CN108712337B - Multipath bandwidth scheduling method in high-performance network - Google Patents
Multipath bandwidth scheduling method in high-performance network Download PDFInfo
- Publication number
- CN108712337B CN108712337B CN201810383730.2A CN201810383730A CN108712337B CN 108712337 B CN108712337 B CN 108712337B CN 201810383730 A CN201810383730 A CN 201810383730A CN 108712337 B CN108712337 B CN 108712337B
- Authority
- CN
- China
- Prior art keywords
- time
- request
- bandwidth
- time windows
- available
- 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.)
- Active
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
- H04L45/12—Shortest path evaluation
- H04L45/125—Shortest path evaluation based on throughput or bandwidth
-
- 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
- H04L45/24—Multipath
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/70—Admission control; Resource allocation
- H04L47/72—Admission control; Resource allocation using reservation actions during connection setup
- H04L47/724—Admission control; Resource allocation using reservation actions during connection setup at intermediate nodes, e.g. resource reservation protocol [RSVP]
Abstract
The invention discloses a multipath bandwidth scheduling method in a high-performance network, which considers the dynamic change condition of bandwidth in the high-performance network with a bandwidth reservation function, aims at maximizing the success rate of user request for a plurality of given bandwidth reservation requests with deadline limit, simultaneously transmits data by a plurality of paths, considers the average completion time and the shortest duration time and improves the success rate of data transmission.
Description
Technical Field
The invention belongs to the technical field of computer networks, and relates to a multipath bandwidth scheduling method in a high-performance network.
Background
The era of big data has come, and a great deal of simulation, observation and experimental data are continuously generated at high speed, such as data of earthquake simulation, computational biology, high-energy physics, environmental monitoring and the like, and after the data are generated, the data need to be timely transmitted to user hosts with different geographical positions for data storage or analysis and other works. In the face of the large PB-level data volume, the traditional internet is not compelling to meet the requirement of high-efficiency transmission of big data, and at this time, a High Performance Network (HPN) with high transmission speed, small delay, reliability and stability becomes the first choice network for big data transmission, and its important significance has been acknowledged by the scientific and network research community. China has great development potential in the field of high-performance networks, and needs a great deal of high-quality related research to enable expensive computing resources and network resources to play the true role.
The high-performance network is a network with bandwidth reservation capability, wherein the bandwidth scheduling problem is an extremely important problem, which is directly related to the utilization rate of network resources and the satisfaction degree of users, and particularly, many scientific application tasks need schedulable and flexible network services. In order to evaluate the performance of the bandwidth scheduling algorithm, the quality of the algorithm should be measured mainly from the aspect of the success rate of the user request.
Disclosure of Invention
Aiming at the problems in the prior art, the invention aims to provide a multi-path bandwidth scheduling method in a high-performance network, which aims to maximize the success rate of user requests, simultaneously transmits data by a plurality of paths, considers the average completion time and the shortest duration time and improves the success rate of data transmission.
In order to achieve the purpose, the invention adopts the following technical scheme:
the multipath bandwidth scheduling method in the high-performance network comprises the following steps:
In the step 4, the step of,calculate the ith request QiJ (th) time window WjAccumulated value of available bandwidthIf B issum,ij<Bmin,ijThen step 5 is performed, wherein Bmin,ijIn order to minimize the bandwidth requirements,tl jfor maximum data transmission time length, DiFor the ith request QiThe amount of data transmission of (a) is,wherein the content of the first and second substances,is the jth time window WjThe starting time of (a) is,is the jth time window WjThe cutoff time of (d); otherwise, the output is at the ith request QiAll available paths and available bandwidth of the available paths are obtained, and step 7 is executed;
wherein A isijkIndicates the ith request QiThe j-th time window W obtainedjActual reserved bandwidth of the kth available path under, BijkIndicates the ith request QiThe j-th time window W obtainedjAvailable bandwidth of the k-th available path ofsum,ijIndicates the ith request QiJ (th) time window WjAccumulated value of available bandwidth of bijIndicates the ith request QiThe next jth time window WjTotal reserved bandwidth over all available paths obtained down, bij=Bmin,ij;
Optionally, in the step 2, the multiple time windows in the high-performance network are sorted according to their deadlines and durations to obtain a sorted time window sequence W1,W2,…Wj,…WJThe method comprises the following steps:
arranging a plurality of time windows in the high-performance network according to the ascending order of the cut-off time of the time windows, and arranging the time windows with the same cut-off time according to the descending order of the duration time of the time windows to obtain a sequence W of the time windows after the sequencing1,W2,…Wj,…WJWherein W isjRepresents the jth request and J represents the total number of time windows.
Optionally, in the step 2, the multiple time windows in the high-performance network are sorted according to their deadlines and durations to obtain a sorted time window sequence W1,W2,…Wj,…WJThe following method can also be adopted:
for high performance networkThe plurality of time windows are arranged according to the descending order of the duration time of the time windows, and the time windows with the same cut-off time are arranged according to the ascending order of the cut-off time of the time windows to obtain the ordered time window sequence W1,W2,…Wj,…WJWherein W isjRepresents the jth request and J represents the total number of time windows.
Compared with the prior art, the invention has the following technical effects: the method of the invention considers the dynamic change situation of the bandwidth in the high-performance network with the bandwidth reservation function, and simultaneously transmits data by a plurality of paths for a plurality of given bandwidth reservation requests with the deadline limit with the aim of maximizing the success rate of the user request, and considers the average completion time and the shortest duration time, thereby improving the success rate of data transmission.
The embodiments of the invention will be explained and explained in further detail with reference to the figures and the detailed description.
Drawings
Fig. 1 is a schematic diagram of a high-performance network topology and corresponding available bandwidth of each link, wherein (a) represents the high-performance network topology, and (b) represents the available bandwidth of each link;
FIG. 2 is a diagram of available bandwidth for each link after scheduling multiple requests in turn, where (a) represents a scheduling request brr2The available bandwidth of each link, (b) represents a scheduling request brr3The available bandwidth of each link, (c) represents a scheduling request brr0The available bandwidth of each link, and (d) represents a scheduling request brr1The available bandwidth of each link.
FIG. 3 is a graph showing a comparison of success rates of the method (MINBP-ECT) and FBR-ECT in accordance with an embodiment of the present invention.
FIG. 4 is a graph comparing the earliest completion time of the method (MINBP-ECT) and FBR-ECT in accordance with an embodiment of the present invention.
FIG. 5 is a graph showing the success rate comparison between the method (MINBP-SD) and FBR-SD) according to an embodiment of the present invention.
FIG. 6 is a graph comparing the method (MINBP-SD) and FBR-SD minimum duration in one embodiment of the invention.
Detailed Description
A multi-path bandwidth scheduling method in a high-performance network comprises the following steps:
Request for (v)s,vd,Bmax,D,[tS,tE]) Is shown in the formula, wherein vsRepresenting the source node, vdRepresenting the destination node, BmaxRepresenting the maximum bandwidth of the local area network, D is the data transmission quantity, and the data with the data transmission quantity D needs to be in [ t ]S,tE]Is passed over a period of time of, wherein tSIs the earliest transmission time, tEIs the latest transmission time; maximum possible transmission time equal to tE-tS。
For a given high performance network, its time windows are determined, each time window comprising a start time and an end time. FIG. 1 shows a network topology of a high performance network and gives [0,10s ]]The available bandwidth on each link during the time period. Define the timeout as the longest time interval during which the available bandwidth of all links remains unchanged, e.g., there are 3 timepaces in fig. 1: [0,3s ]],[3s,6s],[6s,10s]. Defining a time window timewindow as a time period containing one or more consecutive timepies, similar to a timepies, withRepresents the jth time window timewindow, wherein,is the start time of the jth time window,is the cutoff time of the jth time window, e.g., except for [0,3s ] in FIG. 1],[3s,6s],[6s,10s]These 3 timewindows, also [0,6s ]],[0,10s],[3s,10s]3 timewindows.
wherein A isijkIndicates the ith request QiThe j-th time window W obtainedjActual reserved bandwidth of the kth available path under, BijkIndicates the ith request QiThe j-th time window W obtainedjAvailable bandwidth of the k-th available path ofsum,ijIndicates the ith request QiJ (th) time window WjAccumulated value of available bandwidth of bijIndicates the ith request QiThe next jth time window WjTotal reserved bandwidth over all available paths obtained down, bij=Bmin,ij;
Specifically, in another embodiment, the step 2 of sorting the plurality of time windows in the high-performance network according to their deadlines and durations obtains a sequence W of sorted time windows1,W2,…Wj,…WJThe method comprises the following steps:
arranging a plurality of time windows (timewindows) in a high-performance network in ascending order according to the cut-off time of the time windows, and arranging the time windows with the same cut-off time in descending order according to the duration of the time windows to obtain a sequence W of the time windows after sequencing1,W2,…Wj,…WJWherein W isjRepresents the jth request and J represents the total number of time windows.
In the embodiment, the technical scheme is adopted for bandwidth scheduling, so that the success rate maximization and the earliest average completion time during data transmission can be met.
Fig. 1 (a) is a simple high-performance network topology diagram, fig. 1 (b) is a diagram of available bandwidth of each corresponding link, and table 1 is a plurality of request DCBRRs. The ascending order of time points yields LTP ═ 0,3s,6s,10s, and the ordered list of requests is { brr }2,brr3,brr0,brr1}. Pair brr2,TWL={[0,3s],[0,6s],[3s,6s],[0,10s],[3s,10s]TWL denotes a sequence of time windows, tmin2s, with tminIndicating the minimum duration of the data transmission,[0,3s]when t isl=3s,BminThe path with the largest available bandwidth is B-a-c, B-8 Gb/s>4Gb/s,Bsum=8Gb/s,min(Bsum,Bmin)·tl12Gb ═ D, so path b-aC, data transmission can be completed, the network is updated, and the updated available bandwidth of each link is as shown in (a) in fig. 2; pair brr3,tmin=2s,TWL={[0,10s],[3s,10s]And [6s, 10s ], in [0,10s ]]Then, the paths a-b can complete data transmission and update the network, and the updated available bandwidth of each link is as shown in (b) of fig. 2; pair brr0,tmin3s, the TWL after sorting { [0,3s { ]],[0,6s],[0,10s]At [0,3s ]]When t isl=3s,BminThe path with the largest available bandwidth a-B is 8Gb/s, but when B is 6Gb/s < 8Gb/s, the a-B link is removed from the topology, in the new topology a-c-B is the path with the largest available bandwidth and B is 2Gb/s, Bsum=6Gb/s+2Gb/s=8Gb/s=Bmin,min(Bsum,Bmin)·tlD at 24Gb, so that the path a-b and the path a-c-b can complete data transmission, and update the network, and the updated available bandwidth of each link is as shown in (c) of fig. 2; pair brr1,tmin3s, the TWL after ranking { [0,6s { [],[3s,6s],[0,10s],[3s,10s]At [3s, 6s ]]When the data transmission is completed, the network is updated, and the updated available bandwidth of each link is as shown in (d) of fig. 2. All requests DCBRR may be successfully scheduled with an average deadline ECT of 5.5 s.
TABLE 1
# | υs | υd | Bmax(Gb/s) | D(Gb) | [tS,tE](secs) |
brr0 | a | |
8 | 24 | [0,4] |
brr1 | a | |
9 | 27 | [3,6] |
brr2 | b | c | 6 | 12 | [0,6] |
brr3 | a | |
10 | 20 | [6,10] |
Optionally, the implementation manner of step 2 may also adopt the following scheme:
arranging a plurality of time windows (timewindows) in a high-performance network in a descending order according to the duration of the time windows, and arranging the time windows with the same cut-off time in an ascending order according to the cut-off time of the time windows to obtain the sorted time windowsMouth sequence W1,W2,…Wj,…WJWherein W isjRepresents the jth request and J represents the total number of time windows. In this example, the average SD obtained was 3.25 s. In the embodiment, the technical scheme is adopted for bandwidth scheduling, so that the maximization of the success rate and the shortest duration of data transmission can be met.
Example (b):
the performances of MINBP-ECT/SD and FER-ECT/SD are compared through a large number of simulation experiments, and in order to simulate a real ESnet environment, ESnet network topology is established through real ESnet data obtained on line. Wherein MINBP-ECT represents that in the method of the invention, when step 2 adopts a plurality of time windows (timewindows) in the high-performance network to be arranged according to the ascending order of the cut-off time of the time windows, the time windows with the same cut-off time are arranged according to the descending order of the duration time of the time windows, and the sequenced time window sequence W is obtained1,W2,…Wj,…WJA method of this implementation. MINBP-SD shows that in the method of the invention, when step 2 adopts a plurality of time windows (timewindows) in a high-performance network to be arranged in descending order according to the duration of the time windows, the time windows with the same cut-off time are arranged in ascending order according to the cut-off time of the time windows, and a sequence W of the time windows after sorting is obtained1,W2,…Wj,…WJA method of this implementation. In [0,20s ]]Randomly generating a plurality of request DCBRRs: randomly generating source node vsAnd destination node vd(vs≠vd) Maximum bandwidth BmaxAt [1Gb/s,20Gb/s]Is randomly selected, and starts at time tSIn [0,19s ]]Internal random value, cut-off time tEAt [ t ]S,20s]The internal random value is taken, and the data quantity D is in [ B ]max,Bmax×(tE-tS)]And randomly taking values. A total of 10 experiments were performed, with experiment 1 generating 10 random DCBRRs, after which each group was incremented by 10 random DCBRRs, and all algorithms processed the same random DCBRRs in each group of experiments, with each group of experiments randomized 10 times.
(1) MINBP-ECT and FBR-ECT Performance analysis
First the scheduling success rate ssr is considered, then the deadline ECT is considered. The mean and standard deviation of ssr and ECT were calculated and plotted as shown in figures 3 and 4. The ssr of MINBP-ECT is 12.04% higher than that of FBR-ECT, the average cut-off time of FBR-ECT is 6% higher than that of MINBP-ECT, MINBP-ECT uses shorter completion time while successfully scheduling more DCBRRs, and the performance is obviously better than that of FBR-ECT, and MINBP-ECT has better scheduling performance than that of FBR-ECT.
(2) MINBP-SD and FBR-SD Performance analysis
First the success rate ssr is considered and then the duration SD. The mean and standard deviation of ssr and SD were calculated and plotted as shown in fig. 5 and 6. The ssr of MINBP-SD is 12.10% higher than that of FBR-SD, the average duration of FBR-SD is 22.22% more than that of MINBP-SD, MINBP-SD uses shorter duration while successfully scheduling more DCBRRs, and the performance is obviously better than that of FBR-SD, and MINBP-SD has better scheduling performance than that of FBR-SD.
Claims (3)
1. The multipath bandwidth scheduling method in the high-performance network is characterized by comprising the following steps:
step 1, arranging a plurality of given requests in ascending order according to the size of data transmission quantity, and arranging the requests with the same size of data transmission quantity in ascending order according to the maximum possible transmission time to obtain an ordered request sequence Q1,Q2,…Qi,…QIWherein Q isiRepresents the ith request, I represents the total number of requests; each request needs to be in a set time period tS,tE]Internal transmission is completed, wherein tSIs the earliest transmission time, tEIs the latest transmission time;
step 2, sequencing a plurality of time windows in the high-performance network according to the cutoff time and the duration time of the time windows to obtain a sequenced time window sequence W1,W2,…Wj,…WJWherein W isjRepresents the jth request, J represents the total number of time windows; each time window and the set time period tS,tE]Fully overlapping or partially overlapping;
step 3, in the network topologyDrawing Gk-1For the ith request Q in the request sequenceiAnd the jth time window W in the sequence of time windowsjPerforming the kth processing by using a path searching method, and if an available path P is obtainedijkAnd the available path PijkAvailable bandwidth BijkIf not, executing step 6; the initial value of i is 1, the initial value of j is 1, and the initial value of k is 1; when k is 1, the network topology Gk-1Initial network topology G for a provisioned high performance network0;
Step 4, calculating the ith request QiJ (th) time window WjAccumulated value of available bandwidthIf B issum,ij<Bmin,ijThen step 5 is performed, wherein Bmin,ijIn order to minimize the bandwidth requirements,tl jfor maximum data transmission time length, DiFor the ith request QiThe amount of data transmission of (a) is,wherein the content of the first and second substances,is the jth time window WjThe starting time of (a) is,is the jth time window WjThe cutoff time of (d); otherwise, the output is at the ith request QiAll available paths and available bandwidth of the available paths are obtained, and step 7 is executed;
step 5, from the network topology Gk-1Deleting the available path PijkTo obtain a new network topology graph GkThe network topology map in step 3 is obtainedGk-1Modifying to a new network topology GkAnd executing the step 3;
step 6, the jth time window W in the time window sequence in the step 3 is processedjModified to be the j +1 th time window Wj+1And executing the step 3;
step 7, calculate for ith request QiAnd (3) calculating the obtained actual reserved bandwidth of all available paths by adopting the formula (1):
wherein A isijkIndicates the ith request QiThe j-th time window W obtainedjActual reserved bandwidth of the kth available path under, BijkIndicates the ith request QiThe j-th time window W obtainedjAvailable bandwidth of the k-th available path ofsum,ijIndicates the ith request QiJ (th) time window WjAccumulated value of available bandwidth of bijIndicates the ith request QiThe next jth time window WjTotal reserved bandwidth over all available paths obtained down, bij=Bmin,ij;
Step 8, an initial network topological graph G of the given high-performance network is obtained0For the ith request QiModifying the obtained bandwidth of each available path in the corresponding time window into the actual reserved bandwidth of the available path;
step 9, the ith request Q in the request sequence in the step 3 is transmittediModified to be the (I + 1) th request and step 3 is performed until I ═ I.
2. The method as claimed in claim 1, wherein the step 2 of sorting the plurality of time windows in the high performance network according to their deadlines and durations to obtain a sequence W of sorted time windows1,W2,…Wj,…WJThe method comprises the following steps:
arranging a plurality of time windows in the high-performance network according to the ascending order of the cut-off time of the time windows, and arranging the time windows with the same cut-off time according to the descending order of the duration time of the time windows to obtain a sequence W of the time windows after the sequencing1,W2,…Wj,…WJWherein W isjRepresents the jth request and J represents the total number of time windows.
3. The method as claimed in claim 1, wherein the step 2 of sorting the plurality of time windows in the high performance network according to their deadlines and durations to obtain a sequence W of sorted time windows1,W2,…Wj,…WJThe method comprises the following steps:
arranging a plurality of time windows in a high-performance network according to the descending order of the duration time of the time windows, and arranging the time windows with the same cut-off time according to the ascending order of the cut-off time of the time windows to obtain a sequence W of the time windows after sequencing1,W2,…Wj,…WJWherein W isjRepresents the jth request and J represents the total number of time windows.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810383730.2A CN108712337B (en) | 2018-04-26 | 2018-04-26 | Multipath bandwidth scheduling method in high-performance network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810383730.2A CN108712337B (en) | 2018-04-26 | 2018-04-26 | Multipath bandwidth scheduling method in high-performance network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108712337A CN108712337A (en) | 2018-10-26 |
CN108712337B true CN108712337B (en) | 2020-07-07 |
Family
ID=63866898
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810383730.2A Active CN108712337B (en) | 2018-04-26 | 2018-04-26 | Multipath bandwidth scheduling method in high-performance network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108712337B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110417682B (en) * | 2019-06-24 | 2022-07-29 | 西北大学 | Periodic scheduling method in high-performance network |
CN111131421B (en) * | 2019-12-13 | 2022-07-29 | 中国科学院计算机网络信息中心 | Method for interconnection and intercommunication of industrial internet field big data and cloud information |
CN114826925B (en) * | 2022-04-01 | 2023-05-09 | 西北大学 | Big data transmission bandwidth scheduling method based on cost time sequence among data centers |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103036792A (en) * | 2013-01-07 | 2013-04-10 | 北京邮电大学 | Transmitting and scheduling method for maximizing minimal equity multiple data streams |
CN104023408A (en) * | 2014-06-26 | 2014-09-03 | 北京邮电大学 | Dispatcher and data dispatching method based on network multi-path parallel transmission |
CN106789744A (en) * | 2017-01-13 | 2017-05-31 | 重庆邮电大学 | Handled up resource regulating method with the height that time delay is perceived based on UNE virtualization |
CN106911567A (en) * | 2017-01-13 | 2017-06-30 | 西北大学 | The a plurality of fixed route bandwidth scheduling method towards in the high performance network of big data |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9820182B2 (en) * | 2013-07-12 | 2017-11-14 | Telefonaktiebolaget Lm Ericsson (Publ) | Method for enabling control of data packet flows belonging to different access technologies |
-
2018
- 2018-04-26 CN CN201810383730.2A patent/CN108712337B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103036792A (en) * | 2013-01-07 | 2013-04-10 | 北京邮电大学 | Transmitting and scheduling method for maximizing minimal equity multiple data streams |
CN104023408A (en) * | 2014-06-26 | 2014-09-03 | 北京邮电大学 | Dispatcher and data dispatching method based on network multi-path parallel transmission |
CN106789744A (en) * | 2017-01-13 | 2017-05-31 | 重庆邮电大学 | Handled up resource regulating method with the height that time delay is perceived based on UNE virtualization |
CN106911567A (en) * | 2017-01-13 | 2017-06-30 | 西北大学 | The a plurality of fixed route bandwidth scheduling method towards in the high performance network of big data |
Also Published As
Publication number | Publication date |
---|---|
CN108712337A (en) | 2018-10-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108712337B (en) | Multipath bandwidth scheduling method in high-performance network | |
CN113193984B (en) | Air-space-ground integrated network resource mapping method and system | |
CN104717124B (en) | A kind of friend recommendation method, apparatus and server | |
WO2017067505A1 (en) | Network capacity enlargement method and device | |
CN107193899B (en) | A kind of strongly connected graph division methods that nomography is friendly | |
CN113011602A (en) | Method and device for training federated model, electronic equipment and storage medium | |
WO2021147353A1 (en) | Order dispatch | |
CN106021560B (en) | The low time delay memory B+ tree index structuring method of Data Flow Oriented | |
US10346496B2 (en) | Information category obtaining method and apparatus | |
CN111209095B (en) | Pruning method based on tree search in DAG parallel task scheduling | |
CN109617710B (en) | Large data transmission bandwidth scheduling method with deadline constraint between data centers | |
CN103701469A (en) | Compression and storage method for large-scale image data | |
WO2021259246A1 (en) | Resource scheduling method and apparatus, electronic device, and computer-readable storage medium | |
CN112417500A (en) | Data stream statistical publishing method with privacy protection function | |
CN109146551A (en) | A kind of advertisement recommended method, server and computer-readable medium | |
CN106326005B (en) | Parameter automatic tuning method for iterative MapReduce operation | |
CN110048966B (en) | Coflow scheduling method for minimizing system overhead based on deadline | |
Hsu et al. | Virtual network mapping through path splitting and migration | |
CN111913800B (en) | Resource allocation method for optimizing cost of micro-service in cloud based on L-ACO | |
Wang et al. | Research on multi-agent task optimization and scheduling based on improved ant colony algorithm | |
CN110691000A (en) | Web service combination method based on fusion of FAHP and planning graph | |
CN107544848B (en) | Cluster expansion method, apparatus, electronic equipment and storage medium | |
CN109636709A (en) | A kind of figure calculation method suitable for heterogeneous platform | |
CN113868434A (en) | Data processing method, device and storage medium for graph database | |
CN109450684B (en) | Method and device for expanding physical node capacity of network slicing system |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |