WO2023024219A1 - 云边协同网络中时延和频谱占用联合优化方法及系统 - Google Patents

云边协同网络中时延和频谱占用联合优化方法及系统 Download PDF

Info

Publication number
WO2023024219A1
WO2023024219A1 PCT/CN2021/122981 CN2021122981W WO2023024219A1 WO 2023024219 A1 WO2023024219 A1 WO 2023024219A1 CN 2021122981 W CN2021122981 W CN 2021122981W WO 2023024219 A1 WO2023024219 A1 WO 2023024219A1
Authority
WO
WIPO (PCT)
Prior art keywords
spectrum
cloud
user request
delay
occupancy
Prior art date
Application number
PCT/CN2021/122981
Other languages
English (en)
French (fr)
Inventor
陈伯文
刘玲
梁瑞鑫
王守翠
陈琪
沈纲祥
高明义
邵卫东
陈虹
Original Assignee
苏州大学
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 苏州大学 filed Critical 苏州大学
Priority to US17/915,257 priority Critical patent/US11968122B2/en
Publication of WO2023024219A1 publication Critical patent/WO2023024219A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/11—Identifying congestion
    • 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
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/24—Traffic characterised by specific attributes, e.g. priority or QoS
    • H04L47/2491—Mapping quality of service [QoS] requirements between different networks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W72/00—Local resource management
    • H04W72/04—Wireless resource allocation
    • H04W72/044—Wireless resource allocation based on the type of the allocated resource
    • H04W72/0453—Resources in frequency domain, e.g. a carrier in FDMA
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W72/00—Local resource management
    • H04W72/50—Allocation or scheduling criteria for wireless resources
    • H04W72/54—Allocation or scheduling criteria for wireless resources based on quality criteria
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the present invention relates to the technical field of cloud-edge collaborative network optimization, in particular to a joint optimization method and system for time delay and spectrum occupation in cloud-edge collaborative network.
  • MEC Mobile Edge Computing
  • Computing offloading refers to the technology of uploading the computing data of terminal equipment to the cloud and performing a series of computing processing.
  • In the information age of the Internet of Everything in order to achieve low-latency transmission data, low energy consumption of servers, and high storage of mobile terminal resources, it is necessary to offload complex computing tasks to network edge servers for computing processing.
  • cloud-edge collaboration As a new computing model has become a new research trend.
  • cloud computing With the increase of data-intensive applications and computing-intensive applications, it is necessary to use the powerful computing power of cloud computing, as well as the response characteristics of communication resources and short-term transmission of edge computing to realize and complete corresponding application requests.
  • edge computing and cloud computing collaboration By working together and showing their strengths, the value of edge computing and cloud computing collaboration is maximized, thereby effectively improving the performance of applications.
  • most of the research on cloud-edge collaboration focuses on application scenarios in many fields such as the Internet of Things, industrial Internet, intelligent transportation, and security monitoring. The main purpose is to reduce delay, reduce energy consumption, and improve user experience quality.
  • offloading to the cloud allows users to offload computing-intensive tasks to a cloud server with powerful resources for processing; offloading to the edge is to deploy cloud services at the edge of the network. Offloading to the cloud will not be able to handle delay-sensitive applications well due to the long transmission distance. Offloading to the edge also needs to consider factors such as computing resources, storage resources, energy consumption, and delay. Therefore, how to offload services and where to One end has become a current research hotspot.
  • the technical problem to be solved by the present invention is to overcome the problems of high user request end-to-end delay and spectrum resource occupancy in the prior art, thereby providing a method that effectively reduces user request end-to-end delay and spectrum resource occupancy.
  • a joint optimization method for time delay and spectrum occupancy in a cloud-edge collaborative network of the present invention includes the following steps: initializing the cloud-edge collaborative network, generating a set of user requests; The objective function of time delay and the minimum occupied spectrum slot; based on the objective function, in the process of processing each user request, it is sequentially judged whether the node and path selection uniqueness constraints, mobile edge computing server load constraints, spectrum resource occupancy and If the uniqueness constraint, the spectrum continuity constraint and the spectrum consistency constraint are all satisfied, the user request is successfully processed, and the process goes to step S4; if any one is not satisfied, the user request processing fails; calculate the average end-to-end time of the user request delay and spectrum resource occupancy.
  • the computing resource of the edge computing server is initialized, and the spectrum flexible optical network is initialized.
  • the objective function includes a primary optimization objective and a secondary optimization objective.
  • the formula of the objective function is: Where
  • the node and path selection uniqueness constraints are: in is a binary variable, if the user request (u, s) is processed at the MEC server node e and the user request is transmitted through the kth path, the value of the binary variable is 1, otherwise the value is 0, and y (u, s) is a binary variable , if the user request (u, s) is migrated to the second layer, that is, the MEC server node e in the cloud area, the value of the binary variable is 1, otherwise the value is 0.
  • the MEC server load constraint is: where R (u, s) represents the number of computing resources required for user request (u, s) transmission, Indicates the maximum load of the server, V e indicates the maximum computing resource capacity of the MEC server node e.
  • the spectrum resource occupancy and uniqueness constraints are:
  • the spectrum continuity constraint is:
  • ⁇ represents the total spectrum of the entire network, which is equal to the product of the total number of links and the capacity of the link spectrum slots.
  • the spectrum consistency constraint is: where F (u,s) represents the number of spectrum slots required for user request (u,s) transmission.
  • the present invention also provides a method for joint optimization of time delay and spectrum occupancy in a cloud-edge collaborative network, including: an initialization module for initializing the cloud-edge collaborative network and generating a set of user requests; a modeling module for establishing user request The objective function of the lowest average end-to-end delay and the minimum occupied spectrum slot; the judging module is used to judge whether the node and path selection uniqueness constraints, mobile If all of the edge computing server load constraints, spectrum resource occupancy and uniqueness constraints, spectrum continuity constraints, and spectrum consistency constraints are satisfied, the user request is successfully processed and step S4 is entered; if any item is not satisfied, the user request is processed Failed; the calculation module is used to calculate the average end-to-end delay and spectrum resource occupancy rate of user requests.
  • the method and system for joint optimization of time delay and spectrum occupancy in the cloud-edge collaborative network mainly solve the problem of how to select a suitable MEC server to process user requests in the cloud-edge collaborative network. Because cloud computing takes too long to process services, it cannot meet the low-latency requirements of services, while edge computing deploys servers on the user side, but the computing resources are insufficient to handle services with large amounts of data. Therefore, cloud computing and edge computing Combined, the cloud-edge collaborative network is proposed as an effective way to process business.
  • the present invention proposes an evaluation mechanism for end-to-end time delay and spectrum resource occupancy, and then establishes a joint optimization method with the minimum user request average end-to-end time delay and the minimum occupied spectrum slot as the objective function according to this mechanism, and the integer linear programming method To realize the resource allocation method of computing offloading, routing selection and spectrum allocation of cloud-edge collaborative network. Generate a set of user request sets in the static cloud-edge collaborative network, and set the corresponding computing resources and spectrum resource requirements, and then according to the constraints and optimization objectives, establish the optimization objective method with the lowest end-to-end delay and spectrum occupancy, so as to provide All user requests find the best MEC server to process and allocate resources.
  • the present invention can select the optimal MEC server to process user requests, greatly reducing the data processing delay and data transmission delay generated by processing user requests, thereby improving user service quality; at the same time, it also finds the shortest working path for each user request, The waste of spectrum resources in the network is reduced, and the utilization rate of spectrum resources is greatly improved.
  • Fig. 1 is a flowchart of the joint optimization method for time delay and spectrum occupation in the cloud-edge collaborative network of the present invention
  • Fig. 2 is a six-node network topology diagram of the present invention.
  • Fig. 3 is a schematic diagram of service processing in the cloud-edge collaborative network of the present invention.
  • this embodiment provides a joint optimization method for delay and spectrum occupancy in a cloud-edge collaborative network, including: step S1: initialize the cloud-edge collaborative network, and generate a set of user requests; step S2: establish a user request The objective function of the average end-to-end minimum delay and the minimum occupied spectrum slot; Step S3: Based on the objective function, in the process of processing each user request, sequentially determine whether the node and path selection uniqueness constraints are satisfied, and the mobile edge computing If the server load constraint, spectrum resource occupancy and uniqueness constraint, spectrum continuity constraint, and spectrum consistency constraint are all satisfied, the user request is processed successfully, and step S4 is entered; if any item is not satisfied, the user request processing fails; Step S4: Calculate the average end-to-end delay and spectrum resource occupancy rate of user requests.
  • the cloud-edge collaborative network is initialized to generate a set of user requests, thereby facilitating the establishment of an objective function; in the step S2, Establishing an objective function with the lowest average end-to-end delay and the minimum occupied spectrum slot requested by the user is conducive to the realization of an optimization scheme targeting the lowest delay and spectrum occupancy; in the step S3, based on the objective function, in the processing In the process of each user request, it is sequentially judged whether the node and path selection uniqueness constraints, mobile edge computing server load constraints, spectrum resource occupancy and uniqueness constraints, spectrum continuity constraints, and spectrum consistency constraints are met.
  • step S4 If all are satisfied, then If the user request is successfully processed, go to step S4; if any item is not satisfied, the user request processing fails, which is conducive to realizing the computing offload of the cloud-edge collaborative network, routing selection and resource allocation of spectrum allocation, and reducing the user request end-to-end Time delay and occupancy rate of spectrum resources; in the step S4, calculate the average end-to-end time delay and spectrum resource occupancy rate of user requests, so as to find the best MEC server for all user requests to process and allocate resources, the present invention can choose
  • the optimal MEC server processes user requests, greatly reducing the data processing delay and data transmission delay generated by processing user requests, thereby improving user service quality; at the same time, it also finds the shortest working path for each user request, reducing the frequency spectrum in the network. The waste of resources greatly improves the utilization rate of spectrum resources.
  • the end-to-end delay of user requests in the cloud-edge collaborative network is mainly composed of network transmission delay and computing resource delay.
  • the network transmission delay refers to the shortest path length between the user's service area and the edge computing server, which is calculated by the cumulative delay of the link delay; the calculation resource delay is related to each user's computing resource requirements and the edge computing server's related to computing power.
  • the present invention mainly considers three types of end-to-end time delays for processing user requests, which are local processing, offloading to other areas connected to switches, and offloading to cloud areas connected to switches.
  • the average end-to-end delay of user requests is as follows:
  • represents the total number of user requests, and CR is a set of user requests; is a binary variable, if the user request (u, s) is processed at the MEC server node e and the user request is transmitted through the kth path, E represents a set of MEC server sets, the value of the binary variable is 1, otherwise the value is 0; R (u, s) indicates the number of computing resources required for user request (u, s) transmission; M e is the computing resource capability of MEC server node e; Indicates the distance between the link (k, l) where the user request (u, s) is transmitted from the node s to the MEC server node e through the kth working path, K indicates the set of k paths; c indicates that the user request is in the fiber link The transmission rate in the road is set to 3x10 5 km/s; y (u, s) is a binary variable, if the user requests (u, s) to migrate to the MEC server node e
  • Spectrum resource occupancy refers to the ratio of the total number of spectrum slots occupied by the working paths requested by all users divided by the spectrum slots of all links.
  • the specific calculation formula is as follows:
  • F (u, s) represents the spectrum resource requirement requested by user u
  • CR is a set of user requests
  • LN and SN represent the total number of links and the spectrum resource capacity of each link, respectively.
  • the present invention proposes an integer linear programming method on the basis of the above-mentioned time delay and spectrum occupancy evaluation mechanism, that is, in a static network, the lowest time delay is achieved. and spectrum occupancy as the goal of the optimization scheme.
  • the computing resource of the edge computing server is initialized, and the spectrum flexible optical network is initialized.
  • u represents the serial number of the user request, and s represents the source node that generates the user request.
  • the objective function of joint optimization minimizes the average end-to-end delay of user requests and the occupancy rate of spectrum resources in the cloud-edge collaborative network, that is, the objective function consists of two parts: the main optimization objective and the secondary optimization objective, and can be adjusted by adjusting the parameter ⁇ and ⁇ (0 ⁇ , ⁇ 1), change the weight of the optimization objective, so as to achieve different optimization purposes.
  • the optimization objective function can be expressed by the following formula:
  • the objective G of the integer linear programming model is to minimize the average end-to-end delay and the number of occupied spectrum slots in the cloud-edge collaborative network.
  • the first part represents the average end-to-end delay of user requests.
  • the specific evaluation method is shown in formula (1), and the user’s processing delay and transmission delay are reduced by optimizing the selection of an appropriate server; the second The part represents the total number of spectrum slots occupied in the cloud-edge collaborative network.
  • the user request (u, s) is transmitted to the MEC server node e and the spectrum gap numbered f on the link (k, l) is occupied, then the value of the binary variable is 1, otherwise the value is 0.
  • step S3 based on the objective function, a constraint condition satisfying the objective function optimization method is established.
  • nodes and links need to meet the following conditions:
  • the node and path selection uniqueness constraints are:
  • y (u, s) is a binary variable. If the user request (u, s) is migrated to the second layer, that is, the MEC server node e in the cloud area, the value of the binary variable is 1, otherwise the value is 0.
  • Formulas (4) and (5) guarantee that a user request can only be processed by one server, and each user request will select one of the k working paths to transmit the user request.
  • the load constraint of the mobile edge computing server is:
  • R (u, s) in formula (6) represents the number of computing resources required for user request (u, s) transmission, Indicates the maximum load of the server.
  • V e in formula (7) represents the maximum computing resource capacity of MEC server node e.
  • Formulas (6) and (7) can ensure that the total amount of computing resources requested by each MEC server node cannot exceed the maximum load of the MEC node, and the maximum load of the MEC node cannot exceed the node's computing resource capacity.
  • Formulas (9) and (10) ensure that the number of spectrum slots occupied by the user request (u, s) transmitted to the MEC server node in the fiber link (k, l) is equal to the number of spectrum slots required by the user request, and each link The spectrum slot f of can only be occupied by one user request.
  • ⁇ represents the total spectrum of the entire network, which is equal to the product of the total number of links and the capacity of the link spectrum slots, as shown in formula (13).
  • the spectrum slot allocated to each fiber link must be continuous.
  • formula (11) if and All spectrum gaps with index values higher than f+1 will not allocate spectrum resources on the fiber link (k,l).
  • formula (12) if Then no spectrum gap with an index value lower than f will be allocated to the fiber link (k, l). So formulas (11) and (12) ensure the constraint of spectral continuity.
  • the constraint condition (14) can ensure the consistency of spectrum resources, that is, for each user request, the number of spectrum slots occupied by each link passed by the working path is the same.
  • the MEC server node e is represented by a circle, numbered 0, 1, 2, 3, 4, 5, where The server node is node number 0.
  • the numbers in the dotted circles next to the nodes indicate the computing resource capacity of the server.
  • the nodes are connected by optical fiber links, and each optical fiber link is bidirectional, and the numbers on the links represent the length of the link (in km).
  • each link includes 20 spectrum slot resources.
  • corresponding server and spectrum resources can be assigned to the user requests CR 1 (0,5) and CR 2 (1,1) in the cloud-edge collaborative network based on the target conditions.
  • the present invention considers both the time delay and the spectrum slot when selecting server nodes to process user requests.
  • a grid on the link represents a spectrum slot
  • a red grid indicates that the spectrum slot is occupied
  • a white grid indicates that the spectrum slot is in an idle state, that is, it can be used to allocate spectrum resources.
  • the user request CR 1 (0,5) is transmitted to node 4 for processing, and the user request CR 2 (1,1) is transmitted to node 3 for processing, and the spectrum slot of the corresponding working path turns green. It can be seen from Figure 3 that the above constraints can be satisfied, and the number of hops in the working path is only one hop, and the spectrum resource occupation and delay are relatively low.
  • this embodiment provides a joint optimization system for time delay and spectrum occupancy in a cloud-edge collaborative network. The place will not be repeated.
  • This embodiment provides a joint optimization system for delay and spectrum occupancy in a cloud-edge collaborative network, including:
  • the initialization module is used to initialize the cloud-edge collaborative network and generate a set of user requests
  • the modeling module is used to establish the objective function of the lowest average end-to-end delay and the smallest occupied spectrum slot based on the user request;
  • a judging module configured to sequentially judge whether the node and path selection uniqueness constraints, mobile edge computing server load constraints, spectrum resource occupancy and uniqueness constraints, and spectrum continuity are satisfied in the process of processing each user request based on the objective function. If all constraints and spectrum consistency constraints are satisfied, the user request processing is successful, and the process goes to step S4; if any item is not satisfied, the user request processing fails;
  • the calculation module is used to calculate the average end-to-end delay and spectrum resource occupancy rate of user requests.
  • the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
  • computer-usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • 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 operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions
  • the device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Quality & Reliability (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明涉及一种云边协同网络中时延和频谱占用联合优化方法及系统。所述方法包括：S1：初始化云边协同网络，生成一组用户请求；S2：建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；S3：基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；S4：计算用户请求平均端到端时延和频谱资源占用率。本发明有利于提高频谱资源的利用率以及提升用户服务质量。

Description

云边协同网络中时延和频谱占用联合优化方法及系统 技术领域
本发明涉及云边协同网络优化的技术领域，尤其是指一种云边协同网络中时延和频谱占用联合优化方法及系统。
背景技术
近年来，随着物联网的高速发展，人脸识别、智慧城市、智能交通系统、虚拟现实技术(VR)等大量新兴应用程序场景应运而生。由于这些应用程序场景可以为用户带来更高的带宽速率、更多的网络连接和更低的时延，导致核心网络需要在单位时间内处理的数据大小和业务请求呈指数型增长。因此，随着移动设备和物联网设备承载的网络流量的增加，云计算所采用的集中式处理模式因离终端设备比较远，已不能满足用户的日常需求，并且对时延和能耗等性能要求较高的计算服务来说，这种模式会引起高延时、网络堵塞等问题。
对于这些问题，欧洲电信标准协会(European Telecommunications Standards Institute,简称ETSI)通过将边缘计算融合到移动网络的架构，提出了移动边缘计算(Mobile Edge Computing,简称MEC)。MEC是指将移动终端的计算任务卸载到网络边缘处，在网络边缘执行计算和存储的一种新型计算模型。MEC被认为是蜂窝基站模型现代化演变和5G技术发展的关键因素。MEC将计算和存储资源引入到移动网络的边缘，降低终端设备的计算时延和能耗，提升用户对移动互联网应用的体验质量并减轻了云计算中心高负载情况。与此同时，MEC也需要计算卸载技术的支撑，计算卸载是指将终端设备的计算数据上传至云中并进行一系列计算处理的技术。在万物互联的信息时代下，要想实现传输数据低延时、服务器低能耗、移动终端资源高存储这些情况，需要将复杂的计算任务卸载到网络边缘服务器上进行计算处 理。
为了更好地结合云计算与边缘计算的优势，云边协同作为一种新型计算模式成为了新的研究趋势。随着数据密集型应用与计算密集型应用的增加，需要利用云计算强大的计算能力，以及通信资源与边缘计算短时传输的响应特性来实现，并完成相应的应用请求。通过两者协同工作、各展所长，将边缘计算和云计算协作的价值最大化，从而有效地提高应用程序的性能。目前，针对云边协同的研究大多数集中在物联网、工业互联网、智能交通、安全监控等诸多领域的应用场景上，主要目的是减少时延、降低能耗、以及提高用户体验质量等。
目前主要有两种资源卸载方法：卸载到云端和卸载到边缘端。其中，卸载到云端允许用户将计算密集型任务卸载到资源强大的云服务器上进行处理；卸载到边缘端是在网络边缘部署云服务。卸载到云端会因传输距离远而无法很好地处理时延敏感性应用，卸载到边缘端也需要考虑计算资源、存储资源、能量消耗以及时延等等因素，因此如何卸载业务以及卸载到哪一端成为目前的研究热点。
发明内容
为此，本发明所要解决的技术问题在于克服现有技术中用户请求端到端时延和频谱资源的占用率高的问题，从而提供一种有效降低用户请求端到端时延和频谱资源占用率的云边协同网络中时延和频谱占用联合优化方法及系统。
为解决上述技术问题，本发明的一种云边协同网络中时延和频谱占用联合优化方法，包括如下步骤：初始化云边协同网络，生成一组用户请求；建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；计算用户请求 平均端到端时延和频谱资源占用率。
在本发明的一个实施例中，初始化云边协同网络时，对边缘计算服务器的计算资源进行初始化，以及对频谱灵活光网络进行初始化。
在本发明的一个实施例中，所述目标函数包括主要优化目标和次要优化目标。
在本发明的一个实施例中，所述目标函数的公式为：
Figure PCTCN2021122981-appb-000001
其中|CR|表示用户请求的总个数,CR为一组用户请求集合；
Figure PCTCN2021122981-appb-000002
是二进制变量，如果用户请求(u,s)在MEC服务器节点e处理且通过第k条路径传输用户请求，E表示一组MEC服务器集合，则二进制变量的值为1，否则值为0；R (u,s)表示用户请求(u,s)传输所需的计算资源数；M e为MEC服务器节点e的计算资源能力；
Figure PCTCN2021122981-appb-000003
表示用户请求(u,s)从节点s传输到MEC服务器节点e经过第k条工作路径的链路(k,l)之间的距离
Figure PCTCN2021122981-appb-000004
Figure PCTCN2021122981-appb-000005
K表示k条路径集合；c表示用户请求在光纤链路中的传输速率，设置为3ⅹ10 5km/s；y (u,s)为二进制变量，如果用户请求(u,s)迁移到第二层即云区域的MEC服务器节点e处理，则二进制变量的值为1，否则值为0；τ为用户请求通过交换机传输到云区域的额外交换时延，α和β为参数，
Figure PCTCN2021122981-appb-000006
是二进制变量，用户请求(u,s)输到MEC服务器节点e且链路(k,l)编号f的频谱间隙被占用，则二进制变量的值为1，否则值为0。
在本发明的一个实施例中，所述节点及路径选择唯一性约束为：
Figure PCTCN2021122981-appb-000007
其中
Figure PCTCN2021122981-appb-000008
是二进制变量，如果用户请求(u,s)在MEC服务器节点e处理且通过第k条路径传输用户请求，则二进制变量的值为1，否则值为0，y (u,s)为二进制变量，如果用户请求(u,s) 迁移到第二层，即云区域的MEC服务器节点e处理，则二进制变量的值为1，否则值为0。
在本发明的一个实施例中，所述MEC服务器负载约束为：
Figure PCTCN2021122981-appb-000009
其中R (u,s)表示用户请求(u,s)传输所需的计算资源数，
Figure PCTCN2021122981-appb-000010
表示服务器的最大负载，V e表示MEC服务器节点e的最大计算资源容量。
在本发明的一个实施例中，所述频谱资源占用及唯一性约束为：
Figure PCTCN2021122981-appb-000011
Figure PCTCN2021122981-appb-000012
Figure PCTCN2021122981-appb-000013
其中，|F (k,l)|表示链路(k,l)上的最大频谱间隙数，假设每条链路所能提供的最大频谱间隙数相同，
Figure PCTCN2021122981-appb-000014
表示用户请求(u,s)从节点s传输到MEC服务器节点e的第k条工作路径经过光纤链路(k,l)，F (u,s)表示用户请求(u,s)传输所需的频谱间隙数。
在本发明的一个实施例中，所述频谱连续性约束为：
Figure PCTCN2021122981-appb-000015
Figure PCTCN2021122981-appb-000016
θ＝|F (k,l)|×|L|；
其中，θ表示整个网络的频谱总数，等于链路总数与链路频谱隙容量的乘积。在本发明的一个实施例中，所述频谱一致性约束为：
Figure PCTCN2021122981-appb-000017
其中，F (u,s)表示用户请求(u,s)传输所需的 频谱间隙数。
本发明还提供了一种云边协同网络中时延和频谱占用联合优化方法，包括：初始化模块，用于初始化云边协同网络，生成一组用户请求；建模模块，用于建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；判断模块，用于基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；计算模块，用于计算用户请求平均端到端时延和频谱资源占用率。
本发明的上述技术方案相比现有技术具有以下优点：
本发明所述的云边协同网络中时延和频谱占用联合优化方法及系统，主要解决云边协同网络中如何选取合适的MEC服务器处理用户请求的问题。由于云计算对业务的处理时间过长，无法满足业务低时延的需求，而边缘计算虽然将服务器部署在用户侧，但计算资源不足无法处理数据量大的业务，因此将云计算与边缘计算相结合，提出云边协同网络作为处理业务的有效途径。本发明提出了端到端时延以及频谱资源占用的评估机制，然后根据这个机制建立以最低用户请求平均端到端时延和最小占用频谱隙为目标函数的联合优化方法，以整数线性规划方法来实现云边协同网络的计算卸载、路由选择和频谱分配的资源分配方法。在静态云边协同网络中生成一组用户请求集合，并设置相应的计算资源和频谱资源需求，然后根据约束条件和优化目标，建立端到端时延和频谱占用最低的优化目标方法，从而为所有用户请求找到最佳的MEC服务器处理并分配资源。
本发明可以选择最优的MEC服务器处理用户请求，极大地降低处理用户请求产生的数据处理时延和数据传输时延，从而提升用户服务质量；同时也为每个用户请求寻找最短的工作路径，降低网络中频谱资源的浪费，大大提高了频谱资源的利用率。
附图说明
为了使本发明的内容更容易被清楚的理解，下面根据本发明的具体实施例并结合附图，对本发明作进一步详细的说明，其中
图1是本发明云边协同网络中时延和频谱占用联合优化方法的流程图；
图2是本发明六节点网络拓扑图；
图3是本发明云边协同网络中的业务处理示意图。
具体实施方式
实施例一
如图1所示，本实施例提供一种云边协同网络中时延和频谱占用联合优化方法，包括：步骤S1：初始化云边协同网络，生成一组用户请求；步骤S2：建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；步骤S3：基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；步骤S4：计算用户请求平均端到端时延和频谱资源占用率。
本实施例所述云边协同网络中时延和频谱占用联合优化方法，所述步骤S1中，初始化云边协同网络，生成一组用户请求，从而有利于建立目标函数；所述步骤S2中，建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数，有利于实现以最低时延和频谱占用为目标的优化方案；所述步骤S3中，基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败，有利于实现云边协同网络的计算卸载、路由选择和频谱分配的资源分配，以及降低用户请求端到端时延和频谱资源的占用率；所述步骤S4中，计算用户请求平均端到端时延和频谱资源占用率，从而为所有用户请求找到最佳的MEC服务器处理并分配资源，本发明可以选择最优的MEC服务器处 理用户请求，极大地降低处理用户请求产生的数据处理时延和数据传输时延，从而提升用户服务质量；同时也为每个用户请求寻找最短的工作路径，降低网络中频谱资源的浪费，大大提高了频谱资源的利用率。
本发明中，云边协同网络中用户请求的端到端时延主要由网络传输时延和计算资源时延组成。网络传输时延是指用户的服务区域与边缘计算服务器之间的最短路径长度，以其链路时延的累计时延计算；计算资源时延与每个用户的计算资源需求和边缘计算服务器的计算能力有关。本发明主要考虑三种处理用户请求情况的端到端时延，分别是本地处理、卸载到交换机相连的其他区域以及卸载到交换机相连的云区域。用户请求的平均端到端时延如公式如下：
Figure PCTCN2021122981-appb-000018
其中|CR|表示用户请求的总个数，CR为一组用户请求集合；
Figure PCTCN2021122981-appb-000019
是二进制变量，如果用户请求(u,s)在MEC服务器节点e处理且通过第k条路径传输用户请求，E表示一组MEC服务器集合，则二进制变量的值为1，否则值为0；R (u,s)表示用户请求(u,s)传输所需的计算资源数；M e为MEC服务器节点e的计算资源能力；
Figure PCTCN2021122981-appb-000020
表示用户请求(u,s)从节点s传输到MEC服务器节点e经过第k条工作路径的链路(k,l)之间的距离，K表示k条路径集合；c表示用户请求在光纤链路中的传输速率，设置为3ⅹ10 5km/s；y (u,s)为二进制变量，如果用户请求(u,s)迁移到第二层即云区域的MEC服务器节点e处理，则二进制变量的值为1，否则值为0；τ为用户请求通过交换机传输到云区域的额外交换时延。
频谱资源占用率是指所有用户请求的工作路径占用的频谱隙总数除以所有链路频谱隙的比例，具体计算公式如下：
Figure PCTCN2021122981-appb-000021
其中，F (u,s)代表用户请求u的频谱资源需求，CR为一组用户请求集合，LN和SN分别表示链路总数以及每条链路的频谱资源容量。
为了解决云边协同网络中如何选择合适的服务器处理业务的问题，本发明在上述时延和频谱占用评估机制的基础上，提出了整数线性规划方法，即在静态网络中，实现以最低时延和频谱占用为目标的优化方案。
所述步骤S1中，初始化云边协同网络时，对边缘计算服务器的计算资源进行初始化，以及对频谱灵活光网络进行初始化。在云边协同网络G(CR，E)中，其中CR＝{1,2,...,(u,s),...}表示一组用户请求，E＝{1,2,...,e,...}表示一组边缘计算服务器节点。每一个用户请求CR(u,s)∈CR,u表示用户请求的编号，s表示产生用户请求的源节点。
所述步骤S2中，建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数时，由于本发明主要解决云边协同网络中如何选择合适的边缘计算服务器处理业务的问题，因此联合优化的目标函数使云边协同网络中用户请求平均端到端时延和频谱资源占用率最小化，即该目标函数由主要优化目标和次要优化目标两部分组成，并可通过调节参数α和β(0≤α,β≤1)的大小，改变优化目标的权重，从而实现不同的优化目的。当α＝1和β＝0时，优化目标变为实现网络中平均端到端时延的最小值；当α＝0和β＝1时，优化目标用于优化网络中频谱资源的占用率，实现网络中频谱利用的最优化。
优化目标函数可用如下式子表示：
最小化
Figure PCTCN2021122981-appb-000022
其中，该整数线性规划模型的目标G是使云边协同网络中的平均端到端时延和占用的频谱隙个数达到最小化。在公式(3)中，第一部分表示用户请求的平均端到端时延，具体评估方法如公式(1)所示，通过优化选择合适的服务器降低用户的处理时延和传输时延；第二部分代表云边协同网络中占用的频谱隙总个数，通过优化
Figure PCTCN2021122981-appb-000023
来减少连接请求占用频谱隙数目，
Figure PCTCN2021122981-appb-000024
是二 进制变量，用户请求(u,s)传输到MEC服务器节点e且链路(k,l)上编号f的频谱间隙被占用，则二进制变量的值为1，否则值为0。
所述步骤S3中，基于所述目标函数，建立满足目标函数优化方法的约束条件。
在给用户请求分配合理的MEC服务器处理时，节点和链路需要满足以下条件：
所述节点及路径选择唯一性约束为：
Figure PCTCN2021122981-appb-000025
Figure PCTCN2021122981-appb-000026
其中，
Figure PCTCN2021122981-appb-000027
是二进制变量，如果用户请求(u,s)在MEC服务器节点e处理且通过第k条路径传输用户请求，则二进制变量的值为1，否则值为0。y (u,s)为二进制变量，如果用户请求(u,s)迁移到第二层，即云区域的MEC服务器节点e处理，则二进制变量的值为1，否则值为0。公式(4)和(5)保证了一个用户请求只能由一个服务器处理，并且每个用户请求都会选择k条工作路径中的一条路径来传输用户请求。
所述移动边缘计算服务器负载约束为：
Figure PCTCN2021122981-appb-000028
Figure PCTCN2021122981-appb-000029
公式(6)中R (u,s)表示用户请求(u,s)传输所需的计算资源数，
Figure PCTCN2021122981-appb-000030
表示服务器的最大负载。公式(7)中V e表示MEC服务器节点e的最大计算资源容量。公式(6)和(7)能够确保每个MEC服务器节点处理的用户请求的计算资源总量不能超过MEC节点的最大负载，并且MEC节点的最大负载不能超过节点的计算资源容量。
所述频谱资源占用及唯一性约束为：
Figure PCTCN2021122981-appb-000031
Figure PCTCN2021122981-appb-000032
Figure PCTCN2021122981-appb-000033
其中，|F (k,l)|表示链路(k,l)上的最大频谱间隙数，假设每条链路所能提供的最大频谱间隙数相同，
Figure PCTCN2021122981-appb-000034
表示用户请求(u,s)从节点s传输到MEC服务器节点e的第k条工作路径经过光纤链路(k,l)，F (u,s)表示用户请求(u,s)传输所需的频谱间隙数。公式(8)确保每条光纤链路占用的频谱隙的数量不能超过这条链路频谱隙的总数量。公式(9)和(10)确保传输到MEC服务器节点的用户请求(u，s)在光纤链路(k,l)占用的频谱隙数等于用户请求所需频谱隙数，并且每条链路的频谱隙f只能被一个用户请求占用。
所述频谱连续性约束为：
Figure PCTCN2021122981-appb-000035
Figure PCTCN2021122981-appb-000036
θ＝|F (k,l)|×|L|      (13)
其中，θ表示整个网络的频谱总数，等于链路总数与链路频谱隙容量的乘积，如公式(13)所示。在工作路径上，每条光纤链路分配的频谱隙必须具有连续性。在公式(11)中，如果
Figure PCTCN2021122981-appb-000037
并且
Figure PCTCN2021122981-appb-000038
所有索引值高于f+1的频谱间隙都不会在光纤链路(k,l)上分配频谱资源。在公式(12)中，如果
Figure PCTCN2021122981-appb-000039
则索引值低于f的频谱间隙都不会被分配给光纤链路(k，l)。所以公式(11)和(12)确保了频谱连续性的约束。
所述频谱一致性约束为：
Figure PCTCN2021122981-appb-000040
所述约束条件(14)能够保证频谱资源的一致性，即对于每个用户请求，工作路径经过的每条链路占用的频谱隙的编号是相同的。
通过以上约束条件，可以找出基于云边协同网络的计算卸载、路由选择和频谱分配的资源分配方法，从而实现该发明整数线性规划的联合优化目标函数。
为了进一步理解该发明中提出的优化方法，下面结合相关的实例对本发明中具体的实施方法进行详细阐述，具体实例步骤如下所示：
以图2所示网络拓扑图为例，是一个6节点、8条链路的云边协同网络，MEC服务器节点e用圆圈表示，编号为0、1、2、3、4、5，其中云服务器节点为节点编号0。节点旁边虚线圆圈内的数字表示服务器具有的计算资源容量。节点之间用光纤链路连接，每条光纤链路都是双向的，链路上的数字代表链路的长度(单位为km)。这里，设置每条链路包含20个频谱隙资源。
在云边协同网络中生成一组用户请求集合CR＝{CR 1(0,5),CR 2(1,1)}，用户请求的计算资源和频谱资源需求分别为R (0,5)＝5、R (1,1)＝4、F (0,5)＝2和F (1,1)＝3。
确立并执行该发明中提出的最低用户请求平均端到端时延和最小占用频谱隙为目标函数，参考公式(3)。
确立并执行该发明中提出的基于云边协同网络中时延和频谱占用联合优化方法的约束条件。在处理每个用户请求的过程中，需要满足节点及路径选择唯一性约束条件，参考公式4和公式5、MEC服务器负载约束条件，参考公式6和公式7、频谱资源占用及唯一性约束条件参考公式8至公式10、频谱连续性约束条件，参考公式11至公式13、频谱一致性约束，参考公式14。
经过上述步骤，即可基于目标条件下为云边协同网络中的用户请求CR 1(0,5)和CR 2(1,1)分配相应的服务器和频谱资源。为了达到最低用户请求平均端到端时延和最小占用频谱隙的目标，本发明在选择服务器节点处理用户请求时同时考虑时延和频谱隙。如图3所示，链路上一个格子表示一个频谱隙，红色的格子表示频谱隙已被占用，白色的格子表示频谱隙处于空闲状态， 即可以用来分配频谱资源。根据上述的约束条件和目标函数，将用户请求CR 1(0,5)传输到节点4处理，用户请求CR 2(1,1)传输到节点3处理，对应工作路径频谱隙变为绿色。从图3可以看出，以上的约束条件都可以满足，并且工作路径跳数只有一跳，频谱资源占用和时延都相对较低。
实施例二
基于同一发明构思，本实施例提供了一种云边协同网络中时延和频谱占用联合优化系统，其解决问题的原理与所述云边协同网络中时延和频谱占用联合优化方法类似，重复之处不再赘述。
本实施例提供一种云边协同网络中时延和频谱占用联合优化系统，包括：
初始化模块，用于初始化云边协同网络，生成一组用户请求；
建模模块，用于建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；
判断模块，用于基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；
计算模块，用于计算用户请求平均端到端时延和频谱资源占用率。
本领域内的技术人员应明白，本申请的实施例可提供为方法、系统、或计算机程序产品。因此，本申请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。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本申请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器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器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，上述实施例仅仅是为清楚地说明所作的举例，并非对实施方式的限定。对于所属领域的普通技术人员来说，在上述说明的基础上还可以做出其它不同形式变化或变动。这里无需也无法对所有的实施方式予以穷举。而由此所引伸出的显而易见的变化或变动仍处于本发明创造的保护范围之中。

Claims (10)

  1. 一种云边协同网络中时延和频谱占用联合优化方法，其特征在于，包括如下步骤：
    步骤S1：初始化云边协同网络，生成一组用户请求；
    步骤S2：建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；
    步骤S3：基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；
    步骤S4：计算用户请求平均端到端时延和频谱资源占用率。
  2. 根据权利要求1所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：初始化云边协同网络时，对边缘计算服务器的计算资源进行初始化，以及对频谱灵活光网络进行初始化。
  3. 根据权利要求1所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述目标函数包括主要优化目标和次要优化目标。
  4. 根据权利要求1或3所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述目标函数的公式为：
    Figure PCTCN2021122981-appb-100001
    其中|CR|表示用户请求的总个数,CR为一组用户请求集合；
    Figure PCTCN2021122981-appb-100002
    是二进制变量，如果用户请求(u,s)在MEC服务器节点e处理且通过第k条路径传输用户请求，E表示一组MEC服务器集合，则二进制变量的值为1，否则值为0；R (u,s)表示用 户请求(u,s)传输所需的计算资源数；M e为MEC服务器节点e的计算资源能力；
    Figure PCTCN2021122981-appb-100003
    表示用户请求(u,s)从节点s传输到MEC服务器节点e经过第k条工作路径的链路(k,l)之间的距离
    Figure PCTCN2021122981-appb-100004
    K表示k条路径集合；c表示用户请求在光纤链路中的传输速率，设置为3ⅹ10 5km/s；y (u,s)为二进制变量，如果用户请求(u,s)迁移到第二层即云区域的MEC服务器节点e处理，则二进制变量的值为1，否则值为0；τ为用户请求通过交换机传输到云区域的额外交换时延，α和β为参数，
    Figure PCTCN2021122981-appb-100005
    是二进制变量，用户请求(u,s)输到MEC服务器节点e且链路(k,l)编号f的频谱间隙被占用，则二进制变量的值为1，否则值为0。
  5. 根据权利要求4所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述节点及路径选择唯一性约束为：
    Figure PCTCN2021122981-appb-100006
    其中，
    Figure PCTCN2021122981-appb-100007
    是二进制变量，如果用户请求(u,s)在MEC服务器节点e处理且通过第k条路径传输用户请求，则二进制变量的值为1，否则值为0，y (u,s)为二进制变量，如果用户请求(u,s)迁移到第二层，即云区域的MEC服务器节点e处理，则二进制变量的值为1，否则值为0。
  6. 根据权利要求4所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述MEC服务器负载约束为：
    Figure PCTCN2021122981-appb-100008
    Figure PCTCN2021122981-appb-100009
    其中R (u,s)表示用户请求(u,s)传输所需的计算资源数，
    Figure PCTCN2021122981-appb-100010
    表示服务器的最大负载，V e表示MEC服务器节点e的最大计算资源容量。
  7. 根据权利要求4所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述频谱资源占用及唯一性约束为：
    Figure PCTCN2021122981-appb-100011
    Figure PCTCN2021122981-appb-100012
    Figure PCTCN2021122981-appb-100013
    其中，|F (k,l)|表示链路(k,l)上的最大频谱间隙数，假设每条链路所能提供的最大频谱间隙数相同，
    Figure PCTCN2021122981-appb-100014
    表示用户请求(u,s)从节点s传输到MEC服务器节点e的第k条工作路径经过光纤链路(k,l)，F (u,s)表示用户请求(u,s)传输所需的频谱间隙数。
  8. 根据权利要求7所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述频谱连续性约束为：
    Figure PCTCN2021122981-appb-100015
    Figure PCTCN2021122981-appb-100016
    θ＝|F (k,l)|×|L|；其中θ表示整个网络的频谱总数，等于链路总数与链路频谱隙容量的乘积。
  9. 根据权利要求4所述的云边协同网络中时延和频谱占用联合优化方法，其特征在于：所述频谱一致性约束为：
    Figure PCTCN2021122981-appb-100017
    其中，F (u,s)表示用户请求(u,s)传输所需的频谱间隙数。
  10. 一种云边协同网络中时延和频谱占用联合优化系统，其特征在于，包括：
    初始化模块，用于初始化云边协同网络，生成一组用户请求；
    建模模块，用于建立以用户请求平均端到端的最低时延和最小占用频谱隙的目标函数；
    判断模块，用于基于所述目标函数，在处理每个用户请求的过程中，依次判断是否满足节点及路径选择唯一性约束、移动边缘计算服务器负载约束、频谱资源占用及唯一性约束、频谱连续性约束以及频谱一致性约束，若均满足，则用户请求处理成功，进入步骤S4；若有任意一项不满足，则用户请求处理失败；
    计算模块，用于计算用户请求平均端到端时延和频谱资源占用率。
PCT/CN2021/122981 2021-08-24 2021-10-11 云边协同网络中时延和频谱占用联合优化方法及系统 WO2023024219A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/915,257 US11968122B2 (en) 2021-08-24 2021-10-11 Joint optimization method and system for delay and spectrum occupation in cloud-edge collaborative network

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110975856.0 2021-08-24
CN202110975856.0A CN113784373B (zh) 2021-08-24 2021-08-24 云边协同网络中时延和频谱占用联合优化方法及系统

Publications (1)

Publication Number Publication Date
WO2023024219A1 true WO2023024219A1 (zh) 2023-03-02

Family

ID=78839357

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/122981 WO2023024219A1 (zh) 2021-08-24 2021-10-11 云边协同网络中时延和频谱占用联合优化方法及系统

Country Status (3)

Country Link
US (1) US11968122B2 (zh)
CN (1) CN113784373B (zh)
WO (1) WO2023024219A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116208669A (zh) * 2023-04-28 2023-06-02 湖南大学 基于智慧灯杆的车载异构网络协同任务卸载方法及系统
CN116431349A (zh) * 2023-04-13 2023-07-14 山东华科信息技术有限公司 分布式配电网云边端数据协同方法及其系统
CN116582836A (zh) * 2023-07-13 2023-08-11 中南大学 一种任务卸载与资源分配方法、设备、介质和系统
CN116668447A (zh) * 2023-08-01 2023-08-29 贵州省广播电视信息网络股份有限公司 一种基于改进自学习权重的边缘计算任务卸载方法
CN116996198B (zh) * 2023-09-25 2023-12-19 之江实验室 一种灵活以太网双向时延对称小颗粒时隙分配方法及装置

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114363984B (zh) * 2021-12-16 2022-11-25 苏州大学 一种云边协同光载网络频谱资源分配方法及系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110891093A (zh) * 2019-12-09 2020-03-17 中国科学院计算机网络信息中心 一种时延敏感网络中边缘计算节点选择方法及系统
US20200351337A1 (en) * 2019-05-02 2020-11-05 EMC IP Holding Company LLC Resource Allocation and Provisioning in a Multi-Tier Edge-Cloud Virtualization Environment
CN111901424A (zh) * 2020-07-28 2020-11-06 苏州大学 云边协同网络资源平滑迁移与重构方法及系统

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7418494B2 (en) * 2002-07-25 2008-08-26 Intellectual Ventures Holding 40 Llc Method and system for background replication of data objects
EP3356934A1 (en) * 2015-10-02 2018-08-08 IDAC Holdings, Inc. Methods, apparatus and systems for information-centric networking (icn) based surrogate server management under dynamic conditions and varying constraints
CN106992810B (zh) * 2017-01-23 2020-02-18 苏州大学 考虑联合故障概率约束的共享保护路由和频谱分配方法
US11284307B2 (en) * 2020-04-09 2022-03-22 Tmobile Usa, Inc. Enhancing telecommunication quality of service
CN112134916B (zh) * 2020-07-21 2021-06-11 南京邮电大学 一种基于深度强化学习的云边协同计算迁移方法
CN112689303B (zh) * 2020-12-28 2022-07-22 西安电子科技大学 一种边云协同资源联合分配方法、系统及应用

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200351337A1 (en) * 2019-05-02 2020-11-05 EMC IP Holding Company LLC Resource Allocation and Provisioning in a Multi-Tier Edge-Cloud Virtualization Environment
CN110891093A (zh) * 2019-12-09 2020-03-17 中国科学院计算机网络信息中心 一种时延敏感网络中边缘计算节点选择方法及系统
CN111901424A (zh) * 2020-07-28 2020-11-06 苏州大学 云边协同网络资源平滑迁移与重构方法及系统

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
KOVACEVIC IVANA; HARJULA ERKKI; GLISIC SAVO; LORENZO BEATRIZ; YLIANTTILA MIKA: "Cloud and Edge Computation Offloading for Latency Limited Services", IEEE ACCESS, IEEE, USA, vol. 9, 8 April 2021 (2021-04-08), USA , pages 55764 - 55776, XP011849280, DOI: 10.1109/ACCESS.2021.3071848 *
LIU LING, MA WEIKE, CHEN BOWEN, GAO MINGYI, CHEN HONG, WU JINBING: "Network Resource Optimization with Latency Sensitivity in Collaborative Cloud-Edge Computing Networks", ASIA COMMUNICATIONS AND PHOTONICS CONFERENCE/INTERNATIONAL CONFERENCE ON INFORMATION PHOTONICS AND OPTICAL COMMUNICATIONS 2020 (ACP/IPOC), OPTICA PUBLISHING GROUP, WASHINGTON, D.C., 1 January 2020 (2020-01-01), Washington, D.C., pages T4C.3, XP093039634, ISBN: 978-1-943580-82-8, DOI: 10.1364/ACPC.2020.T4C.3 *
TANG, LUN; HU, YANJUAN; LIU, TONG; CHEN, QIANBIN: "Task Offloading and Resource Allocation Algorithm Based on Lyapunov in Mobile Edge Computing", COMPUTER ENGINEERING, SHANGHAI JISUANJI XUEHUI, CN, vol. 47, no. 3, 31 March 2021 (2021-03-31), CN , pages 29 - 36, XP009543864, ISSN: 1000-3428, DOI: 10.19678/j.issn.1000-3428.0058268 *
TANG, LUN; XIAO, JIAO; WEI, YANNAN; ZHAO, GUOFAN; CHEN, QIANBIN: "Joint Resource Allocation Algorithms Based on Mixed Cloud/Fog Computing in Vehicular Network", JOURNAL OF ELECTRONICS & INFORMATION TECHNOLOGY, ZHONGGUO KEXUEYUAN DIANZIXUE YANJIUSUO,CHINESE ACADEMY OF SCIENCES, INSTITUTE OF ELECTRONICS, CN, vol. 42, no. 8, 31 August 2020 (2020-08-31), CN , pages 1926 - 1933, XP009543866, ISSN: 1009-5896, DOI: 10.11999/JEIT190306 *
YANG, PENG; ZHANG, YI-FU; LI, ZHI-DU; WU, DA-PENG; WANG, RU-YAN: "Latency Guarantee Model for Reliable Edge-Node Cooperation", JOURNAL OF BEIJING UNIVERSITY OF POSTS AND TELECOMMUNICATIONS, GAI-KAN BIANJIBU, BEIJING, CN, vol. 44, no. 2, 30 April 2021 (2021-04-30), CN , pages 47 - 53, XP009543865, ISSN: 1007-5321, DOI: 10.13190/j.jbupt.2020-139 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116431349A (zh) * 2023-04-13 2023-07-14 山东华科信息技术有限公司 分布式配电网云边端数据协同方法及其系统
CN116431349B (zh) * 2023-04-13 2023-11-03 山东华科信息技术有限公司 分布式配电网云边端数据协同方法及其系统
CN116208669A (zh) * 2023-04-28 2023-06-02 湖南大学 基于智慧灯杆的车载异构网络协同任务卸载方法及系统
CN116208669B (zh) * 2023-04-28 2023-06-30 湖南大学 基于智慧灯杆的车载异构网络协同任务卸载方法及系统
CN116582836A (zh) * 2023-07-13 2023-08-11 中南大学 一种任务卸载与资源分配方法、设备、介质和系统
CN116582836B (zh) * 2023-07-13 2023-09-12 中南大学 一种任务卸载与资源分配方法、设备、介质和系统
CN116668447A (zh) * 2023-08-01 2023-08-29 贵州省广播电视信息网络股份有限公司 一种基于改进自学习权重的边缘计算任务卸载方法
CN116668447B (zh) * 2023-08-01 2023-10-20 贵州省广播电视信息网络股份有限公司 一种基于改进自学习权重的边缘计算任务卸载方法
CN116996198B (zh) * 2023-09-25 2023-12-19 之江实验室 一种灵活以太网双向时延对称小颗粒时隙分配方法及装置

Also Published As

Publication number Publication date
CN113784373A (zh) 2021-12-10
US11968122B2 (en) 2024-04-23
US20230421501A1 (en) 2023-12-28
CN113784373B (zh) 2022-11-25

Similar Documents

Publication Publication Date Title
WO2023024219A1 (zh) 云边协同网络中时延和频谱占用联合优化方法及系统
WO2022021176A1 (zh) 云边协同网络资源平滑迁移与重构方法及系统
CN108260169B (zh) 一种基于QoS保障的服务功能链动态部署方法
Sun et al. Autonomous resource slicing for virtualized vehicular networks with D2D communications based on deep reinforcement learning
CN113364850B (zh) 软件定义云边协同网络能耗优化方法和系统
WO2023039965A1 (zh) 流量疏导的云边计算网络计算资源均衡调度方法及系统
Lin et al. Three-tier capacity and traffic allocation for core, edges, and devices for mobile edge computing
CN113055487B (zh) 一种基于vmec服务网络选择的迁移方法
CN112020103A (zh) 一种移动边缘云中的内容缓存部署方法
CN111538587A (zh) 一种基于负载均衡的服务功能链重配置方法
CN113076177B (zh) 一种边缘计算环境下的虚拟机动态迁移方法
CN113115256B (zh) 一种在线vmec服务网络选择的迁移方法
Zhang et al. DMRA: A decentralized resource allocation scheme for multi-SP mobile edge computing
WO2023108718A1 (zh) 一种云边协同光载网络频谱资源分配方法及系统
CN109743217B (zh) 一种基于svra算法的自适应资源调节方法
Wang et al. Task allocation mechanism of power internet of things based on cooperative edge computing
Li et al. Deep reinforcement learning-based resource allocation and seamless handover in multi-access edge computing based on SDN
WO2023019604A1 (zh) 基于流量疏导的最小网络能耗优化方法及系统
Lin et al. Storage-assisted optical upstream transport scheme for task offloading in multi-access edge computing
WO2023222061A1 (zh) 意图驱动的无线网络资源冲突解决方法及其装置
Gatzianas et al. Energy-efficient joint computational and network resource planning in Beyond 5G networks
Liu et al. Joint task offloading and dispatching for mec with rational mobile devices and edge nodes
Guan et al. Multidimensional Resource Fragmentation-Aware Virtual Network Embedding for IoT Applications in MEC Networks
Li et al. Suitability-based edge server placement strategy in 5G ultra-dense networks
Yıldız et al. Deep Q-Learning based resource allocation and load balancing in a mobile edge system serving different types of user requests

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: 21954739

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE