CN116132403A - Route distribution method and device of computing power network, electronic equipment and storage medium - Google Patents
Route distribution method and device of computing power network, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116132403A CN116132403A CN202211534614.9A CN202211534614A CN116132403A CN 116132403 A CN116132403 A CN 116132403A CN 202211534614 A CN202211534614 A CN 202211534614A CN 116132403 A CN116132403 A CN 116132403A
- Authority
- CN
- China
- Prior art keywords
- route
- computing
- antibody
- power network
- time delay
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/14—Routing performance; Theoretical aspects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/74—Address processing for routing
-
- 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
Abstract
The invention provides a route distribution method, a device, electronic equipment and a storage medium of a computing power network, wherein the method comprises the steps of obtaining a plurality of calculation requests of a plurality of user services in the computing power network; determining a computing node in a computing power network; obtaining a route distribution result under the condition of meeting a plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm; the route distribution result comprises optimal time delay from user service to each computing node and a scheduling path corresponding to the optimal time delay; and performing route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result. By means of the method, multi-objective joint optimization computing power perception route distribution can be provided for computing power network scenes, ubiquitous and heterogeneous computing power resources are efficiently and cooperatively perceived and utilized, endophytic perception and computing power self-adaption capacity of a communication network are improved, and network energy efficiency is improved.
Description
Technical Field
The present invention relates to the field of mobile communications technologies, and in particular, to a routing allocation method and apparatus for a computing network, an electronic device, and a storage medium.
Background
With the accelerated development and maturity of novel network services and applications, cloud computing, edge computing and intelligent terminal equipment are rapidly developed, computing resources show a ubiquitous deployment trend, and how to efficiently and cooperatively utilize the ubiquitous computing resources becomes an important new subject in the current network field research. Under the background, the concept of the computational power network is proposed and draws a great deal of attention, and the basic idea is to integrate computational power and network depth, cooperate with distributed computing resources, improve the utilization rate of the computing resources, and improve the network service experience of users. The computing power network is used as a novel infrastructure for providing computing power and network depth fusion and integration services, and provides important support for the construction of network strong countries, digital China and intelligent society. The resource management method is different from the traditional cloud computing resource management or intensive resource supply, and the influence of network delay and network loss on the aspect of resource scheduling is more considered in the resource management of the computing network.
With the increasing demand of new business applications such as augmented reality, autopilot, smart city, industrial internet and the like on network computing power, the edge computing power network system faces the problem of network coexistence, namely unbalanced load. This results in that a part of the edge servers cannot meet the processing requirements of the business application, and the computing power resources of another part of the edge servers are in an idle state.
Disclosure of Invention
The invention provides a routing distribution method, a routing distribution device, electronic equipment and a storage medium of a computing power network, which are used for solving the defect of uneven load of the computing power network in the prior art.
The invention provides a route distribution method of a computing power network, which comprises the following steps: acquiring a plurality of calculation requests of a plurality of user services in a computing power network; determining a computing node in a computing power network; obtaining a route distribution result under the condition of meeting a plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm; the route distribution result comprises optimal time delay from user service to each computing node and a scheduling path corresponding to the optimal time delay; and performing route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result.
According to the route distribution method of the power calculation network provided by the invention, the user service comprises service priority; the optimized routing model comprises a service priority model, and the optimized routing model is utilized to obtain a routing distribution result under the condition of meeting a plurality of calculation requests through an immune genetic algorithm, and the method comprises the following steps: the optimized routing model calculates time delay from the user service to each calculation node by utilizing an immune genetic algorithm and a scheduling path corresponding to the time delay; based on the service priority, the service priority model determines the optimal time delay from the user service to each computing node and the scheduling path corresponding to the optimal time delay from the user service to each computing node in the scheduling paths corresponding to the time delay.
According to the route distribution method of the computational power network, provided by the invention, the route distribution result under the condition of meeting a plurality of calculation requests is obtained by utilizing an optimized route model through an immune genetic algorithm, and the route distribution method comprises the following steps: setting a computing node distribution vector of user service as an antigen of an immune genetic algorithm, and setting a routing node path matrix of a path from the user service to the computing node as an antibody of the immune genetic algorithm; determining initial parameters and generating an initial population of antibodies for the immune genetic algorithm based on the initial parameters; evaluating each antibody in the initial population of antibodies to obtain an affinity for each antibody, wherein the affinity represents a degree of matching between the antibody and the antigen; calculating the concentration of the antibody according to the affinity degree to promote and inhibit the antibody so as to obtain the polyfitness; optimizing the antibody population according to the polymerization adaptability to obtain an optimal solution of an immune genetic algorithm; and outputting the optimal solution of the immune genetic algorithm as a route allocation result.
According to the route distribution method of the computational power network, which is provided by the invention, the antibody population is optimized according to the aggregation adaptability to obtain the optimal solution of the immune genetic algorithm, and the route distribution method comprises the following steps: directly inheriting individuals with the fitness lower than a preset value to the next generation or generating new individuals through pairing and crossing and inheriting the new individuals to the next generation; storing an individual with the aggregation adaptability higher than a preset value as a good solution into a memory unit; calculating the fit degree and the corresponding good solution of the next generation until the preset termination condition is met, and outputting the optimal solution; wherein the optimal solution is the optimal solution of the good solution in the memory cell.
According to the route allocation method of the computational power network provided by the invention, before generating the initial antibody population of the immune genetic algorithm based on the initial parameters, the route allocation method further comprises the following steps: comparing the similarity of the antigen and the historical antigen to obtain a comparison result; judging whether the first response occurs or not based on the comparison result; if the answer is the primary answer, randomly generating an initial antibody group by a chaos optimization algorithm; if the primary responses are not, randomly generating a first initial antibody subgroup by a chaos optimization algorithm, taking a memory unit antibody of part of historical antigens from memory data as a second initial antibody subgroup, and forming an initial antibody subgroup by the first initial antibody subgroup and the second initial antibody subgroup.
According to the route distribution method of the power calculation network provided by the invention, the individuals with the fitness lower than the preset value are directly inherited to the next generation or new individuals are generated through pairing and crossing and inherited to the next generation, and the route distribution method comprises the following steps: selecting two antibodies from individuals with aggregation fitness lower than a preset value to perform cross probability calculation to obtain a cross probability result; determining whether to perform cross pairing according to the cross probability result; if cross pairing is carried out, carrying out variation judgment on the result generated by each cross pairing to obtain a variation judgment result; and carrying out variation adjustment on the result generated by the cross pairing based on the variation judgment result.
The invention also provides a route distribution device of the computing power network, which comprises: the user service module is used for acquiring a plurality of calculation requests of a plurality of user services in the computing power network; a computing node module for determining computing nodes in the computing power network; the optimized route model module is used for obtaining route distribution results under the condition of meeting a plurality of calculation requests through an immune genetic algorithm by utilizing the optimized route model; the route distribution result comprises optimal time delay from user service to each computing node and a scheduling path corresponding to the optimal time delay; and the distribution module is used for carrying out route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and operable on the processor, the processor implementing a route allocation method for any one of the aforementioned computing networks when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of route allocation for a power network as any one of the above.
The invention also provides a computer program product comprising a computer program which when executed by a processor implements a method of route allocation for a power network as any one of the above.
The route distribution method, the route distribution device, the electronic equipment and the storage medium of the computing power network, provided by the invention, determine the computing nodes in the computing power network by acquiring a plurality of computing requests of a plurality of user services; obtaining route distribution results under the condition of meeting a plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm; and performing route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result. The route distribution result comprises optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay. By means of the method, multi-objective joint optimization computing power perception route distribution can be provided for computing power network scenes, ubiquitous and heterogeneous computing power resources are efficiently and cooperatively perceived and utilized, endophytic perception and computing power self-adaption capacity of a communication network are improved, and network energy efficiency is improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of an embodiment of a routing distribution method for a power network according to the present invention;
FIG. 2 is a schematic diagram of an embodiment of a route distribution device of the power network of the present invention;
fig. 3 is a schematic structural diagram of an embodiment of the electronic device of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention provides a route distribution method of a computing power network. The invention comprises an immunity genetic-based computational power perception route allocation strategy which effectively carries out collaborative management on computational power and a network and relieves the problem of uneven network load.
Referring to fig. 1, fig. 1 is a flow chart of an embodiment of a routing distribution method of a power network according to the present invention, in this embodiment, the routing distribution method of the power network may include steps S110 to S140, where each step is specifically as follows:
s110: a plurality of computing requests for a plurality of user services in a computing power network are obtained.
Assuming that there are I user traffic in the power network, the set of user traffic is denoted i= {1,2, …, I }, I e I denotes one user traffic. The ith user traffic may be expressed as:
wherein N is i A routing node accessed to the user service i; s is S i The value of the storage resource required for the user service i; c (C) i Calculating resource values required for user service i;
the maximum service processing time delay allowed for the user service i; />
For user traffic i to routing node N i Is connected with the time delay of the access; v (V) i Is the size of the user service i.
User service i is accessed to route node N i Signal to noise ratio of transmission signal to noise ratio of (2)
Can be expressed as formula (1):
wherein eta 0 Is white gaussian noise, which is a white gaussian noise,
is the transmit power, +.>
Is the path loss. />
Is related to the distance between the user and the node.
Data transmission rate
According to the formula (2):
wherein the method comprises the steps of
Accessing to a routing node N for user traffic i i Is used for the channel bandwidth of the mobile station. />
User service i is accessed to route node N i Is to be connected to the access delay of (a)
Is of formula (3):
after user service is accessed through route nodes, data transmission is carried out between route nodes. In a power aware network, the data transmission rate of a dynamic link may be perceived. By w jk Representing the data transfer rate between the different routing nodes j and k. If there is no link w between route node j and route node k jk =0. Thus the transmission delay of traffic i between routing node j and routing node k
Is of formula (4):
s120: computing nodes in the computing power network are determined.
In a power network, there are L computing nodes, and the set of computing nodes is denoted as l= {1,2, …, L }, L e L denotes one computing node. The first compute node may be represented as:
Z l (N l ,S l ,C l ,U l )
wherein N is l A routing node j linked to the computing node l; s is S l Storing a resource amount for the compute node l; c (C) l Calculating a resource amount for a calculation node l; u (U) l For calculating the transmission rate of node l to a routing node。
User traffic i reaches computing node l via a fixed link, and the data transfer rate from routing node j to computing node l is u jl And (3) representing. Therefore, after selecting the computing node l, the user service i reaches the computing node l through the routing node j for reaching the time delay
Is of formula (5):
the total delay of the service i reaching the computing node l comprises: access delay, transmission delay, arrival delay, service processing delay, and service waiting delay. The service processing time delay and the service waiting time delay are fixed values under the determined calculation force requirement, and are not considered in the optimization. Total time delay t for user service i to reach computing node l il Is of formula (6):
when the storage resource and the computing resource of the computing node l meet the requirement of the user service i, the computing delay of the user service i
Is of formula (7):
s130: and obtaining a route distribution result under the condition of meeting a plurality of calculation requests by using an optimized route model through an immune genetic algorithm.
The route distribution result comprises optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay.
The optimized routing model can calculate the time delay from the user service to each calculation node by utilizing the immune genetic algorithm and time delay corresponding scheduling paths; and selecting the optimal time delay and the scheduling path corresponding to the optimal time delay from the user service to each computing node and the scheduling path corresponding to the time delay as a routing distribution result.
In some embodiments, the user traffic includes traffic priority; the optimized routing model comprises a service priority model; the step of obtaining route distribution results under the condition of meeting a plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm specifically comprises the following steps:
the optimized routing model calculates time delay from the user service to each calculation node by utilizing an immune genetic algorithm and a scheduling path corresponding to the time delay; based on the service priority, the service priority model determines the optimal time delay from the user service to each computing node and the scheduling path corresponding to the optimal time delay from the user service to each computing node in the scheduling paths corresponding to the time delay.
In the computing power network, a large amount of services exist, in order to meet the experience of more users, reasonable computing power resources are divided, a service priority model is built, and the following aspects are considered: number N of same service user demands UE (i) User satisfaction rate P Sat (i) Business importance level L Imp (i)。
And when the service obtains the required computational power resource and the total time delay for completing the task processing does not exceed the allowable maximum time delay, the service processing is considered to be successful. Wherein the epsilon (x) function is a step function when
When epsilon (x) has a value of 1; otherwise, ε (x) is 0./>
The function is a step function, when x is more than or equal to C i When (I)>
Has a value of 1; otherwise, go (L)>
Is 0. User satisfaction rate P Sat (i) Is of formula (8):
taking the above factors into consideration, the service i priority model G (i) is built as formula (9):
G(i)＝P Sat (i)L Imp (i)……(9)
and (9) optimizing the calculation task route and the calculation power resource allocation by comprehensively considering the service priority and the time delay, and establishing a task scheduling target. Wherein I is l Representing all traffic sets scheduled to compute node l. The constraint conditions are as follows: the storage resources available for all user traffic scheduled to a computing node do not exceed the storage resources S that the computing node can provide l The method comprises the steps of carrying out a first treatment on the surface of the The computing resources available for all user traffic scheduled to a computing node do not exceed the computing resources C that the computing node can provide l The method comprises the following steps:
in this embodiment, under constraint of computing resources and storage resources, the scheduling transmission delay of user service is minimized, and the priority level of joint service is used as an optimization target, so as to optimize route control, selection and storage of computing nodes, and allocation of computing resources.
Optionally, the step of obtaining the route allocation result under the condition of meeting a plurality of calculation requests by using an optimized route model through an immune genetic algorithm specifically includes:
setting a computing node distribution vector of user service as an antigen of an immune genetic algorithm, and setting a routing node path matrix of a path from the user service to the computing node as an antibody of the immune genetic algorithm; determining initial parameters and generating an initial population of antibodies for the immune genetic algorithm based on the initial parameters; evaluating each antibody in the initial population of antibodies to obtain an affinity for each antibody, wherein the affinity represents a degree of matching between the antibody and the antigen; calculating the concentration of the antibody according to the affinity degree to promote and inhibit the antibody so as to obtain the polyfitness; optimizing the antibody population according to the polymerization adaptability to obtain an optimal solution of an immune genetic algorithm; and outputting the optimal solution of the immune genetic algorithm as a route allocation result.
The method comprises the steps of optimizing an antibody population according to polymerization fitness to obtain an optimal solution of an immune genetic algorithm, and comprises the following steps:
directly inheriting individuals with the fitness lower than a preset value to the next generation or generating new individuals through pairing and crossing and inheriting the new individuals to the next generation; storing an individual with the aggregation adaptability higher than a preset value as a good solution into a memory unit; calculating the fit degree and the corresponding good solution of the next generation until the preset termination condition is met, and outputting the optimal solution; wherein the optimal solution is the optimal solution of the good solution in the memory cell.
In some embodiments, the step prior to generating the initial population of antibodies of the immune genetic algorithm based on the initial parameters further comprises:
comparing the similarity of the antigen and the historical antigen to obtain a comparison result; judging whether the first response occurs or not based on the comparison result; if the answer is the primary answer, randomly generating an initial antibody group by a chaos optimization algorithm; if the primary responses are not, randomly generating a first initial antibody subgroup by a chaos optimization algorithm, taking a memory unit antibody of part of historical antigens from memory data as a second initial antibody subgroup, and forming an initial antibody subgroup by the first initial antibody subgroup and the second initial antibody subgroup.
In some embodiments, the step of directly inheriting individuals with aggregate fitness below a predetermined value to the next generation or generating new individuals by crossover pairing to inherit to the next generation specifically comprises:
selecting two antibodies from individuals with aggregation fitness lower than a preset value to perform cross probability calculation to obtain a cross probability result; determining whether to perform cross pairing according to the cross probability result; if cross pairing is carried out, carrying out variation judgment on the result generated by each cross pairing to obtain a variation judgment result; and carrying out variation adjustment on the result generated by the cross pairing based on the variation judgment result.
S140: and performing route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result.
In summary, the present embodiment provides a route allocation method of a computing power network, which determines a computing node in the computing power network by acquiring a plurality of computing requests of a plurality of user services; obtaining route distribution results under the condition of meeting a plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm; and performing route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result. The route distribution result comprises optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay. By means of the method, multi-objective joint optimization computing power perception route distribution can be provided for computing power network scenes, ubiquitous and heterogeneous computing power resources are efficiently and cooperatively perceived and utilized, endophytic perception and computing power self-adaption capacity of a communication network are improved, and network energy efficiency is improved.
The method comprises the steps of firstly calculating the shortest time delay from user service to each computing node and the corresponding specific scheduling path, and then selecting the computing node which accords with the joint optimization target as an optimal computing node, and taking the path obtained through an algorithm as an optimal routing strategy. The immune genetic algorithm is one of the cores of the present invention, and the immune genetic algorithm is further described below.
Optionally, a computing node allocation vector for user traffic is set
As an antigen for immune genetic algorithms. Setting user service toRouting node path matrix of computing node paths>
As antibodies to immune genetic algorithms. Wherein N is ij Indicating that the user traffic numbered i passes through the routing node N in the jth route ij . Assume that the maximum number of routing nodes of a routing path is Ja. Converting the routing node path matrix into a vector containing i×ja elements, ++>
The whole population is expressed as->
I.e. a matrix of results sets of N rows I x Ja columns of N antibodies.
Because the antigen computing node distributes the situation, namely the computing node distribution result corresponding to each service currently, the two antigens are identical when the two distribution situations are identical. Antigens
The hamming distance shown in the difference expression (11) between them is calculated. Description->
The two antigens compared were identical.
In recognizing an antigen, the current antigen is compared to the historical antigens in all memory cells. Thus, the current antigen
And memory cell->
After comparison of all historical antigens, the result is calculated by the formula (12)To the minimum value D of the degree of difference q,m . Not only D q,m >0, judging the antigen primary response; d (D) q,m When=0, the antigen is judged as non-primary response.
The immune genetic algorithm introduces a concentration regulation mechanism, optimizing the selection strategy as: the higher the fitness, the higher the probability of antibody selection with lower concentration; the lower the fitness, the lower the probability that an individual with a higher concentration will be selected. Thus avoiding blindness of crossover variation of the genetic algorithm, maintaining diversity of population, reducing repeated work and improving algorithm efficiency. The degree of approximation between antibodies is first calculated from the similarity function. And calculating the concentration of the antibody according to the proportion of the similar antibody in the population. Finally, the concentration and the affinity are combined to obtain a polymerization fitness function which is used as the basis for finally selecting the antibody.
Antibody G x ，G y The degree of difference between them is calculated with the Euclidean distance as shown in expression (13):
the similarity calculation between antibodies is shown in formula (14):
the concentration of the kth antibody in the population popu is calculated as shown in formula (15):
wherein the epsilon (x) function is a step function when
When epsilon (x) has a value of 1; otherwise, ε (x) is 0. And Θ is an antibody similarity judgment threshold.
In the route allocation strategy, the optimization objective is to minimize
The affinity Ag of the antibody was calculated as formula (16):
the higher the affinity and the higher the polymerization adaptability of the antibody at a lower concentration, the greater the probability of selection.
Each operation firstly selects two antibodies through selection operation, then judges whether to cross according to the cross probability, and finally adjusts the crossed result according to the variation probability. Wherein Θ is slct ，Θ crss ，Θ muta Random numbers generated by three operators respectively, R c ，R m Probability values set for crossover and mutation.
(i) Selecting: the selection mechanism uses a roulette mechanism, PN of formula (18) k The value of (2) represents the kth antibody selection probability ratio. The higher the value of the fitness function, the greater the probability that the antibody will be selected for reproduction. When PN is the k-1 ＜Θ slct ＜PN k And when the kth antibody is the selected result, performing the next cross operation.
(ii) Crossing: the two antibodies selected by the previous generation cross-bred the antibodies of the next generation. When theta is crss ≥R c At this time, the selected antibody is directly inherited to the next generation. When theta is crss ＜R c When two selected precursors are crossed, antibodies
Obtaining integer result rg by random function which obeys uniform distribution, and crossing by using the positioning result to generate next antibody result +.>
(iii) Variation: the variation judgment is carried out on the result generated by each crossing, if Θ muta ≥R m If it is not operated muta ＜R m Then, the individual generated by the crossover is finely tuned according to the formula (19). Wherein, psi is g Is a trim value. The probability of each bit of the new antibody is adjusted by the result of the random function rand. In addition, when
And when the value of (2) exceeds the upper limit and the lower limit, taking a threshold value. The value added adjustment range is to calculate the number range of all routing nodes.
The specific steps of the immune genetic algorithm are described as follows:
(1) Function initialization and antigen recognition. Initializing parameter settings and based on the input problem antigens
With historical antigen->
Similarity D of (2) q,m Comparing and judgingWhether an initial response has occurred. Parameters include population size N, maximum number of iterations T, k c ,k' c ,k m ,k' m ,w。
(2) An initial population is generated. Generating an initial population of antibodies based on the determination of step (1)
If the response is not the primary response, the N/2 memory unit antibody is called from the memory data, and the rest N/2 antibodies are randomly generated by a chaos optimization algorithm. If the response is the primary response, the initial antibody is formed by random generation of a chaos optimization algorithm.
(3) The affinity of the antibodies was calculated. Each antibody in the population was evaluated after initialization and encoding. By affinity A g The degree of matching between the antibody and antigen, i.e., the degree of goodness of the feasible solution in the iterative process, is expressed. The higher the affinity, the higher the matching of the antibody.
(4) Antibody concentrations were calculated to promote and inhibit antibodies. The higher the affinity of the antibody, the higher the probability of proceeding to the next generation, which easily leads to population evolution singleness, leading to local optimization. Thus, it is necessary to rely on the concentration C of the antibody g Promoting and inhibiting the reaction to obtain the polymerization fitness F g fit ＝A g /C g 。
(5) The immune operator operates on the antibody population. And directly inheriting the optimized individuals with low aggregation fitness to the next generation or generating new individuals through pairing and crossing to inherit the individuals to the next generation according to the calculated aggregation fitness result. Each operation firstly selects two antibodies by a selection operation, and then the cross probability R c Judging whether to cross or not, and finally according to the variation probability R m And adjusting the result after the crossing.
(6) Updating the population and the memory unit. And updating the population according to the immune operator result, and storing the individuals with high poly-fitness to the memory unit to ensure the retention of the excellent solution.
(7) And judging to end according to the termination condition. And if the termination condition is met, outputting the optimal solution as a final result. Otherwise, jumping to the step (3), and continuing iteration.
In summary, according to the route distribution method of the power network in the embodiment, an optimized route model is established through time delay and service priority, and an immune genetic algorithm is applied to realize route distribution strategies, so that multi-objective combined optimized power-aware route distribution can be provided for a power network scene, power and network collaborative management is realized, and the problem of uneven network load is relieved. Under the constraint limit of computing resources and storage resources, the scheduling transmission delay of user service is minimized, and the priority level of the joint service is used as an optimization target to optimize route control, selection and storage of computing nodes and distribution of computing power resources.
The route allocation device of the power calculation network provided by the invention is described below, and the route allocation device of the power calculation network described below and the route allocation method of the power calculation network described above can be referred to correspondingly.
Referring to fig. 2, fig. 2 is a schematic structural diagram of an embodiment of a routing distribution device of a computing network according to the present invention, where the routing distribution device of the computing network may include:
the user service module 210 is configured to obtain a plurality of computing requests for a plurality of user services in the computing power network.
The computing node module 220 is configured to determine computing nodes in the computing power network.
An optimized routing model module 230, configured to obtain a routing assignment result under the condition that a plurality of calculation requests are satisfied by using an optimized routing model through an immune genetic algorithm; the route distribution result comprises optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay.
The allocation module 240 is configured to perform route allocation on a plurality of computing requests of a plurality of user services based on the route allocation result.
In some embodiments, the user traffic includes traffic priority; the optimized route model module 230 includes an optimized route model and a traffic priority model. Specifically:
the optimized routing model calculates time delay from the user service to each calculation node by utilizing an immune genetic algorithm and a scheduling path corresponding to the time delay; based on the service priority, the service priority model determines the optimal time delay from the user service to each computing node and the scheduling path corresponding to the optimal time delay from the user service to each computing node in the scheduling paths corresponding to the time delay.
In some embodiments, the optimized route model module 230 is to:
setting a computing node distribution vector of user service as an antigen of an immune genetic algorithm, and setting a routing node path matrix of a path from the user service to the computing node as an antibody of the immune genetic algorithm; determining initial parameters and generating an initial population of antibodies for the immune genetic algorithm based on the initial parameters; evaluating each antibody in the initial population of antibodies to obtain an affinity for each antibody, wherein the affinity represents a degree of matching between the antibody and the antigen; calculating the concentration of the antibody according to the affinity degree to promote and inhibit the antibody so as to obtain the polyfitness; optimizing the antibody population according to the polymerization adaptability to obtain an optimal solution of an immune genetic algorithm; and outputting the optimal solution of the immune genetic algorithm as a route allocation result.
In some embodiments, the optimized routing model module 230 is further configured to:
directly inheriting individuals with the fitness lower than a preset value to the next generation or generating new individuals through pairing and crossing and inheriting the new individuals to the next generation; storing an individual with the aggregation adaptability higher than a preset value as a good solution into a memory unit; calculating the fit degree and the corresponding good solution of the next generation until the preset termination condition is met, and outputting the optimal solution; wherein the optimal solution is the optimal solution of the good solution in the memory cell.
In some embodiments, the optimized routing model module 230 is further configured to:
comparing the similarity of the antigen and the historical antigen to obtain a comparison result; judging whether the first response occurs or not based on the comparison result; if the answer is the primary answer, randomly generating an initial antibody group by a chaos optimization algorithm; if the primary responses are not, randomly generating a first initial antibody subgroup by a chaos optimization algorithm, taking a memory unit antibody of part of historical antigens from memory data as a second initial antibody subgroup, and forming an initial antibody subgroup by the first initial antibody subgroup and the second initial antibody subgroup.
In some embodiments, the optimized routing model module 230 is further configured to:
selecting two antibodies from individuals with aggregation fitness lower than a preset value to perform cross probability calculation to obtain a cross probability result; determining whether to perform cross pairing according to the cross probability result; if cross pairing is carried out, carrying out variation judgment on the result generated by each cross pairing to obtain a variation judgment result; and carrying out variation adjustment on the result generated by the cross pairing based on the variation judgment result.
The invention also provides an electronic device, refer to fig. 3, and fig. 3 is a schematic structural diagram of an embodiment of the electronic device. In this embodiment, the electronic device may include a memory (memory) 320, a processor (processor) 310, and a computer program stored on the memory 320 and executable on the processor 310. The processor 310, when executing the program, implements the method of route allocation for the power network provided by the methods described above.
Optionally, the electronic device may further comprise a communication bus 330 and a communication interface (Communications Interface) 340, wherein the processor 310, the communication interface 340, and the memory 320 communicate with each other via the communication bus 330. The processor 310 may invoke logic instructions in the memory 320 to perform a method of route allocation for a computing power network, the method comprising:
acquiring a plurality of calculation requests of a plurality of user services in a computing power network; determining a computing node in a computing power network; obtaining a route distribution result under the condition of meeting a plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm; the route distribution result comprises optimal time delay from user service to each computing node and a scheduling path corresponding to the optimal time delay; and performing route distribution on a plurality of calculation requests of a plurality of user services based on the route distribution result.
Further, the logic instructions in the memory 320 described above may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a stand alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention further provides a non-transitory computer readable storage medium, on which a computer program is stored, where the computer program is implemented when executed by a processor to perform the method for route allocation of a computing power network provided by the foregoing methods, and the steps and principles of the method are described in detail in the foregoing methods and are not repeated herein.
In yet another aspect, the present invention further provides a computer program product, where the computer program product includes a computer program, where the computer program can be stored on a non-transitory computer readable storage medium, where the computer program, when executed by a processor, can perform a method for route allocation of a computing network provided by the above methods, and the steps and principles of the method are described in detail in the above methods and are not described herein.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for route allocation for a power network, comprising:
acquiring a plurality of calculation requests of a plurality of user services in a computing power network;
determining a computing node in the computing power network;
obtaining a route distribution result under the condition of meeting the plurality of calculation requests by utilizing an optimized route model through an immune genetic algorithm; the route distribution result comprises optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay;
and performing route distribution on a plurality of calculation requests of the plurality of user services based on the route distribution result.
2. The method of route allocation for a power network of claim 1, wherein the user traffic comprises traffic priority; the optimized route model comprises a service priority model, the route allocation result under the condition of meeting the plurality of calculation requests is obtained by utilizing the optimized route model through an immune genetic algorithm, and the method comprises the following steps:
the optimized routing model calculates the time delay from the user service to each calculation node by using the immune genetic algorithm and a scheduling path corresponding to the time delay;
and determining the optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay from the user service to each computing node in the time delay of the user service to each computing node and the scheduling path corresponding to the time delay based on the service priority.
3. The route allocation method of a computing power network according to claim 1, wherein the obtaining, by an immune genetic algorithm, route allocation results under satisfaction of the plurality of computing requests using an optimized route model comprises:
setting a computing node distribution vector of the user service as an antigen of an immune genetic algorithm, and setting a routing node path matrix of the user service to the computing node path as an antibody of the immune genetic algorithm;
determining an initial parameter and generating an initial population of antibodies for the immune genetic algorithm based on the initial parameter;
evaluating each antibody in the initial population of antibodies to obtain an affinity for each antibody, wherein the affinity represents a degree of matching between the antibody and the antigen;
calculating the concentration of the antibody according to the affinity to promote and inhibit the antibody so as to obtain the poly-fitness;
optimizing the antibody population according to the polymerization fitness to obtain an optimal solution of the immune genetic algorithm;
and outputting the optimal solution of the immune genetic algorithm as a route allocation result.
4. The method of claim 3, wherein optimizing the population of antibodies according to the aggregate fitness to obtain an optimal solution of the immune genetic algorithm comprises:
directly inheriting the individuals with the aggregation fitness lower than a preset value to the next generation or generating new individuals through pairing and crossing and inheriting the new individuals to the next generation;
storing the individual with the aggregation fitness higher than the preset value as a good solution into a memory unit;
calculating the fit degree and the corresponding good solution of the next generation until the preset termination condition is met, and outputting the optimal solution; wherein the optimal solution is the optimal solution of the good solution in the memory unit.
5. The method of route assignment to a computational power network of claim 3, further comprising, prior to generating the initial population of antibodies to the immune genetic algorithm based on the initial parameters:
comparing the similarity of the antigen and the historical antigen to obtain a comparison result;
judging whether the first response occurs or not based on the comparison result;
if the answer is the primary answer, randomly generating an initial antibody group by a chaos optimization algorithm;
if the primary responses are not, randomly generating a first initial antibody subgroup by a chaos optimization algorithm, taking a memory unit antibody of part of historical antigens from memory data as a second initial antibody subgroup, and forming the initial antibody subgroup by the first initial antibody subgroup and the second initial antibody subgroup.
6. The method of claim 4, wherein directly inheriting individuals with aggregate fitness below a predetermined value to a next generation or generating new individuals by pairing crossing to a next generation comprises:
selecting two antibodies from the individuals with the aggregation fitness lower than a preset value to perform cross probability calculation to obtain a cross probability result;
determining whether to perform cross pairing according to the cross probability result;
if cross pairing is carried out, carrying out variation judgment on the result generated by each cross pairing to obtain a variation judgment result;
and carrying out variation adjustment on the result generated by the cross pairing based on the variation judging result.
7. A routing distribution device for a computing power network, comprising:
the user service module is used for acquiring a plurality of calculation requests of a plurality of user services in the computing power network;
a computing node module for determining computing nodes in the power network;
the optimized route model module is used for obtaining route distribution results under the condition that the plurality of calculation requests are met through an immune genetic algorithm by utilizing an optimized route model; the route distribution result comprises optimal time delay from the user service to each computing node and a scheduling path corresponding to the optimal time delay;
and the distribution module is used for carrying out route distribution on a plurality of calculation requests of the plurality of user services based on the route distribution result.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the route allocation method of the computing power network of any one of claims 1 to 6 when the program is executed by the processor.
9. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a route allocation method of a computing power network according to any of claims 1 to 6.
10. A computer program product comprising a computer program which, when executed by a processor, implements a method of route allocation for a power network according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211534614.9A CN116132403A (en) | 2022-11-29 | 2022-11-29 | Route distribution method and device of computing power network, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211534614.9A CN116132403A (en) | 2022-11-29 | 2022-11-29 | Route distribution method and device of computing power network, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116132403A true CN116132403A (en) | 2023-05-16 |
Family
ID=86293811
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211534614.9A Pending CN116132403A (en) | 2022-11-29 | 2022-11-29 | Route distribution method and device of computing power network, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116132403A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385857A (en) * | 2023-06-02 | 2023-07-04 | 山东协和学院 | Calculation power distribution method based on AI intelligent scheduling |
-
2022
- 2022-11-29 CN CN202211534614.9A patent/CN116132403A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385857A (en) * | 2023-06-02 | 2023-07-04 | 山东协和学院 | Calculation power distribution method based on AI intelligent scheduling |
| CN116385857B (en) * | 2023-06-02 | 2023-08-18 | 山东协和学院 | Calculation power distribution method based on AI intelligent scheduling |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109947545B (en) | Task unloading and migration decision method based on user mobility | |
| CN111372314A (en) | Task unloading method and task unloading device based on mobile edge computing scene | |
| CN112512056B (en) | Multi-objective optimization calculation unloading method in mobile edge calculation network | |
| CN111800828B (en) | Mobile edge computing resource allocation method for ultra-dense network | |
| CN109474980A (en) | A kind of wireless network resource distribution method based on depth enhancing study | |
| CN111586720A (en) | Task unloading and resource allocation combined optimization method in multi-cell scene | |
| CN110492955B (en) | Spectrum prediction switching method based on transfer learning strategy | |
| CN111475274A (en) | Cloud collaborative multi-task scheduling method and device | |
| Zhang et al. | Blockchain-based multi-access edge computing for future vehicular networks: A deep compressed neural network approach | |
| CN109361725A (en) | Car networking cloud system resource allocation methods based on multi-objective genetic algorithm | |
| CN113407249B (en) | Task unloading method facing to position privacy protection | |
| CN109905864A (en) | A kind of cross-layer Resource Allocation Formula towards electric power Internet of Things | |
| CN116132403A (en) | Route distribution method and device of computing power network, electronic equipment and storage medium | |
| Yan et al. | Joint user access mode selection and content popularity prediction in non-orthogonal multiple access-based F-RANs | |
| Qiu et al. | Subchannel assignment and power allocation for time-varying fog radio access network with NOMA | |
| Wang et al. | Multi-objective joint optimization of communication-computation-caching resources in mobile edge computing | |
| CN112566131A (en) | C-RAN network resource allocation method based on time delay constraint | |
| CN114143814B (en) | Multi-task unloading method and system based on heterogeneous edge cloud architecture | |
| CN114615705B (en) | Single-user resource allocation strategy method based on 5G network | |
| Liu et al. | Distributed Computation Offloading with Low Latency for Artificial Intelligence in Vehicular Networking | |
| CN116089091A (en) | Resource allocation and task unloading method based on edge calculation of Internet of things | |
| CN112770343B (en) | D2D-NOMA resource allocation method and system based on HAGA | |
| CN109152060A (en) | Transmitter channel distribution model and method in a kind of shortwave downlink communication | |
| CN113873525A (en) | Task unloading method and terminal for ultra-dense edge computing network | |
| CN109819522B (en) | User bandwidth resource allocation method for balancing energy consumption and user service quality |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |