来演进的公共陆地移动网络(Public Land Mobile Network,PLMN)中的网络设备等。
S-NSSAI:单网络切片选择辅助信息(Single Network Slice Selection Assistance Information,S-NSSAI)指用来标识一个网络切片,其中,S-NSSAI包括切片/服务类型(Slice/ServiceType,SST)和切片差异(Slice Differentiator,SD)两部分:
其中SST是指在特性和服务方面预期的网络切片行为。SST的标准取值范围为1、2、3,取值1表示eMBB、2表示URLLC、3表示MIoT(Massive Internet of Things,大规模物联网)。
SD是一个可选信息,用来补充SST以区分同一个切片/业务类型的多个网络切片。其中,可用SD标识用户要进行的具体业务。
NSSAI:网络切片选择辅助信息(Network Slice Selection Assistance Information,NSSAI)是S-NSSNI的集合。5G网络中使用到的NSSAI有Requested NSSAI、Allowed NSSAI、Configured NSSAI。
其中,Requested NSSAI是UE期望使用的NSSAI,UE在注册流程中提供给网络侧的。
Allowed NSSAI是服务PLMN在注册等流程中提供的给UE的,指示UE在服务PLMN当前注册区域可以使用的S-NSSAI值,UE本地保存。
Configured NSSAI是适用于一个或多个PLMN的NSSAI,AMF在注册接受或配置更新命令等消息中下发给UE,UE本地保存。
DNAI:数据网络访问标识符(DN Access Identifier,DNAI)可以标识APP的部署位置。
本申请的方法可以应用在第五代移动通信技术(5th generation mobile networks或5th generation wireless systems,5G)系统中,也可以应用在演进的分组系统(Evolved Packet System,EPS)中,首先结合图1-图3对本申请的系统和架构进行说明。
图1为本申请适用的一种网络架构的示意图,如图1所示,该网络架构包括用户设备(User Equipment,UE)、接入网(access network,AN)、核心网和数据网络(Data Network,DN)。其中,接入网装置主要用于实现无线物理层功能、资源调度和无线资源管理、无线接入控制以及移动性管理等功能;核心网设备可以包含管理设备和网关设备,管理设备主要用于终端设备的设备注册、安全认证、移动性管理和位置管理等,网关设备主要用于与终端设备间建立通道,在该通道上转发终端设备和外部数据网络之间的数据包;数据网络可以包含网络设备(如:服务器、路由器等设备),数据网络主要用于为终端设备提供多种数据业务服务。需要说明的是,图1仅为示例性架构图,除图1中所示功能单元之外,该网络架构还可以包括其他功能单元,本发明实施例对此不进行限定。
该网络架构可以为5G网络架构,图2为5G网络架构的一种示意图,5G系统也称为新无线通信系统、新接入技术(New Radio,NR)或者下一代移动通信系统。
5G系统中的接入网可以是无线接入网(radio access network,(R)AN),5G系统中的(R)AN设备可以由多个5G-(R)AN节点组成,该5G-(R)AN节点可以包括:非3GPP的接入网络如WiFi网络的接入点(access point,AP)、下一代基站(可统称为新一代无线接入网节点(NG-RAN node),其中,下一代基站包括新空口基站(NR nodeB,gNB)、新一代演进型基站(NG-eNB)、中心单元(central unit,CU)和分布式单元(distributed unit,DU)分离形态的gNB等)、收发点(transmission receive point,TRP)、传输点(transmission point,TP)或其它节点。
如图2所示,5G核心网(5G core/new generation core,5GC/NGC)包括接入和移动性管理功能(Access and Mobility Management Function,AMF)网元、会话管理功能(Session Management Function,SMF)网元、用户面功能(User Plane Function,UPF)网元、鉴权服务器功能(Authentication Server Function,AUSF)网元、策略控制功能(Policy Control Function,PCF)网元、应用功能(Application Function,AF)网元、统一数据管理功能(unified data management,UDM)网元、网络切片选择功能(Network Slice Selection Function,NSSF)网元等多个功能单元。
AMF网元主要负责移动性管理、接入管理等服务。SMF网元主要负责会话管理、UE地址管理和分配、动态主机配置协议功能、用户面功能的选择和控制等。UPF主 要负责对外连接到数据网络(data network,DN)以及用户面的数据包路由转发、报文过滤、执行服务质量(quality of service,QoS)控制相关功能等。AUSF主要负责对终端设备的认证功能等。PCF网元主要负责为网络行为管理提供统一的策略框架、提供控制面功能的策略规则、获取与策略决策相关的注册信息等。需要说明的是,这些功能单元可以独立工作,也可以组合在一起实现某些控制功能,如对终端设备的接入鉴权、安全加密、位置注册等接入控制和移动性管理功能,以及用户面传输路径的建立、释放和更改等会话管理功能。
5GC中各功能单元之间可以通过下一代网络(next generation,NG)接口进行通信,如:UE可以通过NG接口1(简称N1)与AMF网元进行控制面消息的传输,RAN设备可以通过NG接口3(简称N3)与UPF建立用户面数据传输通道,AN/RAN设备可以通过NG接口2(简称N2)与AMF网元建立控制面信令连接,UPF可以通过NG接口4(简称N4)与SMF网元进行信息交互,UPF可以通过NG接口6(简称N6)与数据网络DN交互用户面数据,AMF网元可以通过NG接口11(简称N11)与SMF网元进行信息交互,SMF网元可以通过NG接口7(简称N7)与PCF网元进行信息交互,AMF网元可以通过NG接口12(简称N12)与AUSF进行信息交互。需要说明的是,图2仅为示例性架构图,除图2中所示功能单元之外,该网络架构还可以包括其他功能单元。
图2所示网络架构为基于参考点网络架构,且该网络架构为非漫游场景下的网络架构,当然本申请的方法也可以应用在漫游场景下,并且网络架构也不限于基于参考点的网络架构,也可以采用基于服务化接口的网络架构。
图3为非漫游场景下基于服务化接口的5G网络架构的示意图,如图3所示,基于服务化接口场景下,5G系统的核心网还包括NEF和NRF网元。基于服务化接口的场景下,5G核心网中的部分网元通过总线方式连接,如图3所示,AUSF网元、AMF网元、SMF网元、AF网元、UDM、PCF网元、网络存储功能(Network Repository Function,NRF)网元、网络开放功能(Network Exposure Function,NEF)网元和NSSF网元通过总线互连,所述网元在通过总线互连时,采用服务化接口,例如,AUSF网元通过Nausf接口连接到总线上,AMF网元采用Namf接口连接到总线上,SMF网元通过Nsmf接口连接到总线上,AF网元采用NAF网元接口连接到总线上,UDM采用Nudm接口连接到总线上,PCF网元通过NPCF网元接口连接到总线上,NRF通过Nnrf接口连接到总线上,NEF通过Nnef接口连接到总线上,NSSF通过Nnssf接口连接到总线上。
目前,5G包括三大应用场景:增强移动宽带(Enhanced Mobile Broadband,eMBB)、海量机器类通信(massive Machine Type of Communication,mMTC)和超可靠低时延通信(Ultra-reliable and Low Latency Communications,uRLLC)。
其中,eMBB聚焦对带宽有极高需求的业务,例如高清视频、虚拟现实(Virtual Reality,VR)和增强现实(Augmented Reality,AR)等;mMTC聚焦对连接密度要求较高的业务,例如智慧城市、智慧农业、智能家居等;uRLLC聚焦对时延极其敏感的业务,例如自动驾驶、工业控制、远程医疗等。
多接入边缘计算(Multi-Acess Edge Computing,MEC)是在靠近人、物或数据源头的网络边缘侧,融合网络、计算、存储、应用核心能力的开放平台,MEC能够就近 提供边缘智能服务,满足行业数字化在敏捷联接、实时业务、数据优化、应用智能等方面的关键需求。
在3GPP R15中,基于服务化架构,5G协议模块可以根据业务需求灵活调用,为构建边缘网络提供了技术标准,从而使得MEC可以按需、分场景灵活部署在无线接入云、边缘云或者汇聚云。其中,MEC可提供以下价值:通过对4K/8K、VR/AR等高带宽业务的本地分流,降低对核心网络及骨干传输网络的占用,有效提升运营商网络的利用率;通过内容与计算能力的下沉,运营商网络将有效支撑未来时延敏感型业务(车联网、远程控制等)以及大计算和高处理能力需求的业务(视频监控与分析等)。
下面结合图4对MEC系统进行说明,图4为本申请实施例提供的MEC系统的架构示意图,如图4所示:
其中,无线接入网中部署有网络设备,网络设备可以和终端设备进行交互。
参见图4,MEC部署于无线接入网和核心网之间,其中,MEC包括UPF和移动边缘应用(Mobile Edge Application,ME APP),在一种可能的实现方式中,UPF可以是MEC系统的一个组成网元,UPF负责将边缘网络的流量分发导流到MEC业务系统。
以及,ME APP可以是由第三方开发人员开发的第三方应用程序,例如可以为VR/AR应用程序等,在本申请中,在ME APP处理业务的过程中,根据业务不同,需要消耗相应的MEC服务器的算力资源,其中MEC服务器的算力资源可以包括但不限于如下中的至少一种:神经网络处理器(Neural-network Processing Unit,NPU)资源、中央处理器(Central Processing Unit,CPU)资源、图形处理器(Graphics Processing Unit,GPU)资源。
在可能的实现方式中,比如,人工智能(Artificial Intelligence,AI)业务的处理要使用NPU资源,计算业务的处理要使用CPU资源,图像渲染业务的处理要使用GPU资源。
图4中介绍的MEC的实体可以为MEC服务器,其中,MEC服务器即为部署了MEC平台及接受MEC平台管理的服务器。并且,MEC服务器可以连接云数据中心以及其他网络,例如企业网。从而,MEC服务器利用无线接入网为终端就近提供服务和云端计算功能。
其中,UPF负责对终端设备的业务消息进行转发,在一种可能的实现方式中,UPF可以将终端设备的业务消息转发给其余站点的APP,也就是说跨MEC站点进行转发,例如如图4所示,假设终端设备1向网络设备1发送业务消息,此时选择的UPF例如为MEC1的UPF,其中MEC1的UPF将业务消息转发给了MEC2的ME APP,进行了跨MEC站点的转发,然而,跨MEC站点的转发会增加时延,消耗网络带宽。
在MEC共站点部署UPF、ME APP时,为满足业务时延要求、降低网络带宽占用,通常希望APP业务在MEC站点内完成,则在另一种可能的实现方式中,UPF可以将终端设备的业务消息转发给本站点的APP,下面结合图5进行说明,图5为为本申请实施例提供的UPF同站点转发的架构示意图。
如图5所示,假设终端设备1向网络设备1发送业务消息,此时选择的UPF例如为MEC1的UPF,其中MEC1的UPF将业务消息转发给了MEC1的ME APP,进行了 同MEC站点的转发。
因为APP处理业务需要消耗算力资源,其中UPF负责将业务消息转发给相应的MEC的APP,因此UPF的选择就显得尤为重要,目前,现有技术在实现UPF的选择时,通常是在协议数据单元(Protocol Data Unit,PDU)会话建立过程中,由SMF进行UPF的选择。
下面结合图6对PDU会话建立的过程进行说明,并结合图6对现有技术中SMF选择UPF的实现方式进行介绍,图6为现有技术中UE请求PDU会话建立的过程,如图6所示,包括以下步骤。
步骤1、UE向AMF发送PDU会话建立请求,PDU会话建立请求中包括S-NSSAI(s)(网络切片选择辅助信息,network slice selection assistanceinformation)、DNN(数据网络名称,Data Network Name)、PDU会话ID、请求类型、旧的PDU会话ID。
步骤2、AMF为初始会话选择SMF,为进一步路由SM(会话管理,SessionManagement)信息准备。
步骤3、AMF向SMF发送SM请求消息。SM请求消息可以包括订户永久ID、DNN、S-NSSAI、PDU会话ID、AMF ID、N1SM信息、用户位置信息和接入技术类型。N1SM信息可以包括PDU会话ID和PDU会话建立请求消息。
步骤4、SMF向UDM发送订户数据请求消息。订户数据请求消息可以包括订户永久ID和DNN。以及UDM可以向SMF发送订阅数据响应数据。
订阅数据可以包括关于经认证的请求类型、经认证的SSC模式和基本QoS配置文件的信息。
步骤5、SMF向AMF发送SM响应消息。
步骤6、可选的,PDU会话的认证/授权。
步骤7a、SMF选择PCF。
步骤7b、SMF可以在PCF的方向上启动PDU-CAN会话建立,以便获得相对于PDU会话的基本PCC规则。如果过程3期间的请求类型表示“现有PDU会话”,则PCF可以启动PDU-CAN会话修改。
步骤8、SMF选择UPF。
步骤9、SMF可以启动PDU-CAN会话。
步骤10a、当请求类型表示“初始请求”并且不执行步骤5时,SMF使用所选择的UPF启动N4会话建立过程。否则,SMF可以使用所选择的UPF启动N4会话修改过程。
步骤10b、若SMF使用所选择的UPF启动N4会话建立过程,则UPF向SMF返回N4会话建立响应消息;若SMF使用所选择的UPF启动N4会话修改过程,则UPF向SMF返回N4会话修改响应消息。
步骤11、SMF向AMF发送SM响应消息。SM响应消息可以包括原因、N2SM信息和N1SM信息。
此外,N1SM信息包括将从AMF提供到UE的PDU会话接受消息。
步骤12、AMF向RAN发送N2PDU会话请求消息。N2PDU会话请求消息可以 包括N2SM信息和NAS消息。NAS消息可以包括PDU会话ID和PDU会话建立接受消息。
AMF可以发送包括PDU会话ID和PDU会话建立接受消息的NAS消息。另外,AMF将来自SMF的N2SM信息添加到N2PDU会话请求消息中,以向RAN发送N2PDU会话请求消息。
步骤13、RAN可以指定与和从SMF接收到的信息关联的UE进行信号交换。
步骤14、RAN向AMF发送N2PDU会话响应消息。N2PDU会话响应消息可以包括PDU会话ID、原因和N2SM信息。N2SM信息可以包括PDU会话ID、(AN)隧道信息和被允许/拒绝的QoS配置文件列表。
步骤15、AMF可以向SMF发送SM请求消息。SM请求消息可以包括N2SM信息。这里,AMF可以将从RAN接收到的N2SM信息传送到SMF。
步骤16a、SMF向UPF发送N4会话修改请求。
步骤16b、UPF向SMF发送N4会话修改响应。
步骤17、SMF可以向AMF发送SM响应消息。如果以上过程终止,则AMF可以将相关事件传送到SMF。当RAN隧道信息发生改变时或者在进行重新设置AMF的切换时,发生相关事件。
步骤18、SMF向AMF发送会话管理上下文状态通知Nsmf_PDUSession_SM ContextStatusNotify(Release)。
步骤19、SMF通过UPF向UE发送消息。详细地,在PDU类型IPv6的情况下,SMF创建IPv6路由器通告,以通过N4和UPF发送IPv6路由器通告。
步骤20、当由于3GPP接入和非3GPP接入之间的切换而引起PDU会话建立请求时,也就是说,如果请求类型被设置为“现有PDU会话”,SMF通过源接入(3GPP接入或非3GPP接入)来释放用户平面。
如果在流程步骤4之后PDU会话建立失败,则SMF去注册PDU会话。
如果在过程期间PDU会话建立失败,则SMF将以上内容报告给AMF。
基于上述介绍的PDU会话建立过程,可以确定的是,目前的UPF选择是在PDU会话建立过程中,由SMF选择的UPF,可以参见上述步骤8。
在目前的实现方式中,SMF在选择UPF时,通常是根据用户接入位置、UPF本身动态负载和UPF本身能力来选择UPF,在一种可能的实现方式中,SMF可以根据以下参数和信息进行UPF的选择或者重新选择:
UPF的动态负载。
-UPF在支持相同DNN的UPF中的相对静态容量。
-UPF位于SMF。
-UE位置信息。
-PF的功能和特定UE会话所需的功能:可以通过匹配UE所需的功能和特性来选择适当的UPF。
-数据网络名称(DNN)。
-PDU会话类型(即IPv4,IPv6,IPv4v6,以太网类型或非结构化类型)以及静态IP地址/前缀(如果适用)。
-为PDU会话选择会话和服务连续(Session and Service Continuity,SSC)模式。
-UDM中的UE用户配置文件。
-数据网络访问标识符(DN Access Identifier,DNAI)包含在策略控制和计费(Policy Control and Charging,PCC)规则中,并在第5.6.7节中描述。
-本地运营商策略。
-单网络切片选择辅助信息(Single Network Slice Selection Assistance Information,S-NSSAI)。
-UE使用接入技术。
-与用户平面拓扑和用户平面终端相关的信息,可以从以下方面推断:
-AN提供的身份(例如CellID,TAI),可用的UPF和DNAI(s);
-关于UPF的用户平面接口的信息。该信息可以由SMF使用N4获取;
-有关为UE服务的AN的N3用户平面终止的信息。这可以从AN提供的身份(例如CellID,TAI)推断出来;
-有关UPF的N9用户平面终止的信息(如果需要);
-关于与DNAI相对应的用户平面终止的信息
基于上述介绍可以确定的是,目前SMF在选择UPF时,通常是考虑用户接入位置、UPF本身动态负载和UPF本身能力,基于这种方式选择的UPF,因为没有考虑算力资源的数量,因此在使用同站点的APP处理业务时,很有可能出现由于同站点的算力资源不够,导致业务处理失败的情况。
基于现有技术中的问题,本申请提出了如下技术构思:既然要使用UPF同站点的APP处理业务,那么在选择UPF时,就考虑算力资源的数量,从而保证选择的UPF同站点的APP有足够的算力资源可以处理业务。
下面结合具体的实施例对本申请提供的UPF选择方法进行介绍,首先结合图7进行说明,图7为本申请其中一实施例提供的UPF选择方法的流程图。
如图7所示,该方法包括:
S701、第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息。
在本实施例中,第一网元为用于选择UPF的网元,第一网元例如可以为AF,还例如可以为PCF,还例如可以为SMF,其中,AF为用于智能调度的AF,或者,第一网元还可以为其他实现类似功能的网元,本实施例对第一网元的具体实现方式不做限制,只要第一网元可以获取各MEC的业务处理装置的算力资源信息,并根据各MEC的业务处理装置的算力资源信息选择UPF即可。
在其余可能的实现方式中,本实施例中的第一网元也可以称为选择实体,其具体的实现可以根据实际需求进行选择。
其中,业务处理装置可以是各种服务器,服务器可以为APP处理业务提供各种物理算力资源,物理算力资源可以包括GPU、CPU、NPU等,其中一种服务器可以提供一种或者多种物理算例资源。
在网络中,可以部署有多个MEC,则第一网元可以获取至少一个MEC的业务处理装置的算力资源信息,在一种可能的实现方式中,MEC的业务处理装置的算力资源 信息可以包括如下中的至少一种:算力资源类型、各算力资源类型的总资源数量,各算力资源类型的可用资源数量。
其中,算力资源类型可以为上述介绍的CPU、NPU、GPU,或者还可以为任一种可能的算力资源的类型,可以理解的是,各个MEC包括的算力资源类型可以不同。
例如当前存在MEC1和MEC2,假设MEC1的算力资源类型可以包括CPU和NPU,以及假设MEC2的算力资源类型可以包括NPU和GPU,各MEC的业务处理装置的算力资源信息取决于各个MEC的实现,本实施例对此不做限定。
以上述的MEC1为例,则MEC1的算力资源信息可以包括:CPU、CPU的总资源数量、CPU的可用资源数量、NPU、NPU的总资源数量、NPU的可用资源数量。
下面对获取至少一个MEC的业务处理装置的算力资源信息的可能的实现方式进行说明:
在一种可能的实现方式中,MEC可以上报算力资源信息,则第一网元可以接收各个MEC按照预设周期上报的算力资源信息。
在另一种可能的实现方式中,第一网元可以针对每一个MEC发送查询请求,则第一网元可以获取MEC根据查询请求返回的算力资源信息。
S702、第一网元根据获取到的MEC的业务处理装置的算力资源信息选择第一MEC。
S703、第一网元确定第一MEC对应的用户面功能UPF。
下面对S702和S703一起进行介绍:
在本实施例中,为了避免在使用选择的UPF的同站点的APP处理业务时,MEC的算力资源不够的情况,第一网元可以根据获取到的MEC的业务处理装置的算力资源信息选择第一MEC,使得选择的第一MEC的算力资源满足APP处理至少一次业务,以避免了使用同站点的APP处理业务,出现业务处理失败的情况。
在一种可能的实现方式中,例如可以在各个MEC的业务处理装置的算力资源信息中选择可用资源数量最大的MEC作为第一MEC,或者还可以按照各个MEC的业务处理装置的算力资源信息中的可用资源数量的比例,依次在各个MEC中按照比例选择第一MEC,本实施例对选择第一MEC的具体实现方式不做特别限制,只要选择的第一MEC可以满足APP处理至少一次业务即可。
其中,UPF负责将终端设备的业务数据报文转发给相应的APP,则本实施例中,第一网元可以确定第一MEC对应的UPF,从而使得UPF可以将业务数据报文转发给同站点的APP,其中同站点的APP有合适的算力资源用于处理业务,使得业务可以在MEC站点内完成,有效避免了同站点算力资源导致的业务处理失败。
本申请实施例提供的UPF选择方法,包括:第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息。第一网元根据获取到的MEC的业务处理装置的算力资源信息选择第一MEC。第一网元确定第一MEC对应的用户面功能UPF。通过根据各个MEC的业务处理装置的算力资源信息选择第一MEC,并且第一网元确定第一MEC对应的UPF,从而使得UPF可以将业务数据报文转发给同站点的APP,其中选择的第一MEC有足够的算力资源供APP处理至少一次业务,从而可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。
在上述实施例的基础上,下面结合具体的实施例对本申请中选择第一MEC的可能的实现方式进行说明:
首先,本实施例还可以获取终端设备的算力资源要求,以根据算力资源要求和各MEC的业务处理装置的算力资源信息,共同选择第一MEC。
其中,终端设备的算力资源要求可以包括如下中的至少一种:第二算力资源类型,各第二算力资源类型的要求数量。
其中,终端设备的算力资源要求用于指示终端设备当前请求的业务所需要的算力资源类型和算力资源数量,第二算力资源类型是指终端设备当前请求的业务所需要的算力资源的类型,在实际实现过程中,第二算力资源类型可以仅包括一个算力资源类型,或者还可以包括多个算力资源类型,其取决于终端设备当前请求的业务所需要的算力资源。
以及,各第二算力资源类型的要求数量是指终端设备当前请求的业务所需要的第二算力资源类型的数量。
比如,第二算力资源类型可以为NPU,第二算力资源类型的要求数量例如可以为2TFLOPS。其中,一个TFLOPS(teraFLOPS)等于每秒万亿(=10^12)次的浮点运算。FLOPS(Floating-point operations per second),即每秒浮点运算次数。
基于此,下面介绍选择第一MEC的实现方式:
在一种可能的实现方式中,可以针对每一个第二算力资源类型,在各个MEC中获取第二算力资源类型的可用资源数量满足预设阈值信息的MEC,以得到第二算力资源类型对应的至少一个第二MEC,并在第二MEC中选择第一MEC。
其中,若第二算力资源类型的数量为1个,则在第二算力资源类型对应的至少一个第二MEC中进行轮选,以选择第一MEC;或者
若第二算力资源类型的数量大于1个,则取各个第二算力资源类型对应的至少一个第二MEC的交集,并在交集中进行轮选,以选择第一MEC。
下面结合图8和图9A以一个具体的示例对第二算力资源类型的数量为1个的实现方式进行说明,图8为本申请实施例提供的MEC的档次划分示意图一,图9A为本申请实施例提供的MEC的档次划分示意图二,图9B为本申请实施例提供的MEC的档次划分示意图三。
假设当前第二算力资源类型的数量为1个,例如以第二算力资源类型包括CPU为例进行介绍,则可以根据各个MEC的CPU的可用资源数量和预设阈值信息,得到各个MEC对应的档次。
例如可以参见图8,当前存在MEC1、MEC2、MEC3、MEC4和MEC5,假设这5个MEC均包括CPU资源,以及假设这5个MEC的CPU的可用资源数量如图8中的5个柱状所示,同时假设本实施例中的预设阈值信息包括第一预设阈值和第二预设阈值,例如可以将可用资源数量大于第一预设阈值的MEC划分为第一档次,那么第一档次中的MEC就满足第一预设阈值,可以将可用资源数量大于第二预设阈值,小于等于第一预设阈值的MEC划分为第二档次,那么第一档次中的MEC就满足第二预设阈值,以及可以将可用资源数量小于等于第二预设阈值的划分为第三档次,从而可以得到各个MEC对应的档次。
在本实施例中,第一档次可以为有充足的算力资源处理新业务的档次,第二档次可以为有少量的资源处理新业务的档次,第三档次可以为无法处理新业务的档次。
则在选择第一MEC时,若第一档次中有对应的MEC,则可以在第一档次对应的MEC中进行轮选,以选择第一MEC,以图8为例,例如在处理当前的一次业务时,可以选择MEC1,在处理下一次业务时,可以选择MEC2,在处理再下一次的业务时,可以选择MEC3,再下一次时可以选择MEC1,依次进行轮选。
或者,若第一档次中没有对应的MEC,例如参加图9A,当前的第一档次不存在MEC,则可以在第二档次中进行轮选,也就是说在MEC1、MEC2、MEC3、MEC4中进行轮选。
在本实施例中,第一预设阈值和第二预设阈值可以是根据经验值设置的,在本实施例中,设置的第二预设阈值例如可以为完成一次CPU业务最少需要的资源数量,则通过将可用资源数量小于第二预设阈值的MEC划分为第三档次,从而可以避免选择CPU资源数量不足的MEC,能够有效保证在第一档次或者第二档次中选择的MEC的CPU的可用资源数量,起码能够完成一个终端设备的一次业务,从而能够有效保证选择的第一MEC有足够的算力资源供APP处理业务,有效避免了同站点的算力资源不足,导致业务处理失败的情况。
上述是以CPU资源为例进行的说明,本实施例中针对任意一种第二算力资源类型,均可以执行上述操作,例如当前终端设备的业务需要的种第二算力资源类是NPU资源,则可以根据各个MEC的NPU资源的可用资源数量和NPU对应的预设阈值信得到至少一个第二MEC,并在第二MEC中选择第一MEC,其实现方式与上述介绍的CPU的实现方式类似,各个算力资源类型的实现方式均类似,此处不再赘述。
下面再结合图9B以一个具体的示例对第二算力资源类型的数量大于1个的实现方式进行说明,图9B为本申请实施例提供的确定第一MEC的实现示意图。
假设当前第二算力资源类型的数量大于1个,例如以第二算力资源类型包括CPU、NPU为例进行介绍,则可以根据各个MEC的CPU的可用资源数量和预设阈值信息,得到各个MEC对应的档次,以及可以根据各个MEC的NPU的可用资源数量和预设阈值信息,得到各个MEC对应的档次,并且在对应的档次中取MEC的交集,以确定第一MEC。
例如可以参见图9B,对于CPU资源,第一档次中对应的MEC有MEC1、MEC2、MEC3对于NPU资源,第一档次中对应的MEC有MEC1、MEC2,则可以取CPU资源对应的MEC和NPU对应的MEC的交集,得到MEC1和MEC2,接着在MEC1和MEC2中进行轮选,以实现对第一MEC的选择。
或者,若第一档次中没有对应的MEC,则可以在第二档次中选择MEC并取交集,以及在交集中进行轮选,以选择第一MEC。
通过根据档次的划分进行第一MEC的选择,能够快速高效的选择到第一MEC,并且档次的划分能够有效保证选择到的第一MEC的算力资源数量足够APP进行业务处理。
在另一种可能的实现方式中,针对第二算力资源类型,选择第二算力资源类型的可用资源数量最大的MEC为第一MEC。
当前的实现方式仅适用于第二算力资源类型的数量为1个的实现方式,同样以CPU资源为例进行说明,假设第二算力资源类型包括CPU资源,则可以将各个MEC的CPU资源的可用资源数量进行排序,例如可以结合图8中各MEC的CPU资源的可用资源数量,当前根据CPU的可用资源数量对各MEC进行排序可以得到:MEC2、MEC1、MEC3、MEC4、MEC5,则选择可用资源数量最大的MEC2为第一MEC。
可以理解的是,选择可用资源数量最大的MEC作为第一MEC,能够极大概率的保证第一MEC的CPU的可用资源数量足够APP处理CPU业务,在极端的情况中,若可用资源数量最大的MEC的仍然无法提供给APP足够用的算力资源已处理业务时,则表明当前的任意一个MEC均无法满足APP处理业务所需要的算力资源。
在本实施例中,通过选择可用算力资源数量最大的MEC为第一MEC,其实现逻辑简单,并且有效保证选择的第一MEC的算力资源能够满足APP处理业务所需的算力资源。
在再一种可能的实现方式中,针对每一个第二算力资源类型,将第二算力资源类型对应的可用资源数量小于第二算力资源类型的要求数量的MEC排除,得到剩余MEC;
若第二算力资源类型的数量为1个,则根据剩余MEC中各个MEC的第二算力资源类型对应的可用资源数量的比例,得到各个MEC对应的权重,并根据各个MEC对应的权重选择第一MEC;或者
若第二算力资源类型的数量大于1个,则取各个第二算力资源类型的剩余MEC的交集,根据交集中的各个MEC的第二算力资源类型对应的可用资源数量的比例,得到交集中的各个MEC对应的权重,并根据各个MEC对应的权重选择第一MEC。
首先对第二算力资源类型的数量为1个的实现方式进行说明,以第二算力资源类型包括CPU为例,假设算力资源要求中,CPU资源的要求数量为2TFLOPS,则可以将CPU资源的可用资源数量小于2TFLOPS的MEC首先排除,因为这部分MEC无法满足终端设备当前的业务所需要的CPU资源,此时得到CPU资源对应的剩余MEC。
假设当前得到的剩余MEC包括MEC1和MEC2,则可以根据MEC1和MEC2的CPU的可用资源数量的比例,得到MEC1和MEC2的权重。
例如,MEC1的CPU的可用资源数量为1000,MEC2的CPU的可用资源数量为500,则可以得到MEC1和MEC2的比例为2:1,例如根据比例可以得到MEC1的权重为2,MEC2的权重为1。
接着根据MEC1和MEC2的权重选择第一MEC,例如当前存在3个业务,按照权重选择第一MEC时,则可以选择2个MEC1,以及选择1个MEC2。
例如,第一个业务选择MEC1作为第一MEC,第二个业务选择MEC2作为第一MEC,第三个业务选择MEC1作为第一MEC;或者,还可以第一个业务选择MEC1作为第一MEC,第二个业务选择MEC1作为第一MEC,第三个业务选择MEC2作为第一MEC,也就是说选择的MEC的次数满足上述的比例即可。
接着对第二算力资源类型的数量大于1个的实现方式进行说明,以第二算力资源类型包括CPU和NPU为例,假设算力资源要求中,CPU资源的要求数量为2TFLOPS,则可以将CPU资源的可用资源数量小于2TFLOPS的MEC首先排除,得到CPU资源 对应的剩余MEC,以及假设算力资源要求中,NPU资源的要求数量为4TFLOPS,则可以将NPU资源的可用资源数量小于4TFLOPS的MEC首先排除,得到NPU资源对应的剩余MEC。
假设CPU资源对应的剩余MEC包括MEC1、MEC2和MEC4,以及假设NPU资源对应的剩余MEC包括MEC1、MEC2和MEC3,则可以取两个剩余MEC的交集,得到MEC1和MEC2。
此时可以根据MEC1和MEC2中NPU资源和CPU资源的数量,得到MEC1和MEC2各自的权重,例如,MEC1的CPU和NPU的可用资源数量均为1000,MEC2的CPU和NPU的可用资源数量均为500,则可以得到MEC1和MEC2的比例为2:1,例如根据比例可以得到MEC1的权重为2,MEC2的权重为1。
接着根据MEC1和MEC2的权重选择第一MEC,其实现方式与上述介绍的类似,此处不再赘述。
在本实施例中,通过首先排除可用资源数量不足的MEC,从而在剩余的MEC中选择第一MEC,能够有效保证选择的第一MEC能够满足终端设备的业务所需要的算力资源,同时,本实施例中根据剩余的各个MEC的算力资源数量的比例得到各个MEC的权重,从而能够实现对剩余的各MEC的均衡选择。
上述实施例介绍了根据至少一种MEC的业务处理装置的算力资源信息选择第一MEC的可能的实现方式,基于上述介绍的内容可以确定的是,本实施例中的第一网元存在多种可能的实现方式,下面对第一网元的各种可能的实现方式分别进行详细介绍。
首先结合图10对第一网元为AF的一种可能的实现方式进行说明,图10为本申请实施例提供的在PDU会话建立过程中AF选择UPF的流程示意图。
当第一网元为AF时,在一种可能的实现方式中,可以在PDU会话建立过程中由AF选择UPF。
如图10所示:
101、智能调度AF和各个MEC交互,以获取各MEC的业务处理装置的算力资源信息。
图10中示意的情况是当前存在MEC1和MEC2,则智能调度AF可以获取MEC1和MEC2的算力资源信息,当还存在其余的MEC时,其实现方式类似。
在AF获取各MEC的业务处理装置的算力资源信息的一种可能的实现方式中,可以是针对每一个MEC,AF接收MEC按照预设周期上报的MEC的业务处理装置的算力资源信息;或者,还可以是针对每一个MEC,AF向MEC发送查询信息,其中,查询信息用于指示MEC返回算力资源信息,则AF可以接收MEC根据查询请求返回的MEC的业务处理装置的算力资源信息,在实际实现过程中,获取各MEC的业务处理装置的算力资源信息的实现方式可以根据实际需求进行选择。
102、UE向AMF发送PDU会话建立请求,PDU会话建立请求中包括DNN、S-NSSAI、算力资源要求。
在本实施例中,算力资源要求直接携带在PDU会话建立请求中。
103、AMF向SMF发送SM请求消息,以及SMF向AMF返回SM响应消息。
其中,
104、SMF向PCF发送SM策略关联建立。
105、PCF向SMF发送SM策略关联建立响应。
其中,103-105的实现方式与上述介绍的PDU会话建立过程中的实现方式相同,此处不再赘述。
106、SMF请求AF根据各MEC的业务处理装置的算力资源信息选择UPF。
在本实施例中,SMF需要选择UPF时,可以根据用户接入位置、UPF能力确定第一候选UPF列表,在一种可能的实现方式中,在SMF判断需要根据算力资源信息选择UPF时,SMF可以向AF发送第一请求消息,其中,第一请求消息用于请求AF根据各MEC的业务处理装置的算力资源信息选择UPF。
第一请求消息例如可以包括第一UPF候选列表,还可以包括终端设备的算力资源要求,在本实施例中,SMF和智能调度AF间的消息交互,可以通过NEF进行转发处理,也可以由SMF和智能调度AF直接进行消息交互。
在一种可能的实现方式中,SMF判断需要根据算力资源信息选择UPF的实现方式可以为,SMF确定PDU会话建立请求中携带有算力资源要求,或者可以预先指定针对某一算力资源类型,需要根据算力资源信息选择UPF。
107、AF根据各MEC的业务处理装置的算力资源信息选择UPF,并且AF向SMF返回选择的UPF。
在本实施例中,AF可以根据各个MEC的业务处理装置的算力资源信息,获取第一候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息,并且AF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在各个UPF对应的MEC中选择第一MEC。
例如第一候选UPF列表中包括UPF1、UPF2和UPF4,则可以获取UPF1对应的MEC1的算力资源信息、UPF2对应的MEC2的算力资源信息、UPF4对应的MEC4的算力资源信息,接着AF可以根据MEC1的算力资源信息、MEC2的算力资源信息、MEC4的算力资源信息,在这三个MEC中选择第一MEC,其中在多个MEC中选择第一MEC的实现方式在上述实施例中已经进行了介绍,此处不再赘述。
在确定第一MEC之后,AF可以选择第一MEC对应的UPF,并且向SMF返回选择的UPF,SMF可以根据AF返回的信息确定UPF。
108、SMF向UPF发送N4会话建立请求,以及UPF向SMF发送N4会话建立响应。
109、SMF向AMF发送N1SM信息和N2SM信息。
110、AMF向UE返回NAS消息。
在本申请实施例中,通过直接在PDU会话建立请求中携带算力资源要求,从而可以结合算力资源要求和MEC算力资源信息选择UPF。
在另一种可能的实现方式中,还可以在PDU会话建立请求中携带业务特征,其中,业务特征和终端设备的算力资源要求之间存在第一对应关系,则获取到业务特征,就可以确定对应的算力资源要求。
例如业务特征可以为渲染图像分辨率、帧速率,以及假设在AF中配置有业务特征和算力资源要求的第一对应关系,根据该第一对应关系可以确定渲染图像分辨率、 帧速率对应的算力资源要求。
除了配置算力资源要求的实现方式之外,其余的实现方式与上述直接在PDU会话建立请求中携带算力资源要求的实现方式相同,此处不再赘述。
本申请实施例通过在PDU会话建立请求中携带业务特征,并且根据AF中配置的第一对应关系,确定业务特征对应的算力资源要求,从而能够有效提升获取算力资源要求的灵活性。
在再一种可能的实现方式中,还可以在PDU会话建立请求中携带S-NSSAI,其中,S-NSSAI和终端设备的算力资源要求之间存在第二对应关系,则获取到S-NSSAI,就可以确定对应的算力资源要求。
在本实施例中,S-NSSAI可用于标识用户要进行的业务,对于有些算力资源要求固定的业务,可通过S-NSSAI获取算力资源要求,因此可以配置S-NSSAI和算力资源要求的第二对应关系。
除了配置算力资源要求的实现方式之外,其余的实现方式与上述直接在PDU会话建立请求中携带算力资源要求的实现方式相同,此处不再赘述。
本申请实施例中,通过根据PDU会话建立请求中S-NSSAI,确定S-NSSAI对应的算力资源要求,从而无需携带额外的信息。
在本实施例中,在选择UPF之后,可以基于选择的UPF实现同MEC站点的APP处理业务。
在一种可能的实现方式中,若各MEC的同类型APP对外呈现相同IP(AnyCast IP),则UPF在转发报文时,就近转发给本站点的APP。
在另一种可能的实现方式中,若各MEC的同类型APP对外呈现不同IP,则在选择提供服务的APP时,可以选择和UPF同站点的APP,其中选择APP例如可以通过HTTPDNS流程,HTTPDNS还可以向终端设备返回该站点APP的IP,则终端设备发送的业务数据报文中就可以包括该站点APP的IP,以及UPF可以根据业务数据报文中的IP,将业务数据报文转发给相应的APP。
在本申请实施例中,采用AF作为第一网元,根据各个MEC的业务处理装置的算力资源信息进行UPF的选择,可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。
上述实施例介绍的是在PDU会话建立过程中,由AF作为第一网元,进行UPF的选择,在AF作为第一网元的另一种可能的实现方式中,还可以在PDU会话建立完成之后,进行UPF的选择,下面结合图11对AF作为第一网元,在PDU会话建立完成之后,选择UPF的实现方式进行介绍,其中,AF可以接收来自于终端设备的第二请求消息,以获取算力资源要求、终端设备的位置信息等,在一种可能的实现方式中,第二请求消息可以为HTTPDNS请求,或者第二请求消息还可以为DNS查询请求,下面结合图11A和图11B对两种实现方式分别进行介绍。
首先结合图11A对AF可以接收来自于终端设备的HTTPDNS请求的实现方式进行介绍,图11A为本申请实施例提供的在PDU会话建立过程完成之后AF选择UPF的一种流程示意图。
如图11A所示:
11A1、智能调度AF和各个MEC交互,以获取各MEC的业务处理装置的算力资源信息。
其中,11A1的实现方式与上述101的实现方式类似,此处不再赘述。
11A2、UE向HTTPDNS服务器发送HTTPDNS请求。
在本实施例中,终端设备已经建立了PDU会话,在进行因特网(internet)业务,此时终端设备可以向HTTPDNS服务器发送HTTPDNS请求,其中,HTTPDNS请求中包括终端设备的算力资源要求,以及终端设备的位置信息。
11A3、HTTPDNS服务器将HTTPDNS请求转发给AF。
其中,HTTPDNS服务器可以根据HTTPDNS请求确定当前的HTTPDNS请求需要ME APP,则HTTPDNS服务器可以将HTTPDNS请求转发给AF。
11A4、AF确定MEC候选列表,并在MEC候选列表中选择第一MEC。
智能调度AF根据终端设备的位置信息确定就近部署APP的MEC候选列表,在一种可能的实现方式中,AF可以根据终端设备的位置信息,确定终端设备的位置信息对应的MEC候选列表,例如针对各个区域预先设置有各自对应的MEC候选列表,如终端设备的位置信息所在的区域为深圳,则可以获取深圳对应的MEC候选列表。
以及,AF根据MEC候选列表中的各个MEC的业务处理装置的算力资源信息,在MEC候选列表中选择第一MEC,其中,选择第一MEC的实现方式与上述介绍的选择第一MEC的实现方式类似,此处不再赘述。
11A5、AF通过NEF请求5G核心网(5G Core Network,5GC)调整承载路径。
其中,请求5GC调整承载路径的信息中携带的信息包括:终端设备的IP,选定的第一MEC对应的DNAI,APP的IP。
参见图11可以确定的是,当前存在UPF1和UPF2,因为本实施例中终端设备已经完成了PDU会话建立,因此在PDU会话建立过程中,SMF已经选择了一个UPF,即为图11中的UPF2,其中UPF2可以提供上网服务,其中上网服务对算力资源没有特殊的要求,对是否就近转发也没有特殊要求,因此UPF2的选择无需考虑算力资源。
然而,图11中的UPF1是考虑算力资源选择的第一MEC对应的UPF,为了保证业务处理成功,并将业务数据报文就近转发给同站点的APP,则需要将选择的UPF1插入承载路径中。
11A6、通过NEF向AF返回请求5GC调整承载路径的响应。
11A7、PCF通知SMF调整承载路径。
其中,SMF需要根据DNAI插入就近分流的UPF,其中携带的信息包括:DNAI、分流的过滤信息(APP对应的IP)。
具体的,本实施例中的DNAI为第一MEC的DNAI,则SMF根据DNAI,可以确定第一MEC,并确定第一MEC对应的UPF,从而可以确定需要插入的UPF。
11A8、SMF和UPF交互更新用户面路径。
11A9、AF向HTTPDNS服务器发送HTTPDNS请求响应。
11A10、HTTPDNS服务器将HTTPDNS请求响应转发给UE。
具体的,AF可以通过HTTPDNS服务器向终端设备返回APP IP。
后续承载交互,可以由插入的分流UPF1根据过滤规则进行分流处理,对于目的 IP为就近APP的IP的报文,可以在边缘区域内就近处理。
即终端设备的报文首先到达根据算力资源信息选择的UPF1,由UPF1判断将报文转发给同站点的APP,还是转发给UPF2。
在一种可能的实现方式中,终端设备可以接收到HTTPDNS服务器返回的APP IP,则终端设备向该APP发送的业务数据报文的目的IP可以为该APP IP。
在本实施例中,针对不同的业务可以有不同的目的IP,例如对于需要就近处理的VR业务,其目的IP可以为VR APP的IP,或者对于访问internet(如淘宝网)的业务,其目的IP可以为淘宝网的IP,本实施例对目的IP的实现方式不做限制,其可以取决于实际的业务需求。
以上述介绍的VR业务和访问internet(如淘宝网)的业务为例,下面对UPF1转发业务数据报文的可能的实现方式进行说明:
在一种可能的实现方式中,对于需要就近处理的VR业务,其目的IP可以为VR APP的IP,则UPF1可以根据目的IP,将该业务的业务数据报文转发给VR APP。
在另一种可能的实现方式中,对于不需要就近处理的业务,例如上述介绍的访问internet(如淘宝网)业务,其目的IP可以为淘宝网的IP,则UPF1可以将该业务的业务数据报文转发给UPF2,由UPF2再转发到internet(如淘宝网)。
上述介绍的是在HTTPDNS请求中携带有算力资源要求,在另一种可能的实现方式中,还可以在HTTPDNS请求中携带业务有业务特征,业务特征和终端设备的算力资源要求之间存在第三对应关系。
其各种实现方式与上述介绍的在HTTPDNS请求中携带有算力资源要求的实现方式相同,此处不再赘述。
其次结合图11B对AF可以接收来自于终端设备的DNS查询请求的实现方式进行介绍,图11B为本申请实施例提供的在PDU会话建立过程完成之后AF选择UPF的另一种流程示意图。
如图11B所示:
11B1、智能调度AF和各个MEC交互,以获取各MEC的业务处理装置的算力资源信息。
11B2、UE向DNS服务器发送DNS查询请求。
11B3、DNS服务器将DNS查询请求转发给AF。
11B4、AF确定MEC候选列表,并在MEC候选列表中选择第一MEC。
11B5、AF通过NEF请求5G核心网(5G Core Network,5GC)调整承载路径。
11B6、通过NEF向AF返回请求5GC调整承载路径的响应。
11B7、PCF通知SMF调整承载路径。
11B8、SMF和UPF交互更新用户面路径。
11B9、AF向DNS服务器发送DNS查询响应。
11B10、DNS服务器将DNS查询响应转发给UE。
其中,11B1~11B10的实现方式与上述介绍的11A1~11A10的实现方式均类似,不同之处在于,上述图11A中的实现方式为HTTPDNS请求,而当前图11B中的实现方式为DNS查询请求,此处对各种可能的实现方式不再赘述,详细内容可以参照上述图 11A的介绍。
在本实施例中,可以在终端设备建立PDU会话之后,可以通过终端设备的HTTPDNS请求或者DNS查询请求获取算力资源要求,并且基于AF选择第一MEC,并选择第一MEC对应的UPF,从而可以在终端设备建立PDU会话之后,同样可以根据算力资源信息选择UPF,保证了应用场景的全面性。
上述实施例介绍的是AF作为第一网元的实现方式,下面结合图12对PCF作为第一网元的实现方式进行介绍,图12为本申请实施例提供的PCF选择UPF的流程示意图:
如图12所示:
121、PCF和各个MEC交互,以获取各MEC的业务处理装置的算力资源信息。
122、UE向AMF发送PDU会话建立请求,PDU会话建立请求中包括DNN、S-NSSAI、算力资源要求。
123、AMF向SMF发送SM请求消息,以及SMF向AMF返回SM响应消息。
124、SMF选择PCF。
125、SMF向PCF发送第二请求消息。
在本实施例中,SMF选择PCF,并向选择的PCF申请策略信息,具体的,向PCF发送第二请求消息,其中,第二请求消息中包括终端设备的算力资源要求。
在本实施例中,PCF中配置有DNAI和MEC对应关系,例如对一个MEC分配有一个DNAI,也就是说配置有各个MEC分别对应的DNAI。
在一种可能的实现方式中,第二请求消息可以为新增的消息,即为图12中的125所示意的第二请求消息。
在另一种可能的实现方式中,第二请求消息还可以为已有的消息,则无需新增专门的请求消息,其中第二请求消息例如可以为图12中的SM策略关联建立的消息,即图12中的127所示意的消息,当第二请求消息为已有的消息时,则本实施例中的方法中就可以不包括图12中的步骤125。
126、PCF向SMF发送第二请求消息的响应消息。
在一种可能的实现方式中,第二请求消息的响应消息中例如可以包括第一MEC对应的DNAI,SMF可以根据DNAI确定选择的UPF。
例如,PCF可以根据DNN/S-NSSAI、终端设备的位置信息,确定终端设备的位置信息对应的DNAI候选列表;
PCF根据DNAI候选列表中的各个DNAI对应的MEC的业务处理装置的算力资源信息,选择第一MEC。
其中确定DANI候选列表的实现方式与上述介绍的确定MEC候选列表的实现方式类似,可以为预先设置有各个区域对应的DANI候选列表。
假设当前的DNAI候选列表中包括DNAI1、DNAI2和DNAI4,则可以获取DNAI1对应的MEC1的算力资源信息、DNAI2对应的MEC2的算力资源信息、DNAI4对应的MEC4的算力资源信息,接着PCF可以根据MEC1的算力资源信息、MEC2的算力资源信息、MEC4的算力资源信息,在这三个MEC中选择第一MEC,其中在多个MEC中选择第一MEC的实现方式在上述实施例中已经进行了介绍,此处不再赘述。
以及,本实施例中PCF可以将第一MEC对应的DNAI确定为第一DNAI,接着,PCF可以通过第二请求消息的响应消息,将确定的第一DNAI通知给SMF,其中,在SMF中配置有各个DNAI对应的UPF,则SMF可以根据DNAI确定UPF,从而实现对UPF的选择。
127、SMF向PCF发送SM策略关联建立。
128、PCF向SMF发送SM策略关联建立响应。
129、SMF向UPF发送N4会话建立请求,以及UPF向SMF发送N4会话建立响应。
1210、SMF向AMF发送N1SM信息和N2SM信息。
1211、AMF向UE返回NAS消息。
上述介绍的是在PDU会话建立请求中携带有算力资源要求,在另一种可能的实现方式中,还可以在PDU会话建立请求中携带业务特征,或者,还可以在PDU会话建立请求中携带S-NSSAI,其各种实现方式与上述直接在PDU会话建立请求中携带算力资源要求的实现方式相同,此处不再赘述。
在本申请实施例中,采用PCF作为第一网元,根据各个MEC的业务处理装置的算力资源信息进行UPF的选择,可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。
上述实施例介绍的是PCF作为第一网元的实现方式,下面结合图13对SMF作为第一网元的实现方式进行介绍,图13为本申请实施例提供的SMF选择UPF的流程示意图:
如图13所示:
131、AF获取至少一个MEC的业务处理装置的算力资源信息。
132、SMF和AF同步至少一个MEC的业务处理装置的算力资源信息。
在一种可能的实现方式中,可以在AF获取至少一个MEC的业务处理装置的算力资源信息之后,SMF和AF同步AF获取的至少一个MEC的业务处理装置的算力资源信息;或者
还可以在PCF获取至少一个MEC的业务处理装置的算力资源信息之后,SMF和PCF同步PCF获取的至少一个MEC的业务处理装置的算力资源信息。
其中,AF、PCF获取至少一个MEC的业务处理装置的算力资源信息的实现方式在上述实施例中已经进行了介绍,此处不再赘述。
133、UE向AMF发送PDU会话建立请求,PDU会话建立请求中包括DNN、S-NSSAI、算力资源要求。
在本实施例中,算力资源要求直接携带在PDU会话建立请求中。
134、AMF向SMF发送SM请求消息,以及SMF向AMF返回SM响应消息。
135、SMF向PCF发送SM策略关联建立。
136、PCF向SMF发送SM策略关联建立响应。
137、SMF选择UPF。
在一种可能的实现方式中,所述SMF可以根据所述终端设备的接入位置和各UPF的能力,确定第二候选UPF列表;
所述SMF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第二候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述SMF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
其中,确定第二候选UPF列表的实现方式与上述确定第一候选UPF列表的实现方式类似,假设第二候选UPF列表中包括UPF1、UPF2和UPF4,则可以获取UPF1对应的MEC1的算力资源信息、UPF2对应的MEC2的算力资源信息、UPF4对应的MEC4的算力资源信息,接着AF可以根据MEC1的算力资源信息、MEC2的算力资源信息、MEC4的算力资源信息,在这三个MEC中选择第一MEC,其中在多个MEC中选择第一MEC的实现方式在上述实施例中已经进行了介绍,此处不再赘述。
在确定第一MEC之后,SMF可以选择第一MEC对应的UPF,从而实现对UPF的选择。
138、SMF向UPF发送N4会话建立请求,以及UPF向SMF发送N4会话建立响应。
139、SMF向AMF发送N1SM信息和N2SM信息。
1310、AMF向UE返回NAS消息。
在本申请实施例中,采用SMF作为第一网元,根据各个MEC的业务处理装置的算力资源信息进行UPF的选择,可以有效避免同站点的APP处理业务时,同站点的算力资源不够导致的业务处理失败。
在上述实施例的基础上,可以确定的是,AF和PCF可以直接获取各个MEC的业务处理装置的算力资源信息,SMF可以和AF或者PCF同步以获取各个MEC的业务处理装置的算力资源信息,下面结合一个具体的实施例,以第一网元为AF或者PCF,对获取各个MEC的业务处理装置的算力资源信息的实现方式进行介绍,图14为本申请实施例提供的上报方式获取MEC的业务处理装置的算力资源信息的实现方式示意图,图15为本申请实施例提供的查询响应方式获取MEC的业务处理装置的算力资源信息的实现方式示意图。
在一种可能的实现方式中,参见图14,当前的第一网元可以为AF或者PCF,其中各个MEC可以按照各自的预设周期上报各自的算力资源信息,则第一网元可以接收各个MEC按照预设周期上报的算力资源信息。
在可能的实现方式中,在MEC的档次发生变化时,MEC会立刻向第一网元上报算力资源信息。
其中,通过MEC按照预设周期上报算力资源信息,从而无需额外的信令指示。
在另一种可能的实现方式中,参见图15,第一网元可以针对每一个MEC发送查询请求,其中查询请求用于查询MEC算力资源信息,接着第一网元可以接收MEC根据查询请求返回的算力资源信息,从而获取到各个MEC的业务处理装置的算力资源信息。
其中,通过查询请求获取MEC的业务处理装置的算力资源信息,从而可以根据实际需求获取需要的MEC的业务处理装置的算力资源信息。
在本申请中,终端设备的算力资源要求可以携带在PDU会话建立请求中,或者还 可以携带在HTTPDNS请求中,或者还可以携带在DNS查询请求中,下面对本申请中参数携带的可能的实现方式进行说明:
在一种可能的实现方式中,可以采用文本方式。
例如第二算力资源类型包括GPU,第二算力资源信息的要求数量是2TFLOPS,则文本方式的实现可以为:
ComputingResourceRequirements:
Resource1:Type:GPU;Number:2TFLOPS
在另一种可能的实现方式中,可以采用TLV(Tag(Type)-Length-Value)的方式,下面结合图16-图18进行介绍,图16为本申请实施例提供的TLV示意图,图17为本申请实施例提供的嵌套TVL示意图,图18为本申请实施例提供的算力资源要求的TVL示意图。
其中,TLV是一种简单实用的自定义通信协议,采用ASN.1标准。从TLV的全称中可以知道它包括三个域,如图16所示,分别为:标签域(Tag),长度域(Length),内容域(Value)。Tag保存了标签信息,Length保存了定义数值的长度,Value表示保存字段实际的数值。
TLV可以嵌套,TLTLV,其中嵌套的TVL可以如图17所示,TVL中的value还可以是一个TVL。
以及在一种可能的实现方式中,算力资源要求的TVL的格式可以如图18所示,参见图18,其中的标签域为Tag(ResourceRequire),其中TAG标识是算力资源要求的参数,比如可以约定Tag值为10000;
该参数中携带两个子参数,分别是第二算力资源类型(Type)、具体要求的算力数量(Number)。
其中,subtag(Type)为子参数:用于指示算力资源要求中的第二算力资源类型,比如约定其tag值为10。
TypeValue为约定的算力资源类型的枚举值,比如1表示算力资源类型为GPU,2表示算力资源类型为NPU,等等。
subtag(Number)为子参数:用于指示算力资源要求中的算力资源类型的要求数量,比如约定其tag值为11。
NumberValue为约定的要求数量。
图19为本申请一实施例提供的UPF选择装置的结构示意图。如图19所示,该装置190包括:获取模块1901和处理模块1902。
获取模块1901,用于第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息;
处理模块1902,用于所述第一网元根据获取到的所述MEC的业务处理装置的算力资源信息选择第一MEC;
所述处理模块1902还用于:所述第一网元确定所述第一MEC对应的用户面功能UPF。
在一种可能的实施方式中,所述算力资源信息包括如下中的至少一种信息:算力资源类型、算力资源总资源数量,所述算力资源类型的对应的算力资源的可用资源数量。
在一种可能的实施方式中,所述获取模块1901还用于:
所述第一网元获取终端设备的算力资源要求,其中,所述算力资源要求包括如下中的至少一种:第二算力资源类型,各所述第二算力资源类型的要求数量。
在一种可能的实施方式中,所述处理模块1902具体用于:
针对所述第二算力资源类型,选择所述第二算力资源类型的可用资源数量最大的MEC为所述第一MEC。
在一种可能的实施方式中,所述处理模块1902具体用于:
针对每一个所述第二算力资源类型,在各个所述MEC中获取所述第二算力资源类型的可用资源数量满足预设阈值信息的MEC,以得到所述第二算力资源类型对应的至少一个第二MEC;
若所述第二算力资源类型的数量为1个,则在所述第二算力资源类型对应的至少一个第二MEC中进行轮选,以选择所述第一MEC;或者
若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型对应的至少一个第二MEC的交集,并在所述交集中进行轮选,以选择所述第一MEC。
在一种可能的实施方式中,所述处理模块1902具体用于:
针对每一个所述第二算力资源类型,将所述第二算力资源类型对应的可用资源数量小于所述第二算力资源类型的要求数量的MEC排除,得到剩余MEC;
若所述第二算力资源类型的数量为1个,则根据所述剩余MEC中各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC;或者
若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型的剩余MEC的交集,根据所述交集中的各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到所述交集中的各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC。
在一种可能的实施方式中,所述获取模块1901具体用于:
针对每一个所述MEC,所述第一网元接收所述MEC按照预设周期上报的所述MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述获取模块1901具体用于:
针对每一个所述MEC,所述第一网元向所述MEC发送查询请求;
接收所述MEC根据所述查询请求返回的所述MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述第一网元为应用功能AF;
所述获取模块1901具体用于:
所述AF接收来自于SMF的第一请求消息,其中,所述第一请求消息包括所述终端设备的算力资源要求。
在一种可能的实施方式中,所述第一网元为AF;
其中,所述AF接收来自于SMF的第一请求消息,所述第一请求消息包括第一候选UPF列表;
所述处理模块1902具体用于:
所述AF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第一候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述AF根据所述各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
在一种可能的实施方式中,所述第一网元为AF;
所述获取模块1901具体用于:
所述AF接收来自于终端设备的第二请求消息,其中,所述第二请求消息包括所述终端设备的算力资源要求,
所述第二请求消息为超文本传输协议域名系统HTTPDNS请求,或者所述第二请求消息为域名系统DNS查询请求。
在一种可能的实施方式中,所述第一网元为AF;
其中,所述AF接收来自于终端设备的第二请求消息,所述第二请求消息包括所述终端设备的位置信息;
所述处理模块1902具体用于:
所述AF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的MEC候选列表;
所述AF根据所述MEC候选列表中的各个所述MEC的业务处理装置的算力资源信息,在所述MEC候选列表中选择第一MEC。
在一种可能的实施方式中,所述第一网元为策略控制功能PCF,其中,所述PCF中配置有各个MEC分别对应的数据网络访问标识符DNAI。
在一种可能的实施方式中,所述获取模块1901具体用于:
所述PCF接收来自于会话管理功能SMF的第三请求消息,其中,所述第三请求消息包括所述终端设备的算力资源要求。
在一种可能的实施方式中,所述第一网元为PCF;
所述处理模块1902具体用于:
所述PCF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的DNAI候选列表;
所述PCF根据所述DNAI候选列表中的各个所述DNAI对应的MEC的业务处理装置的算力资源信息,选择第一MEC;
所述处理模块1902还用于:
根据所述第一MEC,在所述DNAI候选列表中选择第一DNAI,并将所述第一DNAI发送给SMF。
在一种可能的实施方式中,所述处理模块1902还用于:
所述PCF将所述第一DNAI发送给所述SMF,以使得所述SMF根据所述第一DNAI确定第一UPF。
在一种可能的实施方式中,所述第一网元为SMF;
所述获取模块1901具体用于:
所述SMF和所述AF同步所述AF获取的所述至少一个MEC的业务处理装置的算力资源信息;或者
所述SMF和所述PCF同步所述PCF获取的所述至少一个MEC的业务处理装置的算力资源信息。
在一种可能的实施方式中,所述第一网元为SMF;
所述处理模块1902具体用于:
所述SMF根据所述终端设备的接入位置和各UPF的能力,确定第二候选UPF列表;
所述SMF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第二候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
所述SMF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
在一种可能的实施方式中,所述算力资源要求为根据所述终端设备发送的PDU会话建立请求确定的。
在一种可能的实施方式中,所述PDU会话建立请求中携带有所述终端设备的算力资源要求;或者
所述协议数据单元PDU会话建立请求中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第一对应关系;或者
所述PDU会话建立请求中携带有单网络切片选择辅助信息S-NSSAI,所述S-NSSAI和所述终端设备的算力资源要求之间存在第二对应关系。
在一种可能的实施方式中,所述算力资源要求为根据所述终端设备发送的第二请求消息确定的。
在一种可能的实施方式中,所述第二请求消息中携带有所述终端设备的算力资源要求;或者
所述第二请求消息中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第三对应关系。
本实施例提供的装置,可用于执行上述方法实施例的技术方案,其实现原理和技术效果类似,本实施例此处不再赘述。
图20为本申请实施例提供的车载计算装置的硬件结构示意图,如图20所示,本实施例的车载计算装置200包括:处理器2001以及存储器2002;其中
存储器2002,用于存储计算机执行指令;
处理器2001,用于执行存储器存储的计算机执行指令,以实现上述实施例中UPF选择方法所执行的各个步骤。具体可以参见前述方法实施例中的相关描述。
可选地,存储器2002既可以是独立的,也可以跟处理器2001集成在一起。
当存储器2002独立设置时,该车载计算装置还包括总线2003,用于连接所述存储器2002和处理器2001。
可选的,上述处理器可以是中央处理单元(Central Processing Unit,CPU),还可以是其他通用处理器、数字信号处理器(Digital Signal Processor,DSP)、专用集成电路(Application Specific Integrated Circuit,ASIC)等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本申请实施例所公开的业务处理方法实施例中的步骤可以直接体现为硬件处理器执行完成,或者用处理器中的硬件及软 件模块组合执行完成。
本申请实施例还提供一种计算机存储介质,包括计算机指令,当所述计算机指令在被处理器运行时,实现如上车载计算装置所执行的UPF选择方法。
本申请实施例提供一种计算机程序产品,当所述计算机程序产品在处理器上运行时,实现如上车载计算装置所执行的UPF选择方法。
实现上述各方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成。前述的程序可以存储于一可读取存储器中。该程序在执行时,执行包括上述各方法实施例的步骤;而前述的存储器(存储介质)包括:只读存储器(英文:read-only memory,缩写:ROM)、RAM、快闪存储器、硬盘、固态硬盘、磁带(英文:magnetic tape)、软盘(英文:floppy disk)、光盘(英文:optical disc)及其任意组合。
本申请实施例是参照根据本申请实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理单元以产生一个机器,使得通过计算机或其他可编程数据处理设备的处理单元执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中,使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品,该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上,使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
显然,本领域的技术人员可以对本申请实施例进行各种改动和变型而不脱离本申请实施例的精神和范围。这样,倘若本申请实施例的这些修改和变型属于本申请实施例权利要求及其等同技术的范围之内,则本申请实施例也意图包含这些改动和变型在内。
在本申请实施例中,术语“包括”及其变形可以指非限制性的包括;术语“或”及其变形可以指“和/或”。本本申请实施例中术语“第一”、“第二”等是用于区别类似的对象,而不必用于描述特定的顺序或先后次序。本申请实施例中,“多个”是指两个或两个以上。“和/或”,描述关联对象的关联关系,表示可以存在三种关系,例如,A和/或B,可以表示:单独存在A,同时存在A和B,单独存在B这三种情况。字符“/”一般表示前后关联对象是一种“或”的关系。
最后应说明的是:以上各实施例仅用以说明本发明实施例的技术方案,而非对其限制;尽管参照前述各实施例对本发明实施例进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分或者全部技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的 本质。

Claims (46)

  1. 一种UPF选择方法,其特征在于,包括:
    第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息;
    所述第一网元根据获取到的所述MEC的业务处理装置的算力资源信息选择第一MEC;
    所述第一网元确定所述第一MEC对应的用户面功能UPF。
  2. 根据权利要求1所述的方法,其特征在于,所述算力资源信息包括如下中的至少一种信息:算力资源类型、算力资源总资源数量,所述算力资源类型的对应的算力资源的可用资源数量。
  3. 根据权利要求2所述的方法,其特征在于,所述方法还包括:
    所述第一网元获取终端设备的算力资源要求,其中,所述算力资源要求包括如下中的至少一种:第二算力资源类型,各所述第二算力资源类型的要求数量。
  4. 根据权利要求3所述的方法,其特征在于,所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    针对所述第二算力资源类型,选择所述第二算力资源类型的可用资源数量最大的MEC为所述第一MEC。
  5. 根据权利要求3所述的方法,其特征在于,所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    针对每一个所述第二算力资源类型,在各个所述MEC中获取所述第二算力资源类型的可用资源数量满足预设阈值信息的MEC,以得到所述第二算力资源类型对应的至少一个第二MEC;
    若所述第二算力资源类型的数量为1个,则在所述第二算力资源类型对应的至少一个第二MEC中进行轮选,以选择所述第一MEC;或者
    若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型对应的至少一个第二MEC的交集,并在所述交集中进行轮选,以选择所述第一MEC。
  6. 根据权利要求3所述的方法,其特征在于,所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    针对每一个所述第二算力资源类型,将所述第二算力资源类型对应的可用资源数量小于所述第二算力资源类型的要求数量的MEC排除,得到剩余MEC;
    若所述第二算力资源类型的数量为1个,则根据所述剩余MEC中各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC;或者
    若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型的剩余MEC的交集,根据所述交集中的各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到所述交集中的各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC。
  7. 根据权利要求2所述的方法,其特征在于,所述第一网元获取至少一个MEC的业务处理装置的算力资源信息,包括:
    针对每一个所述MEC,所述第一网元接收所述MEC按照预设周期上报的所述MEC的业务处理装置的算力资源信息。
  8. 根据权利要求2所述的方法,其特征在于,所述第一网元获取至少一个MEC的业务处理装置的算力资源信息,包括:
    针对每一个所述MEC,所述第一网元向所述MEC发送查询请求;
    接收所述MEC根据所述查询请求返回的所述MEC的业务处理装置的算力资源信息。
  9. 根据权利要求7或8所述的方法,其特征在于,所述第一网元为应用功能AF;
    所述第一网元获取终端设备的算力资源要求,包括:
    所述AF接收来自于SMF的第一请求消息,其中,所述第一请求消息包括所述终端设备的算力资源要求。
  10. 根据权利要求2所述的方法,其特征在于,所述第一网元为AF;
    其中,所述AF接收来自于SMF的第一请求消息,所述第一请求消息包括第一候选UPF列表;
    所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    所述AF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第一候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
    所述AF根据所述各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
  11. 根据权利要求7或8所述的方法,其特征在于,所述第一网元为AF;
    所述第一网元获取终端设备的算力资源要求,包括:
    所述AF接收来自于终端设备的第二请求消息,其中,所述第二请求消息包括所述终端设备的算力资源要求,
    所述第二请求消息为超文本传输协议域名系统HTTPDNS请求,或者所述第二请求消息为域名系统DNS查询请求。
  12. 根据权利要求2所述的方法,其特征在于,所述第一网元为AF;
    其中,所述AF接收来自于终端设备的第二请求消息,所述第二请求消息包括所述终端设备的位置信息;
    所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    所述AF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的MEC候选列表;
    所述AF根据所述MEC候选列表中的各个所述MEC的业务处理装置的算力资源信息,在所述MEC候选列表中选择第一MEC。
  13. 根据权利要求7或8所述的方法,其特征在于,所述第一网元为策略控制功能PCF,其中,所述PCF中配置有各个MEC分别对应的数据网络访问标识符DNAI。
  14. 根据权利要求13所述的方法,其特征在于,所述第一网元获取终端设备的算力资源要求,包括:
    所述PCF接收来自于会话管理功能SMF的第三请求消息,其中,所述第三请求消息包括所述终端设备的算力资源要求。
  15. 根据权利要求2所述的方法,其特征在于,所述第一网元为PCF;
    所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    所述PCF根据终端设备的位置信息,确定所述终端设备的位置信息对应的DNAI候选列表;
    所述PCF根据所述DNAI候选列表中的各个所述DNAI对应的MEC的业务处理装置的算力资源信息,选择第一MEC;
    所述方法还包括:
    根据所述第一MEC,在所述DNAI候选列表中选择第一DNAI,并将所述第一DNAI发送给SMF。
  16. 根据权利要求15所述的方法,其特征在于,所述方法还包括:
    所述PCF将所述第一DNAI发送给所述SMF,以使得所述SMF根据所述第一DNAI确定第一UPF。
  17. 根据权利要求2所述的方法,其特征在于,所述第一网元为SMF;
    所述第一网元获取至少一个MEC的业务处理装置的算力资源信息,包括:
    所述SMF和所述AF同步所述AF获取的所述至少一个MEC的业务处理装置的算力资源信息;或者
    所述SMF和所述PCF同步所述PCF获取的所述至少一个MEC的业务处理装置的算力资源信息。
  18. 根据权利要求2所述的方法,其特征在于,所述第一网元为SMF;
    所述第一网元根据各个所述MEC的业务处理装置的算力资源信息选择第一MEC,包括:
    所述SMF根据终端设备的接入位置和各UPF的能力,确定第二候选UPF列表;
    所述SMF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第二候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
    所述SMF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
  19. 根据权利要求9或14中任一项所述的方法,其特征在于,所述算力资源要求为根据所述终端设备发送的协议数据单元PDU会话建立请求确定的。
  20. 根据权利要求19所述的方法,其特征在于,所述PDU会话建立请求中携带有所述终端设备的算力资源要求;或者
    所述PDU会话建立请求中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第一对应关系;或者
    所述PDU会话建立请求中携带有单网络切片选择辅助信息S-NSSAI,所述S-NSSAI和所述终端设备的算力资源要求之间存在第二对应关系。
  21. 根据权利要求11所述的方法,其特征在于,所述算力资源要求为根据所述终端设备发送的第二请求消息确定的。
  22. 根据权利要求21所述的方法,其特征在于,所述第二请求消息中携带有所述终端设备的算力资源要求;或者
    所述第二请求消息中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第三对应关系。
  23. 一种UPF选择装置,其特征在于,包括:
    获取模块,用于第一网元获取至少一个多接入边缘计算MEC的业务处理装置的算力资源信息;
    处理模块,用于所述第一网元根据获取到的所述MEC的业务处理装置的算力资源信息选择第一MEC;
    所述处理模块还用于:所述第一网元确定所述第一MEC对应的用户面功能UPF。
  24. 根据权利要求23所述的装置,其特征在于,所述算力资源信息包括如下中的至少一种信息:算力资源类型、算力资源总资源数量,所述算力资源类型的对应的算力资源的可用资源数量。
  25. 根据权利要求24所述的装置,其特征在于,所述获取模块还用于:
    所述第一网元获取终端设备的算力资源要求,其中,所述算力资源要求包括如下 中的至少一种:第二算力资源类型,各所述第二算力资源类型的要求数量。
  26. 根据权利要求25所述的装置,其特征在于,所述处理模块具体用于:
    针对所述第二算力资源类型,选择所述第二算力资源类型的可用资源数量最大的MEC为所述第一MEC。
  27. 根据权利要求25所述的装置,其特征在于,所述处理模块具体用于:
    针对每一个所述第二算力资源类型,在各个所述MEC中获取所述第二算力资源类型的可用资源数量满足预设阈值信息的MEC,以得到所述第二算力资源类型对应的至少一个第二MEC;
    若所述第二算力资源类型的数量为1个,则在所述第二算力资源类型对应的至少一个第二MEC中进行轮选,以选择所述第一MEC;或者
    若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型对应的至少一个第二MEC的交集,并在所述交集中进行轮选,以选择所述第一MEC。
  28. 根据权利要求25所述的装置,其特征在于,所述处理模块具体用于:
    针对每一个所述第二算力资源类型,将所述第二算力资源类型对应的可用资源数量小于所述第二算力资源类型的要求数量的MEC排除,得到剩余MEC;
    若所述第二算力资源类型的数量为1个,则根据所述剩余MEC中各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC;或者
    若所述第二算力资源类型的数量大于1个,则取各个所述第二算力资源类型的剩余MEC的交集,根据所述交集中的各个所述MEC的所述第二算力资源类型对应的可用资源数量的比例,得到所述交集中的各个所述MEC对应的权重,并根据各个所述MEC对应的权重选择所述第一MEC。
  29. 根据权利要求24所述的装置,其特征在于,所述获取模块具体用于:
    针对每一个所述MEC,所述第一网元接收所述MEC按照预设周期上报的所述MEC的业务处理装置的算力资源信息。
  30. 根据权利要求24所述的装置,其特征在于,所述获取模块具体用于:
    针对每一个所述MEC,所述第一网元向所述MEC发送查询请求;
    接收所述MEC根据所述查询请求返回的所述MEC的业务处理装置的算力资源信息。
  31. 根据权利要求29或30所述的装置,其特征在于,所述第一网元为应用功能AF;
    所述获取模块具体用于:
    所述AF接收来自于SMF的第一请求消息,其中,所述第一请求消息包括终端设备的算力资源要求。
  32. 根据权利要求24所述的装置,其特征在于,所述第一网元为AF;
    其中,所述AF接收来自于SMF的第一请求消息,所述第一请求消息包括第一候选UPF列表;
    所述处理模块具体用于:
    所述AF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第一候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
    所述AF根据所述各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
  33. 根据权利要求29或30所述的装置,其特征在于,所述第一网元为AF;
    所述获取模块具体用于:
    所述AF接收来自于终端设备的第二请求消息,其中,所述第二请求消息包括所述终端设备的算力资源要求,
    所述第二请求消息为超文本传输协议域名系统HTTPDNS请求,或者所述第二请求消息为域名系统DNS查询请求。
  34. 根据权利要求24所述的装置,其特征在于,所述第一网元为AF;
    其中,所述AF接收来自于终端设备的第二请求消息,所述第二请求消息包括所述终端设备的位置信息;
    所述处理模块具体用于:
    所述AF根据所述终端设备的位置信息,确定所述终端设备的位置信息对应的MEC候选列表;
    所述AF根据所述MEC候选列表中的各个所述MEC的业务处理装置的算力资源信息,在所述MEC候选列表中选择第一MEC。
  35. 根据权利要求29或30所述的装置,其特征在于,所述第一网元为策略控制功能PCF,其中,所述PCF中配置有各个MEC分别对应的数据网络访问标识符DNAI。
  36. 根据权利要求35所述的装置,其特征在于,所述获取模块具体用于:
    所述PCF接收来自于会话管理功能SMF的第三请求消息,其中,所述第三请求消息包括所述终端设备的算力资源要求。
  37. 根据权利要求24所述的装置,其特征在于,所述第一网元为PCF;
    所述处理模块具体用于:
    所述PCF根据终端设备的位置信息,确定所述终端设备的位置信息对应的DNAI候选列表;
    所述PCF根据所述DNAI候选列表中的各个所述DNAI对应的MEC的业务处理装置的算力资源信息,选择第一MEC;
    所述处理模块还用于:
    根据所述第一MEC,在所述DNAI候选列表中选择第一DNAI,并将所述第一DNAI发送给SMF。
  38. 根据权利要求37所述的装置,其特征在于,所述处理模块还用于:
    所述PCF将所述第一DNAI发送给所述SMF,以使得所述SMF根据所述第一DNAI确定第一UPF。
  39. 根据权利要求24所述的装置,其特征在于,所述第一网元为SMF;
    所述获取模块具体用于:
    所述SMF和所述AF同步所述AF获取的所述至少一个MEC的业务处理装置的算力资源信息;或者
    所述SMF和所述PCF同步所述PCF获取的所述至少一个MEC的业务处理装置的算力资源信息。
  40. 根据权利要求24所述的装置,其特征在于,所述第一网元为SMF;
    所述处理模块具体用于:
    所述SMF根据终端设备的接入位置和各UPF的能力,确定第二候选UPF列表;
    所述SMF根据各个所述MEC的业务处理装置的算力资源信息,获取所述第二候选UPF列表中的各个UPF对应的MEC的业务处理装置的算力资源信息;
    所述SMF根据各个UPF对应的MEC的业务处理装置的算力资源信息,在所述各个UPF对应的MEC中选择第一MEC。
  41. 根据权利要求31或36中任一项所述的装置,其特征在于,所述算力资源要求为根据所述终端设备发送的协议数据单元PDU会话建立请求确定的。
  42. 根据权利要求41所述的装置,其特征在于,所述PDU会话建立请求中携带有所述终端设备的算力资源要求;或者
    所述PDU会话建立请求中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第一对应关系;或者
    所述PDU会话建立请求中携带有单网络切片选择辅助信息S-NSSAI,所述S-NSSAI和所述终端设备的算力资源要求之间存在第二对应关系。
  43. 根据权利要求33所述的装置,其特征在于,所述算力资源要求为根据所述终端设备发送的第二请求消息确定的。
  44. 根据权利要求43所述的装置,其特征在于,所述第二请求消息中携带有所述终端设备的算力资源要求;或者
    所述第二请求消息中携带有业务特征,所述业务特征和所述终端设备的算力资源要求之间存在第三对应关系。
  45. 一种UPF选择设备,其特征在于,包括存储器和处理器,所述存储器存储计算机程序指令,所述处理器运行所述计算机程序指令以执行权利要求1-22任一项所述的方法。
  46. 一种计算机存储介质,其特征在于,包括计算机指令,当所述计算机指令在被处理器运行时,实现如权利要求1-22任一项所述的方法。
PCT/CN2021/105253 2020-07-10 2021-07-08 Upf选择方法及装置 WO2022007899A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010663446.8 2020-07-10
CN202010663446.8A CN113986520A (zh) 2020-07-10 2020-07-10 Upf选择方法及装置

Publications (1)

Publication Number Publication Date
WO2022007899A1 true WO2022007899A1 (zh) 2022-01-13

Family

ID=79552808

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/105253 WO2022007899A1 (zh) 2020-07-10 2021-07-08 Upf选择方法及装置

Country Status (2)

Country Link
CN (1) CN113986520A (zh)
WO (1) WO2022007899A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115426320A (zh) * 2022-08-25 2022-12-02 绿盟科技集团股份有限公司 一种安全资源调度方法、装置、电子设备及存储介质
WO2023246756A1 (zh) * 2022-06-21 2023-12-28 维沃移动通信有限公司 算力服务方法、装置、终端及核心网设备

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114980034A (zh) * 2021-02-26 2022-08-30 维沃移动通信有限公司 原生算力业务实现方法、装置、网络设备及终端
CN116669061A (zh) * 2022-02-16 2023-08-29 中国移动通信有限公司研究院 用户平面功能的选择方法及设备
CN114615132A (zh) * 2022-02-25 2022-06-10 亚信科技(中国)有限公司 一种分流upf的故障处理方法、装置、设备和存储介质
CN116709416A (zh) * 2022-02-26 2023-09-05 华为技术有限公司 一种通信方法及装置
CN117278409A (zh) * 2022-06-14 2023-12-22 大唐移动通信设备有限公司 算网融合方法和相关设备
CN115550372B (zh) * 2022-09-15 2024-06-04 中国联合网络通信集团有限公司 业务传输策略的确定方法、装置、设备及存储介质
CN118138982A (zh) * 2022-11-28 2024-06-04 大唐移动通信设备有限公司 分布式网络的会话锚点的确定方法、设备及装置
CN118118883A (zh) * 2022-11-29 2024-05-31 中兴通讯股份有限公司 通信方法、装置、电子设备和计算机可读介质
CN115623602B (zh) * 2022-12-19 2023-04-07 浪潮通信信息系统有限公司 资源重选方法及装置
CN118368655A (zh) * 2023-01-18 2024-07-19 维沃移动通信有限公司 时延测量的指示方法、装置及网络侧设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100145838A1 (en) * 2007-08-23 2010-06-10 Huawei Technologies Co., Ltd. Method, system, and apparatus for opening accounting data capabilities
CN110290561A (zh) * 2018-03-19 2019-09-27 大唐移动通信设备有限公司 一种本地业务的发送方法及网络设备
CN110650096A (zh) * 2018-06-27 2020-01-03 中国电信股份有限公司 基于mcdn的内容分发方法和系统
CN110727511A (zh) * 2019-09-25 2020-01-24 咪咕视讯科技有限公司 应用程序的控制方法、网络侧设备和计算机可读存储介质
CN110851529A (zh) * 2019-11-01 2020-02-28 腾讯科技(深圳)有限公司 算力调度方法及相关设备

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108463805B (zh) * 2016-01-14 2022-05-31 索尼移动通讯有限公司 用于移动边缘计算的用户设备选择
WO2019122497A1 (en) * 2017-12-21 2019-06-27 Nokia Technologies Oy Communication connection control procedure using selected communication connection slices

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100145838A1 (en) * 2007-08-23 2010-06-10 Huawei Technologies Co., Ltd. Method, system, and apparatus for opening accounting data capabilities
CN110290561A (zh) * 2018-03-19 2019-09-27 大唐移动通信设备有限公司 一种本地业务的发送方法及网络设备
CN110650096A (zh) * 2018-06-27 2020-01-03 中国电信股份有限公司 基于mcdn的内容分发方法和系统
CN110727511A (zh) * 2019-09-25 2020-01-24 咪咕视讯科技有限公司 应用程序的控制方法、网络侧设备和计算机可读存储介质
CN110851529A (zh) * 2019-11-01 2020-02-28 腾讯科技(深圳)有限公司 算力调度方法及相关设备

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023246756A1 (zh) * 2022-06-21 2023-12-28 维沃移动通信有限公司 算力服务方法、装置、终端及核心网设备
CN115426320A (zh) * 2022-08-25 2022-12-02 绿盟科技集团股份有限公司 一种安全资源调度方法、装置、电子设备及存储介质
CN115426320B (zh) * 2022-08-25 2023-06-06 绿盟科技集团股份有限公司 一种安全资源调度方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN113986520A (zh) 2022-01-28

Similar Documents

Publication Publication Date Title
WO2022007899A1 (zh) Upf选择方法及装置
US12088501B2 (en) Systems and methods for supporting traffic steering through a service function chain
US11218438B2 (en) System, apparatus and method to support data server selection
CN112314032B (zh) 通过使多个策略以信息元素为条件优化用户设备操作
US20210204148A1 (en) Real-time intelligent ran controller to support self-driving open ran
WO2020224622A1 (zh) 一种信息配置方法及装置
EP4138443A1 (en) Communication method and apparatus
US20240073848A1 (en) Network Slice in a Wireless Network
CN112997529A (zh) 用于处理无线通信网络中的服务质量的策略节点、用户平面节点、控制平面节点及其中的方法
JP2022521088A (ja) ポリシー管理方法及び装置
US20240267993A1 (en) Network Slice Management based on Inactivity
WO2023056342A1 (en) Local breakout of user plan function at enterprise 5g radio access network
US20240049048A1 (en) Traffic engineering in 5g and lte cups architecture
US20240155418A1 (en) Method and apparatus for connecting qos flow based terminal in wireless communication system
WO2022155853A1 (zh) 无线通信方法、通信装置及通信系统
WO2020200057A1 (zh) 一种通信方法及装置
WO2023083138A1 (zh) 用于通信系统的电子设备、方法和存储介质
US20230114188A1 (en) Support for Tunneling
WO2019158034A1 (zh) 一种资源分配方法和装置
US11843971B2 (en) Dynamic enablement of RLC mode based on UE RF conditions
WO2024027256A1 (zh) 一种通信方法及装置
US20230345347A1 (en) Method for determining mec access point and apparatus
US20240007395A1 (en) User equipment route selection policy for multi-link operation over multiple radio access technologies
WO2022222666A1 (zh) 一种通信方法及装置
WO2024072819A1 (en) User plane congestion notification control

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21837311

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21837311

Country of ref document: EP

Kind code of ref document: A1