Resource scheduling method and device

Abstract

Embodiments of the present application provide a resource scheduling method and device, the method comprising. A main routing path and a backup routing path of user service are acquired, the communication time delay of the main routing path is minimum, the node risk value of the backup routing path is minimum, and the node risk value represents the influence caused by node failure; a target routing path is determined according to the main routing path and the backup routing path, and the user service is scheduled to a node of the target routing path for service processing.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of optical communication computing power network, and in particular to a resource scheduling method and device. BACKGROUND

[0002] At present, in the resource configuration of all-optical computing power network, there are still some deficiencies such as imperfect real-time collection of computing power device information, flexible computing power resource scheduling department, etc. At the same time, so far, in order to alleviate network congestion, improve the transmission performance of communication network, and reduce network risk, the routing algorithm proposed by the existing research is mostly focused on determining the reachable path for the service from the network topology, and pays insufficient attention to the use of underlying resources; at the same time, it also lacks consideration of the load of underlying computing power nodes. Once the node is damaged and loses the ability, the planned service will fail, resulting in congestion. SUMMARY

[0003] Embodiments of the present application provide a resource scheduling method and device to solve the problems of unclear service demand computing power and unreasonable computing power resource allocation in existing computing power optical network.

[0004] In a first aspect, a resource scheduling method is provided, comprising:

[0005] obtaining a primary routing path and a backup routing path of a user service, the communication delay of the primary routing path being the smallest, and the node risk value of the backup routing path being the smallest, the node risk value representing the influence caused by node failure;

[0006] determining a target routing path according to the primary routing path and the backup routing path, and scheduling the user service to the node of the target routing path for service processing.

[0007] Optionally, obtaining the primary routing path and the backup routing path of the user service comprises:

[0008] obtaining the primary routing path of the user service according to an optimization target and / or a constraint condition of double routing normalization, and obtaining the backup routing path;

[0009] The optimization target includes at least one of the following: minimizing the total communication delay of the service and balancing the network risk as the optimization target.

[0010] The constraint condition of double routing normalization includes at least one of the following: routing uniqueness, path loop-free, the bandwidth resource size allocated to all services should be less than the link bandwidth threshold, the storage resources available to all user services scheduled to the computing node should not exceed the storage resources that the computing node can provide, and the computing resources available to all user services scheduled to the computing node should not exceed the computing resources that the computing node can provide. Optionally, obtaining the primary routing path of the user service comprises:

[0011] The main routing path of the user service is determined according to at least one of an access delay of the user service to an accessed node, a transmission delay of the user service between adjacent nodes, and a calculation delay of the user service.

[0012] Optionally, the method for obtaining the main routing path of the user service further comprises:

[0013] The at least one of the access delay of the user service to the accessed node, the transmission delay of the user service between the adjacent nodes, and the calculation delay of the user service is determined according to at least one of a size of the user service, a data transmission rate of the user service to the accessed node, a data transmission rate of the user service to the adjacent nodes, and a calculation capability size of the user service to the accessed node.

[0014] Optionally, the method for obtaining the backup routing path of the user service comprises:

[0015] The backup routing path of the user service is determined according to at least one of a failure probability of the user service to the accessed node, an importance degree, and a service load, wherein the importance degree represents a criticality of the node in the optical network.

[0016] Optionally, the method further comprises:

[0017] The importance degree of the node is determined according to a voltage value of the node and a degree-coupling degree index of the node, wherein the degree-coupling degree index of the node is determined based on a node betweenness and a node degree of the node, wherein the greater the node degree is, the stronger the connectivity of the node is, and the greater the node betweenness is, the faster the node failure will affect the network performance.

[0018] Optionally, the target routing path is determined according to the main routing path and the backup routing path, and the user service is scheduled to a node of the target routing path for service processing, comprising:

[0019] The target routing path is determined according to a lowest delay level priority strategy and the main routing path and the backup routing path, and the user service is scheduled to a node of the target routing path for service processing.

[0020] Optionally, the scheduling of the user service to the node of the target routing path for service processing comprises:

[0021] The computing power is scheduled according to routing information of the target routing path, a network connection is established, and corresponding idle resource information is updated.

[0022] If the distance of the scheduled node is greater than the transmission distance threshold, the service priority is lowered, the node is waited to release; the planned route is recalculated until all services are completed.

[0023] Optionally, if the distance of the scheduled node is greater than the transmission distance threshold, the service priority is lowered, the node is waited to release; the planned route is recalculated until all services are completed, including:

[0024] The number of frequency slots required for the currently selected target routing path is calculated.

[0025] It is judged whether the number of frequency slots exceeds the transmission distance threshold.

[0026] If the transmission distance threshold is exceeded, it is checked whether all paths are traversed, and in the case of confirming that all paths are traversed, it is output that the current request is blocked, the current request processing state is obtained, and the number of blocked requests of all requests is updated.

[0027] If the transmission distance threshold is not exceeded, the computing power optical network resource nodes are allocated according to the routing information of the target routing path, the current request is processed, and is not counted into the number of blocked requests.

[0028] The number of blocked requests of all requests is updated, the node is waited to release, and the service request priority is lowered.

[0029] It is checked whether the blocked service is 0, if there is a blocked service, the request information is generated again according to the newly added computing power optical network resources, and the path is reconfigured; if the blocked service is 0, the allocation ends.

[0030] In a second aspect, a resource scheduling device is provided, comprising:

[0031] An acquisition module is configured to acquire a main routing path and a backup routing path of a user service, the communication time delay of the main routing path being the smallest, and the node risk value of the backup routing path being the smallest, the node risk value representing the influence caused by node failure.

[0032] A first determination module is configured to determine a target routing path according to the main routing path and the backup routing path, and schedule the user service to the node of the target routing path for service processing.

[0033] In a third aspect, a communication device is provided, comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, the program or instruction being executed by the processor to implement the steps of the method of the first aspect.

[0034] In a fourth aspect, a readable storage medium is provided, the readable storage medium storing a program or instruction, the program or instruction being executed by a processor to implement the steps of the method of the first aspect.

[0035] In the embodiment of the present application, the main routing path and the backup routing path of the user service are obtained, the communication delay of the main routing path is the smallest, the node risk value of the backup routing path is the smallest, and the node risk value represents the influence caused by node failure; according to the main routing path and the backup routing path, the target routing path is determined, and the user service is scheduled to the node of the target routing path for service processing, so as to avoid the unclear service demand computing power, unreasonable computing power resource allocation existing in the existing computing power optical network, and realize the overall business double routing planning solution. BRIEF DESCRIPTION OF DRAWINGS

[0036] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included only to illustrate preferred embodiments and are not intended to limit the present application. Moreover, the same reference numerals are used throughout the various drawings to designate identical or similar components. In the drawings:

[0037] Figure 1 is a flowchart of a resource scheduling method provided by an embodiment of the present application;

[0038] Figure 2 is a schematic diagram of an overall architecture of a computing power optical network service provided by an embodiment of the present application;

[0039] Figure 3 is a resource scheduling flowchart of a computing power optical network provided by an embodiment of the present application;

[0040] Figure 4 is an NSGA II algorithm execution flowchart provided by the present application;

[0041] Figure 5 is a schematic diagram of a resource scheduling device provided by an embodiment of the present application;

[0042] Figure 6 is a schematic diagram of a communication device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0044] The term "comprise", including any variation thereof, e.g. "comprises" or "comprising", as used in the specification and in claims of the present application, is used in the sense of "including" and / or "warranting" and not by way of "restricted" or "limiting" and shall not exclude other moieties, components, elements, or steps.

[0045] In the embodiments of the present application, the word "exemplary" or "for example" is used to mean "an example of" or "an example, only. Any embodiment or design solution described in the embodiments of the present application as "exemplary" or "for example" should not be construed as preferred or advantageous over other embodiments or design solutions. In fact, the word "exemplary" or "for example" is used to present concepts in a concrete manner.

[0046] Referring to Figure 1 The embodiments of the present application provide a resource scheduling method, and the specific steps include: step 101, step 102.

[0047] Step 101: Obtain a main routing path and a backup routing path of user service, the communication delay of the main routing path is minimum, the node risk value of the backup routing path is minimum, and the node risk value represents the influence caused by node failure;

[0048] Step 102: According to the main routing path and the backup routing path, determine a target routing path, and schedule the user service to the node of the target routing path for service processing.

[0049] It can be understood that the target routing path can be the main routing path or the backup routing path, or other routing paths other than the main routing path and the backup routing path.

[0050] In an embodiment of the present application, obtaining the main routing path and the backup routing path of the user service includes:

[0051] According to the optimization target and / or the constraint condition of the dual routing normalization, the main routing path of the user service is obtained, and the backup routing path is obtained;

[0052] The optimization target includes at least one of the following: the minimum total communication delay of the service and the network risk balance as the optimization target.

[0053] The constraint conditions of the dual-route homogenization include at least one of the following: route uniqueness, path acyclic, bandwidth resource size of all service allocation should be less than link bandwidth threshold, storage resource available for all user services scheduled to a computing node should not exceed storage resource that the computing node can provide, computing resource available for all user services scheduled to a computing node should not exceed computing resource that the computing node can provide. In an embodiment of the present application, the main route path of the user service is obtained, including:

[0054] The main route path of the user service is determined according to at least one of the following: access delay of the user service to the accessed node, transmission delay of the user service between adjacent nodes, and computing delay of the user service.

[0055] In an embodiment of the present application, the main route path of the user service is obtained, further including:

[0056] At least one of the following is determined according to at least one of the following: size of the user service, data transmission rate of the user service to the accessed node, data transmission rate of the user service between adjacent nodes, and computing capacity size of the user service to the accessed node: access delay of the user service to the accessed node, transmission delay of the user service between the adjacent nodes, and computing delay of the user service.

[0057] In an embodiment of the present application, the backup route path of the user service is obtained, including:

[0058] The backup route path of the user service is determined according to at least one of the following: failure probability of the user service to the accessed node, importance, and service load, wherein the importance represents the criticality of the node in the optical network.

[0059] In an embodiment of the present application, the method further includes:

[0060] The importance of the node is determined according to voltage value of the node and degree-median correlation index of the node, wherein the degree-median correlation index of the node is determined based on node median and node degree of the node, wherein the greater the node degree is, the stronger the connectivity of the node is, and the greater the node median is, the faster the node failure will affect the network performance.

[0061] In an embodiment of the present application, the target route path is determined according to the main route path and the backup route path, and the user service is scheduled to the node of the target route path for service processing, including:

[0062] According to the minimum latency level priority strategy, and the main routing path and the standby routing path, a target routing path is determined, and the user service is scheduled to a node of the target routing path for service processing.

[0063] In an embodiment of the present application, the scheduling of the user service to the node of the target routing path for service processing comprises:

[0064] According to the routing information of the target routing path, computing power is scheduled, network connection is established, and corresponding idle resource information is updated;

[0065] If the distance of the scheduled node is greater than the transmission distance threshold, the service priority is reduced, the node is waited to release, the planned routing is recalculated, and all services are completed.

[0066] In an embodiment of the present application, if the distance of the scheduled node is greater than the transmission distance threshold, the service priority is reduced, the node is waited to release, the planned routing is recalculated, and all services are completed, comprising:

[0067] The number of frequency slots required by the currently selected target routing path is calculated;

[0068] It is judged whether the number of frequency slots exceeds the transmission distance threshold;

[0069] If the transmission distance threshold is exceeded, it is checked whether all paths are traversed, and in the case of confirming that all paths are traversed, it is output that the current request is blocked, the current request processing state is obtained, and the number of blocked requests of all requests is updated;

[0070] If the transmission distance threshold is not exceeded, computing power optical network resource nodes are allocated according to the routing information of the target routing path, the current request is processed, and is not counted into the number of blocked requests;

[0071] The number of blocked requests of all requests is updated, the node is waited to release, and the service request priority is reduced;

[0072] It is checked whether the blocked service is 0, if there is a blocked service, the blocked service generates request information again, and the path is reconfigured according to the new computing power optical network resource; if the blocked service is 0, the allocation ends.

[0073] In this embodiment, under the condition of clear network topology and service communication index demand, node load modeling is performed according to network structure, device level and service bearing condition and other factors; at the same time, the business performance is taken into account, the high-reliability service double-routing planning problem is modeled as a multi-objective optimization problem with the network overall node load value and the service end-to-end communication delay as the optimization target, and the service demand, bandwidth resource, double-routing independence and the like as the constraints, and finally the purpose of overall service double-routing planning solution is realized.

[0074] To support the integration of computing power optical network services, the computing power optical network architecture is logically divided into an infrastructure layer, a computing power orchestration layer, and a business application layer. Among them, the computing power orchestration layer is a key capability platform for implementing computing power optical network services, and the overall architecture of the computing power optical network service is as shown in Figure 2

[0075] The computing power optical network architecture can be logically divided into an infrastructure layer (or referred to as an optical network infrastructure layer), a computing power orchestration layer, and a business application layer. Among them, the infrastructure layer is the foundation of the computing power optical network, and functions to form a multi-level, three-dimensional, and ubiquitous distributed optical device network system; the computing power orchestration layer is the scheduling center of the computing power network, and functions to realize the flexible combination of computing power resources of the computing power optical network, and to realize the unified management, orchestration, scheduling, and optimization of the computing network resources by combining artificial intelligence (AI) and big data technology; the business application layer is a service and capability providing platform of the computing power network, and provides one-stop services and intelligent and unobtrusive experiences for customers.

[0076] 1. Infrastructure layer: network infrastructure is used to build network connections, which is the basis for all upper layers, including hardware devices such as servers, storage, network devices, and host operating systems, cloud operating systems, and virtualized network functions deployed on the basis of hardware devices. In terms of physical location, it not only includes the physical infrastructure of a cabinet and a machine room, but also covers the integration of cross-regional resources of a data center and multiple data centers, provides communication to ensure distributed deployment, and provides computing resource information distribution functions to provide overall computing capabilities for the upper layers.

[0077] Optionally, the infrastructure layer includes at least one of the following: optical resource devices, data centers, resource discovery modules, resource modeling modules, resource monitoring modules, and resource configuration optimization modules, which collectively implement the management of basic resources within the full optical computing power network, wherein,

[0078] 1) Resource discovery module, used to discover and identify multi-dimensional heterogeneous optical network computing power resources through active / passive methods;

[0079] 2) Resource modeling module, used to model and jointly model the basic optical network resources through formal expression forms;

[0080] 3) Resource monitoring module, used to track the status of optical network resources, including resource trusted service status, resource consumption, usage rate, and also to track the status of each stage in the use of basic resources in the execution process of each task;

[0081] ​4) Resource optimization configuration module, for reconfiguring the resource deployment of the computing power optical network according to the task characteristic information and the resource state information collected in the optical network resource management process, so as to improve the resource utilization and the task completion efficiency.

[0082] In the infrastructure system, the network control plane needs to perceive the storage resource and computing power resource state of the computing power node in real time. Assuming that there are M terminal devices in the computing power perception network system, the set of terminal devices is represented as M={1, 2, …, M}, and m∈M represents a terminal device. It is assumed that each terminal device initiates at most one computing task processing request at a time.

[0083] Assuming that there are K user services in the computing power optical network system, the set of user services is represented as k={1, 2, …, K}, and k∈K represents a user service. The kth user service can be represented as S k (A k ,X k ,C k ,D kmax ,D ka ,B k ), where A k represents the routing node accessed by the user service k, X k represents the quantized value of the storage resource required by the user service k, C k represents the quantized value of the computing resource required by the user service k (number of revolutions of central processing unit / bits of data per second, i.e., the number of revolutions of central processing unit required for processing unit bits of data per unit time), D kmax represents the maximum service processing delay allowed by the user service k, D ka represents the delay of the user service k accessing the routing node, B k represents the size of the user service k.

[0084] The user service requested by the terminal is transmitted to the edge wireless access point in the computing power perception network through a wireless link. According to the Shannon theorem, the data transmission rate of the user service k to the accessed routing node A k can be represented as The present application mainly studies how the service request issued by the terminal is routed and scheduled to the best computing power node for service processing based on the perceived information in the computing power perception network. Therefore, the location distribution of the user service is not considered, but the data transmission rate of the user service to the accessed routing node is uniformly quantized. In summary, the access delay of the user service k to the accessed routing node A k is:

[0085]

[0086] The user service requested by the terminal transmits data between adjacent routing nodes through dynamic communication links in the computing-aware network. Assuming that the data transmission rate of the dynamic link between routing nodes in the computing-aware network can be perceived, it is represented as:

[0087]

[0088] where w i，j represents the data transmission rate between routing node i and routing node j, satisfying

[0089]

[0090] Therefore, the transmission delay of user service k between routing node i and routing node j is:

[0091]

[0092] In the computing-aware network system, the network control plane needs to perceive the storage resources and computing resources of the computing nodes in real time. Assuming that there are N computing nodes in the network, the set of computing nodes is represented as N = {1, 2,..., N}, and n ∈ N represents a computing node, then the nth computing node can be represented as Z n (A n , X n , C n , R n ), where A n represents the routing node label connected to the computing node, X n represents the storage resource size of the computing node, C n represents the computing capacity size of the computing node (number of central processor revolutions / bit data / second, i.e., the number of central processor revolutions required to process a unit of bit data per unit time), and R n represents the data transmission rate of the computing node to the routing node.

[0093] As mentioned above, when the storage resources and computing resources of the computing nodes meet the user service demand, the computing delay of the user service is:

[0094]

[0095] Under the constraint of computing resources and storage resources, the scheduling transmission delay of user business is minimized as the optimization objective, and the routing control, selection of computing nodes and allocation of storage and computing power optical resources are jointly optimized. The total delay of a task processing includes at least one of task access delay, task transmission delay, task arrival computing node delay, task processing delay and task waiting delay, since the task processing delay and the task waiting delay have uniformity under the condition of determined computing power demand, therefore, they can be not considered in the optimization model, so the transmission delay of user business k to the scheduled computing node is:

[0096] T k,t ka k,i,j k,n (6)

[0097] The node risk value refers to the influence caused by the failure of the optical network node. The present application mainly considers the failure probability, importance and bearing business of the node. Since some equipment in the existing optical network system is old and the failure rate is rising, some equipment has stopped production and after-sales service has been terminated, so the reliability decreases and the risk increases. In addition, factors such as manufacturing defects and human damage also increase the probability of node failure. However, the calculation of node failure probability is not complex, which can be obtained by collecting historical data in the network management system and using statistical methods and formula i v represents the failure probability, represents the number of failures of any node v i in a time period T, which is statistically counted in years, months, weeks, etc. in practice.

[0098] At the same time, the influence of optical communication network failure will further spread and propagate in the network through adjacent nodes or links, continuously cracking the network performance. Intuitively, the nodes with larger degree or the nodes that are more critical in the network play a greater role in failure propagation, and the node importance is higher. The node betweenness is a global characteristic quantity in complex network to describe the node importance and influence, which can be represented by the ratio of the number of shortest paths passing through the node to the total number of shortest paths between any node pair in the network. Obviously, the larger the betweenness of the node, the more critical the position of the node in the network, and the higher the node importance. Further, the following two conclusions can be obtained: 1) the larger the node degree, the stronger the node connectivity, and the wider the range of failure propagation after the failure occurs; 2) the larger the node betweenness, the faster the speed of failure affecting network performance, therefore, in order to comprehensively measure the influence degree of node degree and node betweenness on failure propagation from the network topology angle, the node degree-betweenness correlation degree index (Degree-Betweenness Value, DBV) is introduced, which is defined as follows:

[0099] ​​​​

[0100] Here is the DBV of node v i ; is the node degree of node v i ; τ is an adjustment coefficient, mainly to eliminate the situation that the node betweenness and degree difference is too large, which can be adjusted according to actual situation. is the node betweenness, and the calculation formula is:

[0101]

[0102] wherein, represents the number of shortest paths between any node pair v s and v d ; represents the number of shortest paths passing through node v i .

[0103] Different from general communication networks, different levels of computing devices in the computing power optical network have different roles in the network. The node importance represents the criticality of the node in the optical network. The node importance is related to the device level and the device scale. Generally speaking, the higher the device level is, the wider the device jurisdiction area is, and the higher the control level is. In particular, the dispatching center communication device has the highest importance in the network. Assuming that there are μ different device level values in the network, the node level value set is For convenience of calculation, the voltage values of different sites are standardized:

[0104]

[0105] wherein, is the normalized site voltage value; ξ i is the level value of node v i , ξ min , and ξ max are the minimum value and the maximum value in the level set respectively. The final node importance calculation formula is as follows:

[0106]

[0107] Here, is the importance of node v i ; α1 and β1 are the weight coefficients of the site voltage level and the DBV index respectively, and α1 + β1 = 1. In calculation, α1 and β1 can be valued according to actual demand.

[0108] Furthermore, the service load on the node device also affects the node risk value; therefore, the service load can be represented by the number of services carried by the node. In other words, the more services a node device carries, the greater the node load, and the greater the loss or impact after a node failure. Therefore, the formula for calculating the node risk value is as follows:

[0109]

[0110] Where node v i For business s k Nodes on the routing path; For node v i The service set carried on P i v For node v i Failure probability, Represents node v i The importance, and

[0111] The following constraints should also be satisfied during the dual-route planning process:

[0112] Routing uniqueness: A unique routing path must be assigned to each service within the network. Define y. rk Let y be a binary variable, r be the number of candidate paths for the business office, and if p(k) is determined to be the routing path for the business office, then y rk =1, otherwise y rk =0, and the relevant constraints are described as follows:

[0113]

[0114] Acyclic path: Each service route should satisfy the acyclic constraint, that is:

[0115]

[0116] These are decision variables.

[0117]

[0118] Link bandwidth: In the pre-configured dual-route planning, it is assumed that services use dedicated protection on the backup route. To ensure rapid service recovery after a failure and prevent network congestion, the bandwidth resources allocated to all services should be less than the link bandwidth threshold, which is expressed as follows:

[0119]

[0120] in, For business s k|S| is the total number of services; γ is the bandwidth resource reservation ratio under different protection modes; ζ is the link bandwidth.

[0121] The application takes the service of route planning as the research object, takes the minimization of total communication delay of the service and the balance of network risk as the optimization target, and takes the communication delay of a single service, link bandwidth and double road intersection degree as the constraint to model the route, so as to ensure the real-time performance of the route planning scheme and the balanced distribution of the network risk. Based on the vector-based multi-objective joint optimization problem definition method, the objective function is expressed as:

[0122]

[0123] is a vector composed of two objective functions, and the optimal solution is a one-dimensional column vector composed of k elements, that is,

[0124] The constraint condition of the objective function is as follows:

[0125]

[0126]

[0127]

[0128]

[0129]

[0130] In the formula, K n The constraint condition (20) indicates that the storage resources available to all user services scheduled to the computing node do not exceed the storage resources that can be provided by the computing node, and the constraint condition (21) indicates that the computing resources available to all user services scheduled to the computing node do not exceed the computing resources that can be provided by the computing node.

[0131] 2. The computing power arrangement layer: centralized and unified management and control of computing power resources and network resources, and guaranteeing delay determinacy and path optimization; the computing power management mainly centrally and uniformly manages distributed computing and storage resources, provides elastic resources for function calculation processing by real-time monitoring of resource load conditions, has flexible expansion and contraction capability; interfaces with heterogeneous infrastructure resources, shields the influence of heterogeneity on application deployment, realizes abstraction and management of multi-level ubiquitous computing power; and interfaces with distributed infrastructure resources to realize hierarchical management and control.

[0132] The computing power arrangement layer includes at least one of a computing power deconstruction module, a computing power management module, a resource storage module, a computing power scheduling module, a computing power monitoring module and a computing power integration module.

[0133] 1) Algorithmic power deconstruction module, which decomposes diversified, large-granularity, and complex algorithmic power requirements into small-granularity and simplified algorithmic power requirements according to business logic, resource requirements, performance requirements, service continuity, business flow stickiness, and resource node uniformity;

[0134] 2) Algorithmic power management module, which is used for unified management and control of algorithmic power resources and network set resources, and guarantees latency determinacy and path optimization;

[0135] 3) Resource storage module, which stores data generated by business computing of algorithmic power nodes;

[0136] 4) Algorithmic power scheduling module, which is responsible for publishing current computing capability status and network status as routing information to the network by comprehensively considering network status and computing resource status;

[0137] 5) Algorithmic power monitoring module, which is used for real-time monitoring of load status of algorithmic power resources and network resources;

[0138] 6) Algorithmic power integration module, which is used for centralized and unified management of distributed computing and storage resources;

[0139] Optionally, algorithmic power optical resource scheduling process:

[0140] A user first initiates an application on an algorithmic power transaction platform, including requirements for the size of algorithmic power resources and requirements for latency. After receiving the user transaction application, the algorithmic power transaction platform queries relevant algorithmic power resource information and network resource information from an algorithmic power network management and arrangement system. If necessary, the algorithmic power optical network management and arrangement platform can also initiate an end-to-end latency measurement process to obtain more accurate latency information.

[0141] Step 1: An algorithmic power optical network consumer proposes a business application, such as a site location, an algorithmic power optical resource demand size, and connection service requirements.

[0142] Step 2: The algorithmic power optical network transaction platform generates an algorithmic power optical network resource view according to the business application of the algorithmic power optical network consumer, integrates possible algorithmic power resource pools and related network connection resources in a view with the algorithmic power optical network consumer as the center.

[0143] Step 3: The algorithmic power optical network transaction platform plans a route with minimum communication latency as a business primary route and a route with minimum risk as a business backup route according to a risk perception and double-route allocation algorithm of a non-dominated sorting genetic algorithm II (NSGAII).

[0144] Fourth step: The computing power transaction platform updates the corresponding idle resource information by dispatching computing power resources, establishing network connections, etc. through the computing power network management and scheduling system according to the transaction contract.

[0145] Fifth step: The computing power transaction platform continuously tracks the resource occupation until the transaction end time specified by the transaction contract, and then terminates the service and releases the computing power resources and network resources.

[0146] 3. Business application layer: This part is mainly new business and application deployed in the computing power optical network. These businesses and applications can access and call the function services provided by the platform through the corresponding function application programming interface (API). At the same time, according to the application itself, it can put forward demands to the computing power scheduling layer, control the underlying network and resource scheduling through the computing power scheduling layer, and has the ability of elastic expansion.

[0147] The business application layer is the layer that users of the computing power optical network directly contact. Its main function is to provide users with an interface to access the computing power optical network, collect user business information and match it to improve user satisfaction.

[0148] Optionally, the business application layer includes at least one of a business expression module, a service opening module, a business collection module, a computing power request module, a computing power display module, and a computing power transaction module.

[0149] 1) The business expression module is used to formulate and analyze the business expression template.

[0150] 2) The service opening module is used to display the services that the computing power optical network can provide and the current available service resources, and match them with the user's business to determine whether the corresponding service can be provided.

[0151] 3) The user business collection module is used to record and save the relevant information of the user's business request for the computing power optical network, and conduct user satisfaction survey.

[0152] 4) The computing power request module is used to serve users. Users can select specific computing power nodes according to performance and time needs.

[0153] 5) The computing power display module is used to display the idle computing power resources in the computing power network and the performance characteristics of each computing power resource.

[0154] 6) The computing power transaction module is used for users to complete computing power demand transactions.

[0155] In the embodiments of the present application, firstly, in the case of clear network topology and service communication index demand, the access delay, transmission delay and calculation delay of the computing power service in the computing power optical network are considered according to the communication delay modeling. In the case of ensuring the user experience and the minimum communication delay, a main route is planned for subsequent computing power scheduling.

[0156] Secondly, the bearing capacity of the core node and the load balancing of the overall network are considered because the core node often bears most of the business tasks and once the core node has a problem, it will greatly affect the network operation. Therefore, according to the device scale level and service bearing condition of the optical network node and other factors, the node load modeling is performed to plan a standby route with the minimum load and the highest risk resistance.

[0157] Finally, a multi-objective optimization problem is solved by taking the service demand, bandwidth resource, double-route independence and other constraints as the constraints, and then the routing information obtained by solving is uploaded to the computing power arrangement layer. The control node of the computing power arrangement layer schedules the computing power according to the routing information, establishes the network connection and updates the corresponding idle resource information. If the scheduled computing power node is too far away and exceeds the threshold, the business priority will be reduced and the computing power node will be released. When there are enough idle computing power nodes, the routing is recalculated and planned until all the services are completed.

[0158] In order to better understand the present application, the following will be combined Figure 3 with the embodiments of the present application to make a further detailed description. The embodiments of the present application do not have special restrictions on the optical network topology, node number and terminal device number used by the system; and the access mechanism and access time are also not limited.

[0159] The specific implementation process is as follows:

[0160] Step 1: Initialize the computing power optical network device parameter information.

[0161] Step 2: The user initiates a transaction request and randomly obtains computing power resources.

[0162] Step 3: Generate all transmission paths to form a real-time changing multi-source weighted graph.

[0163] Step 4: Select the shortest delay and the corresponding specific scheduling path.

[0164] Step 5: According to the highest scheduling level strategy, calculate the number of frequency slots required for the currently selected path.

[0165] Step 6: Determine whether the transmission distance threshold is exceeded. If yes, execute step 7, otherwise execute step 12.

[0166] Step 7: Determine whether all paths are traversed. If yes, execute step 8, otherwise return to step 5.

[0167] Step 8: The current request is blocked.

[0168] Step 9: Get the current request processing status, update the number of blocked requests of all requests.

[0169] Step 10: Wait for the release of the computing node, and reduce the priority of the service request.

[0170] Step 11: Determine whether the number of blocked requests is released to 0, if yes, end the process, otherwise return to step 2;

[0171] Step 12: Allocate computing power optical network resources.

[0172] Step 13: The current request is processed, and then step 9 is executed.

[0173] In order to better understand the present application, the present application will be further described in detail below in combination with embodiments. The specific implementation process is as follows: Step 1: The state Z of the computing node in the computing power aware optical network, the number N of computing nodes, the state W of the link in the computing power aware optical network, the computing power demand S of the user service, and the number K of user services.

[0174] Step 2: Initialize the information of the computing power optical network equipment, integrate the cross-regional resources of the data center and multiple data centers, calculate the number N of computing nodes, and the state W of the link in the computing power aware network.

[0175] Step 3: The user initiates an application on the computing power transaction platform, including the demand for the size of the computing power resource and the requirement for the delay, etc.

[0176] Step 4: In order to ensure the fairness of the user, the computing task randomly acquires the computing power resource of the network. According to the accumulation and analysis of the task characteristic data and the resource state information, etc., the layout scheme of the computing power network is reconfigured, including the redistribution of the computing power node density and the redistribution of the network bandwidth.

[0177] Step 5: According to the information calculated by the network awareness, the transmission delay of the user service through any link in the network is calculated, and all terminal devices, routing nodes, computing nodes and dynamic link delays constitute a real-time changing multi-source weighted graph.

[0178] Step 6: Initialize the population, set the iteration number g = 1 and calculate the network risk balance value and the total communication delay value in the initial service distribution.

[0179] Step 7: Perform selection, crossover and mutation operations on P1 and generate a new population N1. Merge the populations P g , N g to get R g . Calculate the new population P gThe corresponding objective function value, perform fast non-dominated sorting and calculate the crowded distance.

[0180] Step 8: Select the best m chromosomes to produce population P g+1 . P g+1 Perform selection, crossover, mutation operations and produce population N g+1 .

[0181] Step 9: The number of iterations reaches the set value, output the path with the minimum communication delay as the service main route planning scheme, and select the route scheme with the minimum risk balance value as the service backup route planning.

[0182] Step 10: According to the lowest delay level priority strategy, calculate the number of frequency slots required by the currently selected path.

[0183] Step 11: Count the number of frequency slots at this time, whether it exceeds the transmission distance threshold.

[0184] Step 12: If it exceeds the transmission distance threshold, check whether all paths are traversed, and output the current request blocked under the condition that all paths are confirmed to be traversed. Finally, get the current request processing state, and update the number of blocked requests of all requests.

[0185] Step 13: If it does not exceed the transmission distance threshold, allocate computing power optical network resource nodes according to the path information, and the current request is processed, not counted into the blocked request number.

[0186] Step 14: Update the number of blocked requests of all requests, wait for the computing power node to be released, and reduce the priority of the service request.

[0187] Step 15: Check if the blocked service is 0, if there is a blocked service, generate request information again, and reconfigure the path according to the new computing power optical network resources. If the blocked service is 0, the allocation is ended.

[0188] In this application, a risk perception and double route allocation algorithm based on NSGA II is adopted.

[0189] Fast NSGA II with elitist strategy. The elitist strategy means that the parent individuals with good performance in the population iteration process will be reserved. This strategy can improve the performance of the population during the execution of the algorithm; and the crowded distance quantifies the distance between the individual in the same front and the other two adjacent individuals. By calculating the crowded distance, the population diversity is maintained and the algorithm convergence is accelerated. However, due to the competitive relationship between multiple objective functions, the decision maker needs to make a trade-off among multiple near-optimal solutions according to specific requirements when determining the final scheme, Figure 4 The execution process of NSGA II algorithm is given, and the specific steps are as follows:

[0190] Step 1: randomly generate initial population P1.

[0191] Step 2: set the number of evolutionary iterations g = 1.

[0192] Step 3: perform selection, crossover and mutation operations on P1 and generate new population N1;

[0193] Step 4: merge populations P g , N g to obtain R g .

[0194] Step 5: calculate the objective function value corresponding to the new population P g ;

[0195] Step 6: perform fast non-dominated sorting.

[0196] Step 7: calculate the crowding distance.

[0197] Step 8: select the best m chromosomes to generate population P g+1 .

[0198] Step 9: perform selection, crossover and mutation operations on P g+1 and generate population N g+1 .

[0199] Step 10: determine whether g is less than the threshold value g max , if yes, return to step 4; otherwise, output the result.

[0200] In NSGA II, the encoding and decoding method of chromosomes will directly affect the performance of the algorithm. To improve the encoding and decoding efficiency and realize unified routing planning for all services, the application adopts a fixed-length integer encoding method. Each chromosome represents a routing planning scheme, and each chromosome has multiple independent chromosome segments. Each chromosome segment corresponds to a service path, and the length of the chromosome segment depends on the number of network nodes. Therefore, the network nodes need to be numbered first, and then the gene bits are randomly generated according to the number of network nodes. From the above encoding process, it can be seen that the length of each chromosome is the product of the number of network nodes and the number of services.

[0201] In the decoding process, a decoding method based on position priority is adopted. The source point of the service is specified as the first node on the path, and then according to the relationship between nodes in the network topology adjacency matrix, the node with high priority, i.e. large position index, is selected as the next hop node. Repeat the process until the destination node of the service is reached.

[0202] The selection operator is a process of selecting excellent individuals from the old population to form a new population with a certain probability. The selection operation is implemented by using the binary tournament method and the elite strategy. In the mutation operator, the diversity of the population is achieved by selecting a mutated chromosome according to a set mutation probability and exchanging the positions of two genes in the same chromosome segment.

[0203] In the embodiment, first, under the condition of clear network topology and service communication index demand, node risk modeling is performed according to network structure, site level, service bearing condition and other factors, and the overall risk balance value of the network is calculated. At the same time, the service performance is considered, the high-reliability service dual-routing planning problem is modeled as a multi-objective optimization problem with the overall risk balance value of the network and the service end-to-end communication delay as the optimization objectives, and the service demand, bandwidth resource, dual-routing independence and the like as the constraints, and an intelligent optimization algorithm is used for solving, so as to finally realize the purpose of overall service dual-routing planning solution. Finally, taking the algorithm as the center and the light as the foundation, the network, the cloud, the data, the intelligence, the security, the edge, the end and the chain are deeply integrated, and a new type of information infrastructure of integrated service is realized. The computing power optical network aims to connect the ubiquitous computing power through the optical transmission network, break through the performance limit of single-point computing power, play the scale efficiency advantage of full-optical computing power, and serve all walks of life in the society.

[0204] Referring to Figure 5 The embodiment of the present application provides a resource processing device, which comprises a device 500.

[0205] The acquisition module 501 is configured to acquire a main routing path and a backup routing path of a user service, the communication delay of the main routing path being the smallest, and the node risk value of the backup routing path being the smallest, the node risk value representing the influence caused by node failure.

[0206] The first determination module 502 is configured to determine a target routing path according to the main routing path and the backup routing path, and schedule the user service to a node of the target routing path for service processing.

[0207] In an embodiment of the present application, the acquisition module 501 is further configured to:

[0208] According to the optimization objective and / or the constraint condition of dual-routing normalization, the main routing path and the backup routing path of the user service are acquired.

[0209] The optimization objective includes at least one of the following: the optimization objective is to minimize the total communication delay of the service and the network risk balance.

[0210] The constraint condition of the dual-route homogenization includes at least one of the following: route uniqueness, path acyclic, bandwidth resource size of all service allocation should be less than a link bandwidth threshold, storage resource available for all user services scheduled to a computing node should not exceed storage resource available for the computing node, computing resource available for all user services scheduled to a computing node should not exceed computing resource available for the computing node. In an embodiment of the present application, the main route path of the user service is obtained, including:

[0211] The main route path of the user service is determined according to at least one of the following: access delay of the user service to an accessed node, transmission delay of the user service between adjacent nodes, and computing delay of the user service.

[0212] In an embodiment of the present application, the main route path of the user service is obtained, further including:

[0213] The access delay of the user service to an accessed node, the transmission delay of the user service between adjacent nodes, and the computing delay of the user service are determined according to at least one of the following: size of the user service, data transmission rate of the user service to the accessed node, data transmission rate of the user service between the adjacent nodes, and computing capacity size of the user service to the accessed node.

[0214] In an embodiment of the present application, the backup route path of the user service is obtained, including:

[0215] The backup route path of the user service is determined according to at least one of the following: failure probability of the user service to an accessed node, importance, and service load, wherein the importance represents a key degree of a node in an optical network.

[0216] In an embodiment of the present application, the apparatus further includes:

[0217] The second determining module is configured to determine the importance of a node according to a voltage value of the node and a degree-cosine correlation index of the node, wherein the degree-cosine correlation index of the node is determined based on a node betweenness and a node degree of the node, wherein the greater the node degree is, the stronger the connectivity of the node is, and the greater the node betweenness is, the faster the node failure will affect the network performance.

[0218] In an embodiment of the present application, the first determining module is further configured to:

[0219] According to the minimum latency level priority strategy, and the main routing path and the standby routing path, a target routing path is determined, and the user service is scheduled to a node of the target routing path for service processing.

[0220] In an embodiment of the present application, the scheduling of the user service to the node of the target routing path for service processing comprises:

[0221] According to the routing information of the target routing path, scheduling computing power, establishing a network connection and updating the corresponding idle resource information;

[0222] If the distance of the scheduled node is greater than the transmission distance threshold, the service priority is reduced, the node is waited to release, the planned routing is recalculated until all services are completed.

[0223] In an embodiment of the present application, if the distance of the scheduled node is greater than the transmission distance threshold, the service priority is reduced, the node is waited to release, the planned routing is recalculated until all services are completed, comprising:

[0224] The number of frequency slots required by the currently selected target routing path is calculated;

[0225] It is judged whether the number of frequency slots exceeds the transmission distance threshold;

[0226] If the transmission distance threshold is exceeded, it is checked whether all paths are traversed, and in the case of confirming that all paths are traversed, it is output that the current request is blocked, the current request processing state is obtained, and the number of blocked requests of all requests is updated;

[0227] If the transmission distance threshold is not exceeded, the computing power optical network resource node is allocated according to the routing information of the target routing path, the current request is processed, and is not counted into the number of blocked requests;

[0228] The number of blocked requests of all requests is updated, the node is waited to release, and the service request priority is reduced;

[0229] It is checked whether the blocked service is 0, if there is a blocked service, the blocked service generates request information again, and the path is reconfigured according to the newly added computing power optical network resource; if the blocked service is 0, the allocation is ended.

[0230] The device provided by the embodiment of the present application can realize Figure 1 The method embodiment shown realizes each process and achieves the same technical effect, and to avoid repetition, details are not repeated here.

[0231] As Figure 6As shown, the embodiments of the present application also provide a communication device 600, which comprises a processor 601, a memory 602, and a program or instruction stored in the memory 602 and executable on the processor 601. The program or instruction is executed by the processor 601 to implement the above-mentioned Figure 1 The various processes of the method embodiments are implemented, and the same technical effects can be achieved. To avoid repetition, details are not described here.

[0232] The embodiments of the present application also provide a readable storage medium, which stores a program or instruction. The program or instruction is executed by a processor to implement the above-mentioned Figure 1 The various processes of the method embodiments are implemented, and the same technical effects can be achieved. To avoid repetition, details are not described here.

[0233] The processor is the processor in the terminal described in the above embodiments. The readable storage medium includes a computer readable storage medium, such as a computer readable memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, etc.

[0234] The steps of the method or algorithm described in connection with the present application can be implemented in hardware, or by executing software instructions in a processor. The software instructions can be composed of corresponding software modules, which can be stored in a RAM, a flash memory, a ROM, an EPROM, an EEPROM, a register, a hard disk, a mobile hard disk, a read-only optical disk, or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor, so that the processor can read information from the storage medium, and can write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and the storage medium can be carried in an ASIC. In addition, the ASIC can be carried in a core network interface device. Of course, the processor and the storage medium can also exist as discrete components in the core network interface device.

[0235] Those skilled in the art should realize that in one or more examples described above, the functions described in the present application can be implemented in hardware, software, firmware or any combination thereof. When implemented in software, these functions can be stored in a computer readable medium or transmitted as one or more instructions or codes on a computer readable medium. The computer readable medium includes a computer storage medium and a communication medium, wherein the communication medium includes any medium that facilitates the transfer of a computer program from one place to another. The storage medium can be any available medium accessible by a general or special purpose computer.

[0236] The specific implementation described above further details the purposes, technical solutions and beneficial effects of the present application. It should be understood that the above description is merely a specific implementation of the present application and is not intended to limit the protection scope of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the present application shall be included in the protection scope of the present application.

[0237] Those skilled in the art should understand that embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, embodiments of the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. In addition, embodiments of the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0238] Embodiments of the present application are described with reference to flowcharts and / or block diagrams according to methods, devices (systems), and computer program products of embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0239] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction apparatus that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0240] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks ​ an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0241] It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover modifications and variations of this application provided they come within the scope of the appended claims and their equivalents.

Claims

1. A resource scheduling method, characterized in that, The method comprises the following steps: obtaining a primary routing path and a backup routing path of user service, the communication delay of the primary routing path being minimum, and the node risk value of the backup routing path being minimum, the node risk value representing the influence caused by node failure; determining a target routing path according to the primary routing path and the backup routing path, and scheduling the user service to a node of the target routing path for service processing; obtaining a primary routing path and a backup routing path of user service, comprising: obtaining the primary routing path and the backup routing path of the user service according to an optimization target and a constraint condition of double routing normalization; wherein the optimization target comprises at least one of the following: minimizing the total communication delay of the service and balancing the network risk as the optimization target; the constraint condition of double routing normalization comprises at least one of the following: routing uniqueness, path loop-free, the bandwidth resource size allocated to all services being less than a link bandwidth threshold, the storage resource available to all user services scheduled to a computing node being less than the storage resource that can be provided by the computing node, and the computing resource available to all user services scheduled to a computing node being less than the computing resource that can be provided by the computing node; determining a target routing path according to the primary routing path and the backup routing path, and scheduling the user service to a node of the target routing path for service processing, comprising: determining a target routing path according to a lowest delay level priority strategy and the primary routing path and the backup routing path; scheduling computing power according to the routing information of the target routing path, establishing a network connection and updating the corresponding idle resource information; calculating the number of frequency slots required by the currently selected target routing path; judging whether the number of frequency slots exceeds a transmission distance threshold; if the number of frequency slots exceeds the transmission distance threshold, checking whether all paths are traversed, and if it is confirmed that all paths are traversed, outputting that the current request is blocked, obtaining the current request processing state, and updating the number of blocked requests of all requests; if the number of frequency slots does not exceed the transmission distance threshold, allocating computing power optical network resource nodes according to the routing information of the target routing path, the current request being processed, and not being counted into the number of blocked requests; updating the number of blocked requests of all requests, waiting for the node to be released, and reducing the priority of the service request; checking whether the blocked service is 0, if there is a blocked service, generating request information again, and reconfiguring the path according to the newly added computing power optical network resource; if the blocked service is 0, the allocation ends.

2. The method of claim 1, wherein, obtaining a primary routing path of user service, comprising: determining the primary routing path of the user service according to at least one of the access delay of the user service to the accessed node, the transmission delay of the user service between adjacent nodes, and the computing delay of the user service.

3. The method of claim 2, wherein, obtaining a primary routing path of user service, further comprising: Determine at least one of an access latency of the user service to the accessed node, a transmission latency of the user service between the accessed neighboring nodes, and a computation latency of the user service according to at least one of a size of the user service, a data transmission rate of the user service to the accessed node, a data transmission rate of the user service between the accessed neighboring nodes, and a size of a computation capability of the user service to the accessed node.

4. The method of claim 1, wherein, The method further comprises: Determine the importance of a node according to a voltage value of the node and a degree-mediacy index of the node, wherein the degree-mediacy index of the node is determined based on a node betweenness and a node degree of the node, wherein the greater the node degree is, the stronger the connectivity of the node is, and the greater the node betweenness is, the faster the node will affect the network performance after a fault occurs.

5. A resource scheduling apparatus, characterized by comprising: Comprise: An acquisition module, configured to acquire a primary routing path and a backup routing path of user service, wherein a communication latency of the primary routing path is the smallest, and a node risk value of the backup routing path is the smallest, wherein the node risk value represents an impact caused by a node fault; A first determination module, configured to determine a target routing path according to the primary routing path and the backup routing path, and schedule the user service to a node of the target routing path for service processing; Acquire a primary routing path and a backup routing path of user service, comprising: Acquire the primary routing path and the backup routing path of the user service according to an optimization target and a constraint condition of double routing normalization; The optimization target comprises at least one of the following: minimizing the total communication latency of the service and balancing the network risk as an optimization target; The constraint condition of double routing normalization comprises at least one of the following: routing uniqueness, path acyclic, the bandwidth resource size allocated to all services should be less than a link bandwidth threshold, the storage resource available to all user services scheduled to a computing node should not exceed the storage resource that the computing node can provide, and the computing resource available to all user services scheduled to a computing node should not exceed the computing resource that the computing node can provide; Determine a target routing path according to the primary routing path and the backup routing path, and schedule the user service to a node of the target routing path for service processing, comprising: Determine a target routing path according to a lowest latency level priority strategy and the primary routing path and the backup routing path; Schedule computing power according to routing information of the target routing path, establish a network connection, and update corresponding idle resource information; Calculate the number of frequency slots required by the currently selected target routing path; Determine whether the number of frequency slots exceeds a transmission distance threshold; If the number of frequency slots exceeds the transmission distance threshold, check whether all paths are traversed, and if it is confirmed that all paths are traversed, output that the current request is blocked, obtain a current request processing state, and update the number of blocked requests of all requests; If the number of frequency slots does not exceed the transmission distance threshold, allocate computing power optical network resource nodes according to routing information of the target routing path, the current request is processed, and is not counted into the number of blocked requests; Update the number of blocked requests of all requests, wait for the node to be released, and reduce the priority of the service request. If the blocked service is 0, the request information is generated again, and the path is reconfigured according to the newly added computing power optical network resource; if the blocked service is 0, the allocation ends.

6. A communication device, characterized by The processor, the memory, and a program or instructions stored on the memory and executable on the processor to implement the steps of the method of any one of claims 1 to 4.

7. A readable storage medium, characterized by, The program or instructions are stored on the readable storage medium, and the program or instructions are executable on the processor to implement the steps of the method of any one of claims 1 to 4.

Images