WO2024103960A1 - 网络切片配置方法及系统、计算机可存储介质 - Google Patents
网络切片配置方法及系统、计算机可存储介质 Download PDFInfo
- Publication number
- WO2024103960A1 WO2024103960A1 PCT/CN2023/120348 CN2023120348W WO2024103960A1 WO 2024103960 A1 WO2024103960 A1 WO 2024103960A1 CN 2023120348 W CN2023120348 W CN 2023120348W WO 2024103960 A1 WO2024103960 A1 WO 2024103960A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- time slot
- network slice
- current time
- user
- slice configuration
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000010801 machine learning Methods 0.000 claims abstract description 20
- 230000015654 memory Effects 0.000 claims description 23
- 238000012795 verification Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 6
- 230000002787 reinforcement Effects 0.000 claims description 4
- 230000006403 short-term memory Effects 0.000 claims description 3
- 238000004891 communication Methods 0.000 abstract description 3
- 238000010200 validation analysis Methods 0.000 abstract 2
- 238000010586 diagram Methods 0.000 description 14
- 238000007726 management method Methods 0.000 description 9
- 238000003860 storage Methods 0.000 description 8
- 101100012902 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) FIG2 gene Proteins 0.000 description 6
- 238000013439 planning Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 101001121408 Homo sapiens L-amino-acid oxidase Proteins 0.000 description 3
- 102100026388 L-amino-acid oxidase Human genes 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000002411 adverse Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 101000827703 Homo sapiens Polyphosphoinositide phosphatase Proteins 0.000 description 1
- 102100023591 Polyphosphoinositide phosphatase Human genes 0.000 description 1
- 101100233916 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) KAR5 gene Proteins 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012384 transportation and delivery Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0823—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0866—Checking the configuration
- H04L41/0869—Validating the configuration within one network element
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
Definitions
- the present disclosure relates to the field of wireless communication technology, and in particular to a network slicing configuration method and system, and a computer storable medium.
- 6G networks need to meet diverse services, and network slicing is a potential method to support this vision.
- Network slicing is to build multiple logically isolated virtual networks for different services on top of a common physical network.
- the QoS (Quality of Service) requirements of different services can be guaranteed through cost-effective slice management strategies in the preparation, planning, and operation stages of the network slice life cycle.
- 6G networks need to support a variety of new services while meeting their different and strict QoS requirements, which further increases the complexity of slice management.
- a network slicing configuration system including: a radio access network RAN layer device, configured to: generate service level information of the current time slot according to user level information of the user layer device in the current time slot; predict the service demand information of the user in the current time slot by using a first machine learning model according to the service level information of the current time slot; send the service level information and service demand information of the current time slot; a software defined network SDN controller, configured to: determine the network slicing configuration strategy of the current time slot by using a second machine learning model according to the service level information and service demand information of the current time slot from the RAN layer device; perform slice configuration verification on the network slicing configuration strategy of the current time slot by using a digital twin based on the user layer device and the RAN layer device; and send the network slicing configuration strategy of the current time slot to the RAN layer device if the verification passes.
- a radio access network RAN layer device configured to: generate service level information of the current time slot according to user level information of the user layer device in the current
- the SDN controller is further configured to: send the network slice configuration policy of the current time slot to the digital twin; deploy and run the network slice configuration policy of the current time slot in the digital twin; obtain the estimated network slice performance after running the network slice configuration policy of the current time slot; and The gain of the estimated network slice performance is estimated, and the slice configuration is verified for the network slice configuration strategy of the current time slot, wherein the verification is passed when the gain of the estimated network slice performance exceeds the gain threshold.
- the RAN layer device is further configured to: receive a network slice configuration policy for a current time slot from the SDN controller; run the network slice configuration policy for the current time slot; measure user feedback information on the network slice configuration policy for the current time slot, wherein the feedback information reflects the actual network slice performance under the network slice configuration policy for the current time slot; and send the feedback information to the SDN controller for the SDN controller to determine the network slice configuration policy for the next time slot of the current time slot.
- the network slice configuration system also includes: a digital twin layer device that deploys the digital twin, configured to receive real-time network information from the user layer device and the RAN layer device, and update the digital twin based on the real-time network information.
- a digital twin layer device that deploys the digital twin, configured to receive real-time network information from the user layer device and the RAN layer device, and update the digital twin based on the real-time network information.
- the uplink between the RAN layer device and the digital twin layer device is composed of a high-bandwidth, high-speed data link, and the downlink between the RAN layer device and the digital twin layer device is a control channel.
- the user level information includes at least one of a service demand pattern of a user, a mobility pattern of a user, location information of a user, and a random channel condition of a user.
- the RAN layer device is also configured to perform at least one of the following: parsing the service demand pattern of the user to obtain the user's demand for latency, bandwidth and service content information as service level information; analyzing the user's mobility pattern to obtain the user's mobility pattern as service level information; obtaining the current channel signal-to-noise ratio based on the user's random channel conditions as service level information; and determining the user's density information based on the user's location information as service level information.
- the first machine learning model includes a long short-term memory network LSTM model; and/or the second machine learning model includes a deep reinforcement learning model.
- the network slice configuration system further includes: the user layer device is configured to send user level information of the current time slot to the RAN layer device.
- the RAN layer device includes an access point device of the RAN layer.
- a network slicing configuration method including: a radio access network RAN layer device generates service level information according to user level information of a user layer device in a current time slot; the RAN layer device predicts the service demand information of the user in the current time slot using a first machine learning model according to the service level information of the current time slot; the RAN layer device sends the service level information and service demand information of the current time slot; a software defined network SDN controller generates service level information according to the service level information and service demand information of the current time slot from the RAN layer device; The SDN controller uses the digital twin based on the user layer device and the RAN layer device to verify the network slice configuration policy of the current time slot. If the verification is successful, the SDN controller sends the network slice configuration policy of the current time slot to the RAN layer device.
- a network slice configuration system comprising: a memory; and a processor coupled to the memory, the processor being configured to execute the network slice configuration method described in any of the above embodiments based on instructions stored in the memory.
- a computer-storable medium on which computer program instructions are stored, and when the instructions are executed by a processor, the network slice configuration method described in any of the above embodiments is implemented.
- FIG1 is a block diagram illustrating a network slice configuration system according to some embodiments of the present disclosure
- FIG2 is a block diagram showing a network slice configuration system according to some other embodiments of the present disclosure.
- FIG3 is a schematic diagram showing the architecture of a network slice configuration system according to some embodiments of the present disclosure.
- FIG4 is a flowchart illustrating a network slice configuration method according to some embodiments of the present disclosure.
- FIG5 is a flowchart illustrating a network slice configuration method according to some other embodiments of the present disclosure.
- FIG6 is a block diagram showing a network slice configuration system according to some further embodiments of the present disclosure.
- FIG. 7 is a block diagram illustrating a computer system for implementing some embodiments of the present disclosure.
- Figure 1 is a block diagram showing a network slice configuration system according to some embodiments of the present disclosure.
- the network slicing configuration system 1 includes a RAN (Radio Access Network) layer device 11 and an SDN (Software Defined Network) controller 13.
- RAN Radio Access Network
- SDN Software Defined Network
- the RAN layer device 11 is configured to generate service level information of the current time slot based on the user level information of the user layer device in the current time slot; predict the service demand information of the user in the current time slot using the first machine learning model based on the service level information of the current time slot; and send the service level information and service demand information of the current time slot to the SDN controller 13.
- the user level information includes at least one of a service demand pattern of a user, a mobility pattern of a user, location information of the user, and a random channel condition of the user.
- the RAN layer device is also configured to perform at least one of the following: parsing the user's service demand pattern to obtain the user's demand for latency, bandwidth and service content information as service level information; analyzing the user's mobility pattern to obtain the user's mobility pattern as service level information; obtaining the current channel signal-to-noise ratio based on the user's random channel conditions as service level information; and determining the user's density information based on the user's location information as service level information.
- the first machine learning model includes an LSTM (Long short-term memory) model.
- the RAN layer device includes an access point (AP) device of the RAN layer.
- AP access point
- the SDN controller 13 is configured to determine the network slice configuration policy of the current time slot based on the service level information and service demand information of the current time slot from the RAN layer device by using a second machine learning model; to verify the slice configuration of the network slice configuration policy of the current time slot by using a digital twin based on the user layer device and the RAN layer device; and to send the network slice configuration policy of the current time slot to the RAN layer device if the verification passes.
- the second machine learning model includes a deep reinforcement learning (DRL) model.
- DRL deep reinforcement learning
- the SDN controller 13 is further configured to send the network slice configuration policy of the current time slot to the digital twin; deploy and run the network slice configuration policy of the current time slot in the digital twin; obtain the estimated network slice performance after running the network slice configuration policy of the current time slot; and perform slice configuration verification on the network slice configuration policy of the current time slot according to the gain of the estimated network slice performance. When the gain of the estimated network slice performance exceeds the gain threshold, the verification is passed.
- the RAN layer device 11 is further configured to receive a network slice configuration policy for a current time slot from the SDN controller; run the network slice configuration policy for the current time slot; measure user feedback information on the network slice configuration policy for the current time slot, wherein the feedback information reflects the actual network slice performance under the network slice configuration policy for the current time slot; and send the feedback information to the SDN controller 13.
- the feedback information is used by the SDN controller to determine the network slice configuration policy for the next time slot of the current time slot.
- intelligent dynamic slice management can be implemented to improve resource utilization and the success rate of network slice configuration, and reduce adverse effects on the physical network.
- FIG2 is a block diagram showing a network slice configuration system according to some other embodiments of the present disclosure.
- FIG2 differs from FIG1 in that FIG2 shows other structures of the network slice configuration system of some other embodiments. Only the differences between FIG2 and FIG1 will be described below, and the similarities will not be repeated.
- the network slice configuration system 1 further includes a digital twin layer device 12 for deploying a digital twin.
- the digital twin layer device 12 is configured to receive real-time network information from user layer devices and RAN layer devices, and update the digital twin according to the real-time network information.
- the digital twin is a digital twin network of a physical network composed of user layer devices and RAN layer devices.
- the uplink between the RAN layer device and the digital twin layer device is composed of a high-bandwidth, high-speed data link, and the downlink between the RAN layer device and the digital twin layer device is a control channel.
- the network slice configuration system 1 also includes a user layer device 10.
- the user layer device 10 is configured to send user level information of the current time slot to the RAN layer device 11.
- DT Digital Twins
- the deep reinforcement learning agent in the SDN controller of the present invention learns the optimal network slicing strategy based on the observation of the real network environment, and forwards the slicing strategy to DT. Before the strategy is sent to the real network, it is performed in DT. Pre-verification avoids the harm caused by direct deployment into the network when the policy performance is unknown, reduces the cost of network slicing configuration, and improves the success rate of network slicing configuration.
- the present disclosure introduces digital twins into the network architecture, and through the design and construction of relevant network intelligent modules, it achieves the effect of predicting user needs and monitoring the performance of slicing strategies without affecting the physical network.
- the present disclosure proposes a wireless communication network architecture (SDN controller, RAN layer equipment, digital twin layer equipment, etc.) that supports digital twins.
- the architecture supports the construction of digital twins of wireless networks to simulate their complex environments and predict the dynamic characteristics of the network.
- Figure 3 is an architectural diagram showing a network slice configuration system according to some embodiments of the present disclosure.
- the architecture of the network slice configuration system also known as the 6G wireless network architecture with endogenous intelligence or the digital twin architecture for network slice configuration
- the user layer and the RAN layer constitute the physical network.
- the RAN layer uploads real-time information of the physical network to the digital twin layer, and the digital twin layer sends network slicing decisions or policies to the RAN layer.
- the RAN layer also uploads predicted user needs and network slice performance to the SDN controller layer at different time slots.
- the information transmitted by the SDN controller layer to the digital twin layer includes the optimized slicing strategy.
- the uplink between the RAN layer and the digital twin layer is composed of a high-bandwidth, high-speed data link, which can meet the needs of rapid upload of real-time network information to the digital twin layer.
- the downlink between the RAN layer and the digital twin layer is a control channel, which meets the needs of timely and accurate delivery of slicing strategies.
- the LSTM-based intelligent algorithm model on the RAN side is trained with data collected from the wireless network environment to learn user demand characteristics.
- the trained model can predict user demand in future time slots, and update the model and improve its performance in the process of continuous interaction with the wireless network environment (including uplink and downlink rates, the number of successful uplink and downlink connection accesses, the congestion of the current wireless network link, and other wireless network status).
- the DRL intelligent agent (DRL model) located at the SDN controller layer makes network slicing decisions based on the predicted data on user demand provided by the RAN side, and receives feedback information on network slicing performance in each time slot to complete its own optimization.
- the present disclosure proposes a network slicing architecture based on digital twins in a 6G network as shown in Figure 3.
- the UE user equipment
- the upper layer of the UE is the access network RAN layer.
- the RAN layer and the UE layer constitute a real physical network.
- the digital twin layer (DT layer) is a mapping of the real physical network. DT creates a digital twin of the physical network entity and network topology, simulates the complex environment of the physical network, and dynamically updates as the physical network changes.
- SDN Software Defined Network
- the top layer of the architecture proposed in this disclosure is the SDN controller.
- An agent based on the DRL intelligent algorithm is deployed in the SDN controller.
- the processed service level information from the RAN layer and the information on user demand prediction by the LSTM algorithm deployed at the RAN layer are received by the SDN controller and stored in the database for the DRL agent to extract data for training and generate the optimal network slicing strategy.
- Figure 4 is a flowchart showing a network slice configuration method according to some embodiments of the present disclosure.
- the network slice configuration method includes steps S410 to S460.
- step S410 the RAN layer device generates service level information according to the user level information of the user layer device in the current time slot.
- step S420 the RAN layer device predicts the service demand information of the user in the current time slot based on the service level information of the current time slot using the first machine learning model.
- step S430 the RAN layer device sends the service level information and service requirement information of the current time slot.
- step S440 the SDN controller determines the network slice configuration strategy for the current time slot based on the service level information and service demand information of the current time slot from the RAN layer device using the second machine learning model.
- step S450 the SDN controller uses the digital twin based on the user layer device and the RAN layer device to verify the slice configuration of the network slice configuration policy of the current time slot.
- step S460 if the verification is successful, the SDN controller sends the network slice configuration policy of the current time slot to the RAN layer device.
- each step in the network slice configuration method can refer to the contents of the aforementioned network slice configuration system and will not be repeated here.
- Figure 5 is a flowchart showing a network slice configuration method according to some other embodiments of the present disclosure.
- the network slice configuration method includes steps S501-S514.
- an access point device (AP) in the RAN layer collects and sends user-level random information from a UE (User Equipment) in the user layer to a RAN layer device.
- the user-level random information includes but is not limited to the end user's service demand pattern, mobility pattern, location information, and random channel conditions.
- the intelligent algorithm i.e., machine learning model deployed on the RAN side converts the user level information into the required service level information. For example, by analyzing the user service demand pattern, the user's demand for latency, bandwidth, and service content information can be obtained. By analyzing the mobility pattern, the user's mobility pattern can be obtained, which is convenient for subsequent prediction of user mobility. Through random channel conditions, the current channel signal-to-noise ratio can be obtained, which is convenient for combining the user's mobility pattern to realize the switching of user access channels and ensure the user's service quality.
- the processing of user location information can obtain user density information. After abstracting, integrating and analyzing the data of service level information, the user's service demand is predicted using intelligent algorithms.
- step S503 the DT layer initializes and generates a digital twin of the physical network consisting of the user layer and the RAN layer.
- step S504 the RAN layer transmits the processed service level information including the prediction of the user service demand to the SDN controller.
- step S505 the SDN controller runs a decision algorithm based on the DRL model according to the predicted service demand to make the optimal decision for network slicing based on the collected service level information (including predicted user service demand information).
- step S506 the SDN controller decides to send it to the DT layer for pre-verification.
- step S507 the planning decision determined by the DRL agent in the SDN controller is executed in the DT for pre-verification to avoid the harm caused by the poor performance of the AI model but directly deployed in the network.
- network slicing decisions are run in DT before being deployed in the real network. If the policy's gain in network slicing performance exceeds a certain threshold, the policy is sent to the RAN side. If the policy does not generate gains or has a bad impact on the performance of the network slice, it is not sent.
- step S508 the verified planning decision is sent back to all access points (APs) in the RAN layer.
- step S509 the access point (AP) in the RAN layer executes the received planning decision, for example, reserving network resources or allocating bandwidth for the corresponding slice.
- step S510 the users in service report their real-time information to the DT layer, such as channel conditions and corresponding service relationships between users and access points (APs).
- APs access points
- step S511 the DT layer is updated according to the real-time information in the wireless network environment.
- step S512 the RAN side evaluates and detects the performance of the network slice by measuring the user's feedback information on all slice decisions in the previous time slot.
- the feedback information includes a satisfaction rate.
- step S513 the access point (AP) in the RAN layer sends the slice performance to the SDN controller.
- step S514 the SDN controller makes a network slice planning decision for the next time slot and adjusts the planning strategy based on the feedback information.
- the present disclosure provides a 6G wireless network architecture with endogenous intelligence.
- a digital twin layer is introduced to enable efficient management of network slices.
- the key step of pre-verification is added to the full life cycle management of network slices to avoid the poor effect of AI models but the direct deployment of them into the network. the hazards.
- This disclosure also provides a detailed process for exchanging slice information in the digital twin architecture for network slices in 6G networks.
- Intelligent algorithms are introduced on the RAN side and the SDN controller side to intelligently manage network slices, and to assist slice management in a dynamic manner by interacting with the physical network, thereby improving resource utilization compared to static solutions.
- Figure 6 is a block diagram showing a network slice configuration system according to some further embodiments of the present disclosure.
- the network slice configuration system 6 includes a memory 61; and a processor 62 coupled to the memory 61.
- the memory 61 is used to store instructions for executing the corresponding embodiment of the network slice configuration method.
- the processor 62 is configured to execute the network slice configuration method in any of some embodiments of the present disclosure based on the instructions stored in the memory 61.
- the memory 61 and the processor 62 are located in the RAN layer device; when executing the method of the SDN controller, they are located in the SDN controller.
- FIG. 7 is a block diagram illustrating a computer system for implementing some embodiments of the present disclosure.
- Computer system 70 may be in the form of a general-purpose computing device.
- Computer system 70 includes a memory 710, a processor 720, and a bus 700 that connects various system components.
- the memory 710 may include, for example, a system memory, a non-volatile storage medium, etc.
- the system memory may store, for example, an operating system, an application, a boot loader, and other programs.
- the system memory may include a volatile storage medium, such as a random access memory (RAM) and/or a cache memory.
- the non-volatile storage medium may store, for example, instructions for executing at least one corresponding embodiment of the network slicing configuration method.
- the non-volatile storage medium includes, but is not limited to, a disk memory, an optical memory, a flash memory, and the like.
- the processor 720 can be implemented by a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistors, etc.
- DSP digital signal processor
- ASIC application-specific integrated circuit
- FPGA field programmable gate array
- each module such as the judgment module and the determination module can be implemented by a central processing unit (CPU) running instructions in a memory that execute corresponding steps, or can be implemented by a dedicated circuit that executes corresponding steps.
- Bus 700 may use any of a variety of bus architectures, including, but not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, and Peripheral Component Interconnect (PCI) bus.
- ISA Industry Standard Architecture
- MCA Micro Channel Architecture
- PCI Peripheral Component Interconnect
- the computer system 70 may also include an input/output interface 730, a network interface 740, a storage interface 750, etc. These interfaces 730, 740, 750, the memory 710, and the processor 720 may be connected via a bus 700.
- the input/output interface 730 may provide a connection interface for input/output devices such as a display, a mouse, and a keyboard.
- the network interface 740 provides a connection interface for various networked devices.
- the storage interface 750 provides a connection interface for external storage devices such as a floppy disk, a USB flash drive, and an SD card. The device provides a connection interface.
- a computer-readable storage medium stores computer program instructions, which, when executed by a processor, implement the steps of the method in the above embodiments.
- a computer program comprising: instructions, which, when executed by a processor, cause the processor to execute the network slice configuration method as described above.
- These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable device to produce a machine, so that the processor executes the instructions to produce means for implementing the functions specified in one or more blocks in the flowchart and/or block diagram.
- These computer-readable program instructions may also be stored in a computer-readable memory, which cause the computer to work in a specific manner to produce an article of manufacture, including instructions for implementing the functions specified in one or more blocks in the flowchart and/or block diagram.
- the present disclosure can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.
- intelligent dynamic slice management can be implemented, resource utilization and the success rate of network slice configuration can be improved, and adverse effects on the physical network can be reduced.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
本公开涉及网络切片配置方法及系统、计算机可存储介质,涉及无线通信技术领域。网络切片配置系统包括:RAN层设备,被配置为:根据用户层设备在当前时隙的用户级别信息,生成当前时隙的服务级别信息;根据当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息;发送当前时隙的服务级别信息和服务需求信息;SDN控制器,被配置为:根据来自RAN层设备的当前时隙的服务级别信息和服务需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略;利用基于用户层设备和RAN层设备的数字孪生体,对当前时隙的网络切片配置策略,进行切片配置验证;在验证通过的情况下,发送当前时隙的网络切片配置策略到RAN层设备。
Description
相关申请的交叉引用
本申请是以CN申请号为202211445997.2,申请日为2022年11月18日的申请为基础,并主张其优先权,该CN申请的公开内容在此作为整体引入本申请中。
本公开涉及无线通信技术领域,特别涉及网络切片配置方法及系统、计算机可存储介质。
6G网络需要满足多样化的服务,网络切片是支持这一愿景实现的潜在方法。网络切片是在公共物理网络之上,为不同服务构建多个逻辑隔离的虚拟网络。不同业务的QoS(Quality of Service,服务质量)要求,可以通过网络切片生命周期的准备、规划和运营阶段的经济高效的切片管理策略来保证。而6G网络需要支持各种新业务,同时满足其不同且严格的QoS要求,这进一步增加了切片管理的复杂性。
发明内容
根据本公开的第一方面,提供了一种网络切片配置系统,包括:无线接入网RAN层设备,被配置为:根据用户层设备在当前时隙的用户级别信息,生成当前时隙的服务级别信息;根据所述当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息;发送所述当前时隙的服务级别信息和服务需求信息;软件定义网络SDN控制器,被配置为:根据来自所述RAN层设备的当前时隙的服务级别信息和服务需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略;利用基于所述用户层设备和所述RAN层设备的数字孪生体,对所述当前时隙的网络切片配置策略,进行切片配置验证;在验证通过的情况下,发送所述当前时隙的网络切片配置策略到所述RAN层设备。
在一些实施例中,SDN控制器还被配置为:发送所述当前时隙的网络切片配置策略到所述数字孪生体;在所述数字孪生体,部署并运行所述当前时隙的网络切片配置策略;获取运行所述当前时隙的网络切片配置策略后的预估网络切片性能;根据所述
预估网络切片性能的增益,对所述当前时隙的网络切片配置策略,进行切片配置验证,其中,在所述预估网络切片性能的增益超过增益阈值的情况下,验证通过。
在一些实施例中,所述RAN层设备还被配置为:接收来自所述SDN控制器的当前时隙的网络切片配置策略;运行所述当前时隙的网络切片配置策略;测量用户对所述当前时隙的网络切片配置策略的反馈信息,其中,所述反馈信息反映所述当前时隙的网络切片配置策略下的实际网络切片性能;发送所述反馈信息到所述SDN控制器,用于所述SDN控制器确定所述当前时隙的下一时隙的网络切片配置策略。
在一些实施例中,网络切片配置系统,还包括:部署所述数字孪生体的数字孪生层设备,被配置为接收来自所述用户层设备和所述RAN层设备的实时网络信息,并根据所述实时网络信息,更新所述数字孪生体。
在一些实施例中,所述RAN层设备与所述数字孪生层设备之间的上行链路由具有高带宽的、高速率的数据链路构成,所述RAN层设备与所述数字孪生层设备之间的下行链路为控制信道。
在一些实施例中,所述用户级别信息包括用户的服务需求模式、用户的移动模式、用户的位置信息和用户的随机信道条件中的至少一种。
在一些实施例中,所述RAN层设备还被配置为执行以下中的至少一种:对所述用户的服务需求模式进行解析,得到用户对时延、带宽和服务内容信息的需求,作为服务级别信息;对所述用户的移动模式进行分析,得到用户的移动规律,作为服务级别信息;根据所述用户的随机信道条件,得到当前信道信噪比,作为服务级别信息;和根据所述用户的位置信息,确定用户的密度信息,作为服务级别信息。
在一些实施例中,所述第一机器学习模型包括长短期记忆网络LSTM模型;和/或所述第二机器学习模型包括深度强化学习模型。
在一些实施例中,网络切片配置系统,还包括:所述用户层设备,被配置为发送所述当前时隙的用户级别信息到所述RAN层设备。
在一些实施例中,所述RAN层设备包括RAN层的接入点设备。
根据本公开第二方面,提供了一种网络切片配置方法,包括:无线接入网RAN层设备根据用户层设备在当前时隙的用户级别信息,生成服务级别信息;RAN层设备根据所述当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息;RAN层设备发送所述当前时隙的服务级别信息和服务需求信息;软件定义网络SDN控制器根据来自所述RAN层设备的所述当前时隙的服务级别信息和服务
需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略;SDN控制器利用基于所述用户层设备和所述RAN层设备的数字孪生体,对所述当前时隙的网络切片配置策略,进行切片配置验证;SDN控制器在验证通过的情况下,发送所述当前时隙的网络切片配置策略到所述RAN层设备。
根据本公开第三方面,提供了一种网络切片配置系统,包括:存储器;以及耦接至所述存储器的处理器,所述处理器被配置为基于存储在所述存储器的指令,执行上述任一实施例所述的网络切片配置方法。
根据本公开的第四方面,提供了一种计算机可存储介质,其上存储有计算机程序指令,该指令被处理器执行时实现上述任一实施例所述的网络切片配置方法。
通过以下参照附图对本公开的示例性实施例的详细描述,本公开的其它特征及其优点将会变得清楚。
构成说明书的一部分的附图描述了本公开的实施例,并且连同说明书一起用于解释本公开的原理。
参照附图,根据下面的详细描述,可以更加清楚地理解本公开,其中:
图1是示出根据本公开一些实施例的网络切片配置系统的框图;
图2是示出根据本公开另一些实施例的网络切片配置系统的框图;
图3是示出根据本公开一些实施例的网络切片配置系统的架构示意图;
图4是示出根据本公开一些实施例的网络切片配置方法的流程图;
图5是示出根据本公开另一些实施例的网络切片配置方法的流程图;
图6是示出根据本公开再一些实施例的网络切片配置系统的框图;
图7是示出用于实现本公开一些实施例的计算机系统的框图。
现在将参照附图来详细描述本公开的各种示例性实施例。应注意到:除非另外具体说明,否则在这些实施例中阐述的部件和步骤的相对布置、数字表达式和数值不限制本公开的范围。
同时,应当明白,为了便于描述,附图中所示出的各个部分的尺寸并不是按照实际的比例关系绘制的。
以下对至少一个示例性实施例的描述实际上仅仅是说明性的,决不作为对本公开及其应用或使用的任何限制。
对于相关领域普通技术人员已知的技术、方法和设备可能不作详细讨论,但在适当情况下,所述技术、方法和设备应当被视为说明书的一部分。
在这里示出和讨论的所有示例中,任何具体值应被解释为仅仅是示例性的,而不是作为限制。因此,示例性实施例的其它示例可以具有不同的值。
应注意到:相似的标号和字母在下面的附图中表示类似项,因此,一旦某一项在一个附图中被定义,则在随后的附图中不需要对其进行进一步讨论。
图1是示出根据本公开一些实施例的网络切片配置系统的框图。
如图1所示,网络切片配置系统1包括RAN(Radio Access Network,无线接入网)层设备11和SDN(Software Defined Network,软件定义网络)控制器13。
RAN层设备11被配置为根据用户层设备在当前时隙的用户级别信息,生成当前时隙的服务级别信息;根据当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息;并发送当前时隙的服务级别信息和服务需求信息到SDN控制器13。
在一些实施例中,用户级别信息包括用户的服务需求模式、用户的移动模式、用户的位置信息和用户的随机信道条件中的至少一种。
在一些实施例中,RAN层设备还被配置为执行以下中的至少一种:对用户的服务需求模式进行解析,得到用户对时延、带宽和服务内容信息的需求,作为服务级别信息;对用户的移动模式进行分析,得到用户的移动规律,作为服务级别信息;根据用户的随机信道条件,得到当前信道信噪比,作为服务级别信息;和根据用户的位置信息,确定用户的密度信息,作为服务级别信息。
在一些实施例中,第一机器学习模型包括LSTM(Long short-term memory,长短期记忆网络)模型。
在一些实施例中,RAN层设备包括RAN层的接入点(Access Point,AP)设备。
SDN控制器13被配置为根据来自RAN层设备的当前时隙的服务级别信息和服务需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略;利用基于用户层设备和RAN层设备的数字孪生体,对当前时隙的网络切片配置策略,进行切片配置验证;并在验证通过的情况下,发送当前时隙的网络切片配置策略到RAN层设备。
在一些实施例中,第二机器学习模型包括深度强化学习(Deep Reinforcement Learning,DRL)模型。
在一些实施例中,SDN控制器13还被配置为发送当前时隙的网络切片配置策略到数字孪生体;在数字孪生体,部署并运行当前时隙的网络切片配置策略;获取运行当前时隙的网络切片配置策略后的预估网络切片性能;根据预估网络切片性能的增益,对当前时隙的网络切片配置策略,进行切片配置验证。在预估网络切片性能的增益超过增益阈值的情况下,验证通过。
在一些实施例中,RAN层设备11还被配置为接收来自SDN控制器的当前时隙的网络切片配置策略;运行当前时隙的网络切片配置策略;测量用户对当前时隙的网络切片配置策略的反馈信息,其中,反馈信息反映当前时隙的网络切片配置策略下的实际网络切片性能;发送反馈信息到SDN控制器13。反馈信息用于SDN控制器确定当前时隙的下一时隙的网络切片配置策略。
上述实施例中,可以实现智能化的动态切片管理,提升资源利用率和网络切片配置的成功率,减少对物理网络的不良影响。
图2是示出根据本公开另一些实施例的网络切片配置系统的框图。图2与图1的不同之处在于,图2示出了另一些实施例的网络切片配置系统的其他结构。下面将仅描述图2与图1的不同之处,相同之处不再赘述。
在一些实施例中,如图2所示,网络切片配置系统1还包括部署数字孪生体的数字孪生层设备12。数字孪生层设备12被配置为接收来自用户层设备和RAN层设备的实时网络信息,并根据实时网络信息,更新数字孪生体。例如,数字孪生体为由用户层设备和RAN层设备构成的物理网络的数字孪生网络。
在一些实施例中,RAN层设备与数字孪生层设备之间的上行链路由具有高带宽的、高速率的数据链路构成,RAN层设备与数字孪生层设备之间的下行链路为控制信道。
在一些实施例中,如图2所示,网络切片配置系统1还包括用户层设备10。用户层设备10被配置为发送当前时隙的用户级别信息到RAN层设备11。
DT(Digital Twins,数字孪生)生成网络的数字孪生体,从真实物理网络接收网络拓扑结构、网络状态观测、用户与基站对应服务关系、用户设备动态QoS要求等信息,本公开SDN控制器中的深度强化学习代理基于对真实网络环境的观察学习最优的网络切片策略,并将切片策略转发给DT,在策略下发到真实网络之前在DT中进行
预验证,避免在策略性能未知的情况下直接部署到网络中带来危害,降低网络切片配置的成本,提高网络切片配置的成功率。
本公开将数字孪生引入网络架构中,并通过相关网络智能化模块的设计、构建,达到预测用户需求的效果,并在不影响物理网络的情况下监控切片策略的性能。另外,本公开提出支持数字孪生的无线通信网络架构(SDN控制器、RAN层设备、数字孪生层设备等),该架构支持构建无线网络的数字孪生体,以模拟其复杂环境并预测网络的动态特性。
图3是示出根据本公开一些实施例的网络切片配置系统的架构示意图。
如图3所示,网络切片配置系统的架构,也称为具有智能内生的6G无线网络架构或用于网络切片配置的数字孪生架构,具有从下到上分别为用户层、RAN层、数字孪生层和SDN控制器层的四层结构。用户层和RAN层构成了物理网络。RAN层向数字孪生层上传物理网络的实时信息,数字孪生层向RAN层下发网络切片决策或策略。RAN层在不同的时隙还会向SDN控制器层分别上传预测的用户需求以及网络切片的性能。SDN控制器层向数字孪生层传递的信息包括优化后的切片策略。
RAN层与数字孪生层之间的上行链路由具有高带宽的、高速率的数据链路构成,能够满足网络实时信息快速上传至数字孪生层,RAN层与数字孪生层之间的下行链路为控制信道,满足切片策略及时准确下发的需求。
位于RAN侧的基于LSTM的智能算法模型采用从无线网络环境中收集的数据进行训练,学习用户需求特征。训练后的模型可以对未来时隙内用户需求进行预测,并在与无线网络环境(包括上下行速率、上下行连接成功接入数、当前无线网络链路的拥塞情况等无线网络的状态)不断交互的过程中,更新模型,提升模型的性能。位于SDN控制器层的DRL智能代理(DRL模型)根据RAN侧提供的对用户需求的预测数据,进行网络切片决策,并在每个时隙接收对网络切片性能的反馈信息,完成自身优化。
为了实现网络切片的智能高效管理,本公开提出了图3所示的6G网络中基于数字孪生的网络切片架构。位于底层的是6G网络中的用户设备(User Equipment,UE),根据6G网络提出的万物互联的愿景,UE有着多样的接入方式、个性化的QoS需求。UE的上一层为接入网RAN层,RAN层与UE层构成了真实的物理网络。数字孪生层(DT层)是真实物理网络的映射,DT创造了物理网络实体与网络拓扑的数字孪生体,模拟物理网络的复杂环境并跟随物理网络的变化动态更新。软件定义网络(SDN)
与网络切片结合有助于实现切片的灵活管理,本公开所提架构的最上层为SDN控制器,在SDN控制器中部署了基于DRL智能算法的代理,来自RAN层的经过处理的服务级别的信息和由部署于RAN层的LSTM算法对用户需求预测的信息,由SDN控制器接收并存储在数据库中,供DRL代理从中抽取数据进行训练,生成最优网络切片策略。
图4是示出根据本公开一些实施例的网络切片配置方法的流程图。
如图4所示,网络切片配置方法包括步骤S410-步骤S460。
在步骤S410中,RAN层设备根据用户层设备在当前时隙的用户级别信息,生成服务级别信息。
在步骤S420中,RAN层设备根据当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息。
在步骤S430中,RAN层设备发送当前时隙的服务级别信息和服务需求信息。
在步骤S440中,SDN控制器根据来自RAN层设备的当前时隙的服务级别信息和服务需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略。
在步骤S450中,SDN控制器利用基于用户层设备和RAN层设备的数字孪生体,对当前时隙的网络切片配置策略,进行切片配置验证。
在步骤S460中,SDN控制器在验证通过的情况下,发送当前时隙的网络切片配置策略到RAN层设备。
网络切片配置方法中每个步骤的相关实施例可以参考前述网络切片配置系统的内容,此处不再赘述。
图5是示出根据本公开另一些实施例的网络切片配置方法的流程图。
如图5所示,网络切片配置方法包括步骤S501-步骤S514。
在步骤S501中,RAN层中的接入点设备(AP)从用户层的UE(User Equipment,用户设备)收集并发送用户级别的随机信息到RAN层设备。例如,用户级别的随机信息包括但不限于最终用户的服务需求模式、移动模式、位置信息和随机信道条件。
在步骤S502中,部署在RAN侧的智能算法(即机器学习模型)将用户级别信息转换为所需的服务级别信息。例如,通过对用户服务需求模式的解析,可以得到用户对时延、带宽和服务内容信息的需求。通过对移动模式的分析,可以得到用户的移动规律,便于后续对用户移动的预测。通过随机信道条件,可以获得当前信道信噪比,便于结合用户的移动规律,实现用户接入信道的切换,保证用户的服务质量。通过对
用户位置信息的处理,可以得到用户密度信息。对服务级别信息的数据进行抽象、融合、分析后,利用智能算法对用户的服务需求进行预测。
在步骤S503中,DT层初始化生成由用户层和RAN层构成的物理网络的数字孪生体。
在步骤S504中,RAN层将处理后的包含对用户服务需求预测的服务级别信息被传送到SDN控制器。
在步骤S505中,SDN控制器根据预测的服务需求,运行基于DRL模型的决策算法,以根据收集的服务级别信息(包括预测的用户服务需求信息)做出网络切片的最优决策。
在步骤S506中,SDN控制器决策下发到DT层,用于预验证。
在步骤S507中,由SDN控制器中DRL代理确定的规划决策在DT中执行进行预验证,避免AI模型效果不佳但直接部署到网络中带来的危害。
在一些实施例中,针对AI模型可解释性差,训练效果未知的情况,网络切片决策在部署到真实网络中之前先在DT中运行。若该策略对网络切片性能的增益超过某一阈值,则将策略下发至RAN侧,若该策略对网络切片的性能不能产生增益或带来恶劣影响,则不下发。
在步骤S508中,通过验证的规划决策被发送回RAN层中的所有接入点(AP)。
在步骤S509中,RAN层中的接入点(AP)执行接收到的规划决策。例如,为相应的切片保留网络资源或分配带宽。
在步骤S510中,服务中的用户向DT层报告它们的实时信息。例如,实时信息包括信道条件和用户与接入点(AP)之间的对应服务关系等。
在步骤S511中,DT层根据无线网络环境中的实时信息,进行更新。
在步骤S512中,RAN侧通过测量用户对上一个时隙中所有切片决策的反馈信息,对网络切片的性能进行评估与检测。例如,反馈信息包括满意率。
在步骤S513中,RAN层中的接入点(AP)将切片性能发送至SDN控制器。
在步骤S514中,SDN控制器对下一个时隙做出网络切片规划决策,并根据反馈信息调整规划策略。
本公开提供了具有智能内生的6G无线网络架构,针对AI模型可解释性差,训练效果未知的情况,引入数字孪生层赋能网络切片的高效管理,在网络切片的全生命周期管理中增加预验证这一关键步骤,避免AI模型效果不佳但直接部署到网络中带来
的危害。
本公开还提供了面向6G网络中网络切片的数字孪生架构中切片信息交换的详细流程。在RAN侧与SDN控制器侧引入智能算法,智能管理网络切片,通过与物理网络交互,以动态的方式协助切片管理,相较于静态方案,提升资源利用率。
图6是示出根据本公开再一些实施例的网络切片配置系统的框图。
如图6所示,网络切片配置系统6包括存储器61;以及耦接至该存储器61的处理器62。存储器61用于存储执行网络切片配置方法对应实施例的指令。处理器62被配置为基于存储在存储器61中的指令,执行本公开中任意一些实施例中的网络切片配置方法。存储器61和处理器62在执行RAN层设备的方法时,位于RAN层设备;在执行SDN控制器的方法时,位于SDN控制器。
图7是示出用于实现本公开一些实施例的计算机系统的框图。
如图7所示,计算机系统70可以通用计算设备的形式表现。计算机系统70包括存储器710、处理器720和连接不同系统组件的总线700。
存储器710例如可以包括系统存储器、非易失性存储介质等。系统存储器例如存储有操作系统、应用程序、引导装载程序(Boot Loader)以及其他程序等。系统存储器可以包括易失性存储介质,例如随机存取存储器(RAM)和/或高速缓存存储器。非易失性存储介质例如存储有执行网络切片配置方法中的至少一种的对应实施例的指令。非易失性存储介质包括但不限于磁盘存储器、光学存储器、闪存等。
处理器720可以用通用处理器、数字信号处理器(DSP)、应用专用集成电路(ASIC)、现场可编程门阵列(FPGA)或其它可编程逻辑设备、分立门或晶体管等分立硬件组件方式来实现。相应地,诸如判断模块和确定模块的每个模块,可以通过中央处理器(CPU)运行存储器中执行相应步骤的指令来实现,也可以通过执行相应步骤的专用电路来实现。
总线700可以使用多种总线结构中的任意总线结构。例如,总线结构包括但不限于工业标准体系结构(ISA)总线、微通道体系结构(MCA)总线、外围组件互连(PCI)总线。
计算机系统70还可以包括输入输出接口730、网络接口740、存储接口750等。这些接口730、740、750以及存储器710和处理器720之间可以通过总线700连接。输入输出接口730可以为显示器、鼠标、键盘等输入输出设备提供连接接口。网络接口740为各种联网设备提供连接接口。存储接口750为软盘、U盘、SD卡等外部存储
设备提供连接接口。
在另一些实施例中,一种计算机可读存储介质,其上存储有计算机程序指令,该指令被处理器执行时实现上述实施例中的方法的步骤。
在本公开的一些实施例中,还提供了一种计算机程序,包括:指令,所述指令当由处理器执行时使所述处理器执行如前所述的网络切片配置方法。
这里,参照根据本公开实施例的方法、装置和计算机程序产品的流程图和/或框图描述了本公开的各个方面。应当理解,流程图和/或框图的每个框以及各框的组合,都可以由计算机可读程序指令实现。
这些计算机可读程序指令可提供到通用计算机、专用计算机或其他可编程装置的处理器,以产生一个机器,使得通过处理器执行指令产生实现在流程图和/或框图中一个或多个框中指定的功能的装置。
这些计算机可读程序指令也可存储在计算机可读存储器中,这些指令使得计算机以特定方式工作,从而产生一个制造品,包括实现在流程图和/或框图中一个或多个框中指定的功能的指令。
本公开可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。
通过上述实施例中的网络切片配置方法及系统、计算机可存储介质,可以实现智能化的动态切片管理,提升资源利用率和网络切片配置的成功率,减少对物理网络的不良影响。
至此,已经详细描述了根据本公开的网络切片配置方法及系统、计算机可存储介质。为了避免遮蔽本公开的构思,没有描述本领域所公知的一些细节。本领域技术人员根据上面的描述,完全可以明白如何实施这里公开的技术方案。
Claims (15)
- 一种网络切片配置系统,包括:无线接入网RAN层设备,被配置为:根据用户层设备在当前时隙的用户级别信息,生成当前时隙的服务级别信息;根据所述当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息;发送所述当前时隙的服务级别信息和服务需求信息;软件定义网络SDN控制器,被配置为:根据来自所述RAN层设备的当前时隙的服务级别信息和服务需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略;利用基于所述用户层设备和所述RAN层设备的数字孪生体,对所述当前时隙的网络切片配置策略,进行切片配置验证;在验证通过的情况下,发送所述当前时隙的网络切片配置策略到所述RAN层设备。
- 根据权利要求1所述的网络切片配置系统,其中,所述SDN控制器还被配置为:发送所述当前时隙的网络切片配置策略到所述数字孪生体;在所述数字孪生体,部署并运行所述当前时隙的网络切片配置策略;获取运行所述当前时隙的网络切片配置策略后的预估网络切片性能;根据所述预估网络切片性能的增益,对所述当前时隙的网络切片配置策略,进行切片配置验证,其中,在所述预估网络切片性能的增益超过增益阈值的情况下,验证通过。
- 根据权利要求1或2所述的网络切片配置系统,其中,所述RAN层设备还被配置为:接收来自所述SDN控制器的当前时隙的网络切片配置策略;运行所述当前时隙的网络切片配置策略;测量用户对所述当前时隙的网络切片配置策略的反馈信息,其中,所述反馈信息 反映所述当前时隙的网络切片配置策略下的实际网络切片性能;发送所述反馈信息到所述SDN控制器,用于所述SDN控制器确定所述当前时隙的下一时隙的网络切片配置策略。
- 根据权利要求1所述的网络切片配置系统,还包括:部署所述数字孪生体的数字孪生层设备,被配置为接收来自所述用户层设备和所述RAN层设备的实时网络信息,并根据所述实时网络信息,更新所述数字孪生体。
- 根据权利要求4所述的网络切片配置系统,其中,所述RAN层设备与所述数字孪生层设备之间的上行链路由具有高带宽的、高速率的数据链路构成,所述RAN层设备与所述数字孪生层设备之间的下行链路为控制信道。
- 根据权利要求1所述的网络切片配置系统,其中,所述用户级别信息包括用户的服务需求模式、用户的移动模式、用户的位置信息和用户的随机信道条件中的至少一种。
- 根据权利要求6所述的网络切片配置系统,其中,所述RAN层设备还被配置为执行以下中的至少一种:对所述用户的服务需求模式进行解析,得到用户对时延、带宽和服务内容信息的需求,作为服务级别信息;对所述用户的移动模式进行分析,得到用户的移动规律,作为服务级别信息;根据所述用户的随机信道条件,得到当前信道信噪比,作为服务级别信息;和根据所述用户的位置信息,确定用户的密度信息,作为服务级别信息。
- 根据权利要求1所述的网络切片配置系统,其中,所述第一机器学习模型包括长短期记忆网络LSTM模型。
- 根据权利要求1或8所述的网络切片配置系统,其中,所述第二机器学习模型包括深度强化学习模型。
- 根据权利要求1所述的网络切片配置系统,还包括:所述用户层设备,被配置为发送所述当前时隙的用户级别信息到所述RAN层设备。
- 根据权利要求1所述的网络切片配置系统,其中,所述RAN层设备包括RAN层的接入点设备。
- 一种网络切片配置方法,包括:无线接入网RAN层设备根据用户层设备在当前时隙的用户级别信息,生成服务级别信息;RAN层设备根据所述当前时隙的服务级别信息,利用第一机器学习模型,预测用户在当前时隙的服务需求信息;RAN层设备发送所述当前时隙的服务级别信息和服务需求信息;软件定义网络SDN控制器根据来自所述RAN层设备的所述当前时隙的服务级别信息和服务需求信息,利用第二机器学习模型,确定当前时隙的网络切片配置策略;SDN控制器利用基于所述用户层设备和所述RAN层设备的数字孪生体,对所述当前时隙的网络切片配置策略,进行切片配置验证;SDN控制器在验证通过的情况下,发送所述当前时隙的网络切片配置策略到所述RAN层设备。
- 一种网络切片配置系统,包括:存储器;以及耦接至所述存储器的处理器,所述处理器被配置为基于存储在所述存储器的指令,执行如权利要求12所述的网络切片配置方法。
- 一种计算机可存储介质,其上存储有计算机程序指令,该指令被处理器执行时实现如权利要求12所述的网络切片配置方法。
- 一种计算机程序,包括:指令,所述指令当由处理器执行时使所述处理器执行根据权利要求12所述的网 络切片配置方法。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211445997.2 | 2022-11-18 | ||
CN202211445997.2A CN115843050B (zh) | 2022-11-18 | 2022-11-18 | 网络切片配置方法及系统、计算机可存储介质 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2024103960A1 true WO2024103960A1 (zh) | 2024-05-23 |
Family
ID=85575686
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2023/120348 WO2024103960A1 (zh) | 2022-11-18 | 2023-09-21 | 网络切片配置方法及系统、计算机可存储介质 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN115843050B (zh) |
WO (1) | WO2024103960A1 (zh) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115843050B (zh) * | 2022-11-18 | 2024-06-28 | 中国电信股份有限公司 | 网络切片配置方法及系统、计算机可存储介质 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109218047A (zh) * | 2017-06-30 | 2019-01-15 | 中国移动通信有限公司研究院 | 网络切片处理方法及装置、通信系统及存储介质 |
CN111817876A (zh) * | 2020-06-04 | 2020-10-23 | 北京思特奇信息技术股份有限公司 | 基于行业客户需求的5g网络切片管理方法 |
CN112152855A (zh) * | 2020-09-25 | 2020-12-29 | 广州爱浦路网络技术有限公司 | 网络切片实例资源动态调整的方法 |
CN113133087A (zh) * | 2019-12-30 | 2021-07-16 | 中国移动通信集团浙江有限公司 | 针对终端设备配置网络切片的方法及装置 |
CN114666222A (zh) * | 2020-12-08 | 2022-06-24 | 中兴通讯股份有限公司 | 网络切片的运维方法、装置、系统和存储介质 |
CN115250500A (zh) * | 2021-04-28 | 2022-10-28 | 中国移动通信集团北京有限公司 | 承载网络的切片分析方法 |
CN115843050A (zh) * | 2022-11-18 | 2023-03-24 | 中国电信股份有限公司 | 网络切片配置方法及系统、计算机可存储介质 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10608895B2 (en) * | 2017-03-31 | 2020-03-31 | At&T Intellectual Property I, L.P. | Quality of service management for dynamic instantiation of network slices and/or applications |
US10601932B2 (en) * | 2017-06-09 | 2020-03-24 | At&T Intellectual Property I, L.P. | Next generation mobility core network controller for service delivery |
CN112382064B (zh) * | 2020-11-12 | 2023-01-20 | 广东电网有限责任公司 | 一种基于数字孪生技术的电力物联网故障预警方法及系统 |
US20220116335A1 (en) * | 2021-12-21 | 2022-04-14 | Vesh Raj Sharma Banjade | End-to-end network slicing (ens) from ran to core network for next generation (ng) communications |
-
2022
- 2022-11-18 CN CN202211445997.2A patent/CN115843050B/zh active Active
-
2023
- 2023-09-21 WO PCT/CN2023/120348 patent/WO2024103960A1/zh unknown
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109218047A (zh) * | 2017-06-30 | 2019-01-15 | 中国移动通信有限公司研究院 | 网络切片处理方法及装置、通信系统及存储介质 |
CN113133087A (zh) * | 2019-12-30 | 2021-07-16 | 中国移动通信集团浙江有限公司 | 针对终端设备配置网络切片的方法及装置 |
CN111817876A (zh) * | 2020-06-04 | 2020-10-23 | 北京思特奇信息技术股份有限公司 | 基于行业客户需求的5g网络切片管理方法 |
CN112152855A (zh) * | 2020-09-25 | 2020-12-29 | 广州爱浦路网络技术有限公司 | 网络切片实例资源动态调整的方法 |
CN114666222A (zh) * | 2020-12-08 | 2022-06-24 | 中兴通讯股份有限公司 | 网络切片的运维方法、装置、系统和存储介质 |
CN115250500A (zh) * | 2021-04-28 | 2022-10-28 | 中国移动通信集团北京有限公司 | 承载网络的切片分析方法 |
CN115843050A (zh) * | 2022-11-18 | 2023-03-24 | 中国电信股份有限公司 | 网络切片配置方法及系统、计算机可存储介质 |
Also Published As
Publication number | Publication date |
---|---|
CN115843050A (zh) | 2023-03-24 |
CN115843050B (zh) | 2024-06-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Fan et al. | Workload allocation in hierarchical cloudlet networks | |
US11523288B2 (en) | Data analysis apparatus, system, and method | |
Chih-Lin et al. | The big-data-driven intelligent wireless network: architecture, use cases, solutions, and future trends | |
WO2024103960A1 (zh) | 网络切片配置方法及系统、计算机可存储介质 | |
WO2022001918A1 (zh) | 构建预测模型的方法、装置、计算设备和存储介质 | |
US20210042578A1 (en) | Feature engineering orchestration method and apparatus | |
CN105814845A (zh) | 软件定义网络中的流量工程框架 | |
CN109088755B (zh) | 一种边缘计算下的复杂事件处理系统部署方法 | |
Quan et al. | Cybertwin-driven DRL-based adaptive transmission scheduling for software defined vehicular networks | |
Wang et al. | Computing aware scheduling in mobile edge computing system | |
Hlophe et al. | QoS provisioning and energy saving scheme for distributed cognitive radio networks using deep learning | |
Xu et al. | Flow-level QoE of video streaming in wireless networks | |
CN114945926A (zh) | 用于传输速率控制的网络实体和用户设备 | |
Nouruzi et al. | Online service provisioning in NFV-enabled networks using deep reinforcement learning | |
US11722429B1 (en) | Resource allocation for extended reality applications | |
US20230060623A1 (en) | Network improvement with reinforcement learning | |
Ren et al. | End-to-end network SLA quality assurance for C-RAN: a closed-loop management method based on digital twin network | |
Alfa et al. | A discrete time queueing model of cognitive radio networks with multi-modal overlay/underlay switching service levels | |
Kocak et al. | Fuzzy logic-based performance improvement on MAC layer in wireless local area networks | |
Ravindran et al. | Efficient service allocation scheduling algorithms for 5g user equipments in slice-in-slice networks | |
Wieselthier et al. | Ordinal optimization of admission control in wireless multihop voice/data networks via standard clock simulation | |
Falahatraftar et al. | A multiple linear regression model for predicting congestion in heterogeneous vehicular networks | |
Mamadou et al. | DURATS: Distributed user-centric radio access technology selection framework | |
WO2020245948A1 (ja) | 要求通信品質推定装置、要求通信品質推定方法、及びプログラム | |
Moura et al. | Estimating quality of service on Wi-Fi stations using recurrent neural networks |
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: 23890397 Country of ref document: EP Kind code of ref document: A1 |