CN114500296A - Communication, storage and computing resource unified characterization method based on function expansion diagram - Google Patents
Communication, storage and computing resource unified characterization method based on function expansion diagram Download PDFInfo
- Publication number
- CN114500296A CN114500296A CN202210087697.5A CN202210087697A CN114500296A CN 114500296 A CN114500296 A CN 114500296A CN 202210087697 A CN202210087697 A CN 202210087697A CN 114500296 A CN114500296 A CN 114500296A
- Authority
- CN
- China
- Prior art keywords
- node
- virtual
- data
- nodes
- functional
- 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.)
- Granted
Links
- 238000003860 storage Methods 0.000 title claims abstract description 84
- 238000004891 communication Methods 0.000 title claims abstract description 67
- 238000010586 diagram Methods 0.000 title claims abstract description 63
- 238000012512 characterization method Methods 0.000 title claims abstract description 8
- 230000006870 function Effects 0.000 claims abstract description 126
- 230000005540 biological transmission Effects 0.000 claims abstract description 90
- 238000004364 calculation method Methods 0.000 claims abstract description 14
- 238000007726 management method Methods 0.000 claims abstract description 7
- 238000000034 method Methods 0.000 claims description 10
- 230000004931 aggregating effect Effects 0.000 claims description 3
- 238000000354 decomposition reaction Methods 0.000 claims description 3
- 238000000638 solvent extraction Methods 0.000 claims description 3
- 238000010295 mobile communication Methods 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 3
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 241000220225 Malus Species 0.000 description 1
- 235000015197 apple juice Nutrition 0.000 description 1
- 235000021016 apples Nutrition 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/12—Discovery or management of network topologies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/142—Network analysis or design using statistical or mathematical methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Mathematical Physics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Computational Mathematics (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a unified characterization method of communication, storage and computing resources based on a function expansion diagram, which mainly solves the problem that the traditional time expansion diagram can not characterize the computing resources, and the realization scheme is as follows: initializing a characterization network parameter, dividing network nodes, and decomposing the network nodes according to the functions of the network nodes; dividing a planning period into T unequal time intervals; initializing a blank T-layer directed graph, and adding non-functional nodes, virtual sub-nodes and virtual computing nodes; adding a transmission link, a storage link and a virtual transmission link in the directed graph to form a function expansion graph; setting communication capacity constraint, storage capacity constraint, calculation capacity constraint and flow conservation constraint, and converting the joint management problem of communication, storage and calculation resources into a data flow problem in a function expansion diagram. The invention can use the function expansion diagram to represent the time-varying property and the relativity of the resources in the network, and can be used for the unified analysis and management of communication, storage and computing resources in the time-varying network.
Description
Technical Field
The invention belongs to the technical field of information, and particularly relates to a communication, storage and computing resource unified characterization method based on a function expansion diagram, which can be used for communication, storage and computing resource analysis and management of time-varying networks such as public transport, communication and supply chains.
Background
In order to model the influence of network topology on data transmission, Fulkerson et al propose a time expansion diagram, and connect discrete time snapshots by introducing a storage link, thereby implementing joint characterization of communication resources and storage resources of network nodes. Time expansion maps are widely used to characterize dynamic networks that change over time, such as multi-commodity problems, evacuation planning problems, spatial information networks, and communication networks. However, the network node has other computing processing functions in addition to the communication function and the storage function. For example, for a communication network, a network node may have an image processing function, and an original image flowing into the node may be compressed into a compressed image and then flows out from the node; for data flow problems, a network node may have processing functionality that converts incoming raw material into product after processing, such as processing apples into apple juice, and flows out of the node. However, such time-expansion diagrams do not characterize the computational processing functions of the nodes in the dynamic network.
For example, in a "Maximum flow routing protocol for space information network with service function constraints" article by husting Yang, a time-varying spatial information network is represented by using a time expansion diagram, and discrete time snapshots are connected by introducing a storage link, so that joint representation of communication resources and storage resources of network nodes is realized. However, nodes in the spatial information network, such as satellites, have other computing processing functions, such as image compression, in addition to communication and storage functions. The time expansion graph cannot take the corresponding computing resources of the spatial information network nodes into account, and therefore cannot be used for joint planning of communication, storage and computation in the spatial information network.
Disclosure of Invention
The invention aims to provide a unified characterization method of communication, storage and computing resources based on a function expansion diagram aiming at the defect that the prior art cannot jointly characterize the communication, storage and computing resources, so as to form a unified function diagram model, describe the connection and transformation relation among different resources and realize the analysis and management of the communication, storage and computing resources in a time-varying network.
In order to achieve the purpose, the technical scheme of the invention comprises the following steps:
(2) dividing network nodes, namely dividing nodes which can not provide a task computing function and only play a role in communication and storage in a network into non-functional nodes, and dividing nodes which can not only provide the communication and storage functions but also provide the computing function in the network into functional nodes; aggregating network nodes according to the partitioningExpressed as:whereinIs a collection of non-functional nodes that,in the form of a collection of functional nodes,denotes the jth non-functional node, N1The number of non-functional nodes is,denotes the ith functional node, N2Is the number of functional nodes, N ═ N1+N2;
(3) Decomposing the network nodes according to the functions of the network nodes:
if function nodeCan provide MiA computing function, then decomposing the node intoA virtual child node viAnd MiA virtual computing nodeAnd two virtual transmission linksAndwherein M isiAs a functional nodeThe total number of computing functions can be provided,functional node of a representationThe mth virtual compute node of the decomposition,representing a slave virtual child node viTo virtual computing nodeThe directional line segment of (a) is,representing slave virtual computing nodesTo virtual child node viIs directed line segment of (M ∈ [1, M)i];
(4) Planning the network period according to the connectivity of the network nodeDivided into T time intervalsWhereinAnd at a time interval tauqThe internal network topology is kept unchanged, and q belongs to [1, T ]];
(5) Constructing a function expansion diagram:
(5a) initializing a blank T-layer directed graph, wherein the time interval of the q-th layer directed graph is tauq;
(5b) At each time interval τ of the directed graphqAdding all non-functional nodes, virtual sub-nodes decomposed by all functional nodes and virtual computing nodes decomposed by all functional nodes in the network respectively to form a functional node graph and obtain three types of node sets of the functional node graph:
wherein,is a set of non-functional nodes of a functional node map,is a set of virtual child nodes of the functional node map,is a virtual compute node set of a functional node map,is expressed in time interval tauqNon-functional node of internal networkA copy of (a) is made of,for a virtual sub-node viA copy of (a) is made of,is expressed in time interval tauqIntra-network virtual compute nodeA copy of (1);
(5c) adding links in the functional node graph:
(5c1) adding a transmission link according to the connectivity of the node:
if at time interval τqThe jth non-functional node in the internal and external networkCan give the kth non-functional nodeTransmitting data, and then transmitting the data in the jth non-functional node in the functional node diagramAnd the kth non-functional nodeAdding a directed line segment between
If at time interval τqInner, j th non-functional nodeCan give the ith functional nodeTransmitting data, and then transmitting the data in the jth non-functional node in the functional node diagramAnd the ith virtual child nodeAdding a directed line segment between
If at time interval τqInner, i-th function nodeCan give the kth function nodeWhen data is transmitted, the ith virtual child node in the functional node diagramAnd the kth virtual child nodeAdding a directed line segment between
If at time interval τqInner, i-th function nodeCan give the jth non-functional nodeWhen data is transmitted, the ith virtual child node in the functional node diagramAnd the jth non-functional nodeAdding a directed line segment between
(5c2) Adding a storage link:
adding a slave node between adjacent time intervals of each non-functional node of the functional node mapTo the nodeDirected line segment of
Adding a slave node between adjacent time intervals of each virtual child node of the functional node graphTo the nodeDirected line segment of
(5c3) Adding a virtual transmission link: at each virtual child node of the functional node mapVirtual sub-computing node corresponding theretoAdding two directed line segments in betweenAndobtaining a function expansion diagram;
(6) setting communication capacity constraint, storage capacity constraint, calculation capacity constraint and flow conservation constraint:
the communication capacity constraint is to limit the sum of the data amount transmitted by all the data streams on the transmission link or the virtual transmission link not to exceed the communication capacity of the transmission link or the virtual transmission link;
the storage capacity constraint is to limit the sum of the data amount stored on the storage link of all the data streams to be not more than the storage capacity of the storage link;
the computing capacity constraint is to restrict the data flowStreaming virtual compute nodesThe consumed computing capacity cannot exceed the virtual computing nodeProvided computing power, whereinTo calculate the function for the upcoming receptionThe data stream of (a) is transmitted,m∈[1,Mi],shown is a virtual flow into the functional expansion diagram through the transmission link and the storage linkPseudo child nodeThe number of types of different data streams;
the flow conservation constraint includes: defining the amount of data flowing into the non-functional node of each data stream to be equal to the amount of data flowing out of the non-functional node; defining the amount of data flowing into the virtual child node for each data stream to be equal to the amount of data flowing out of the virtual child node; qualifying instant data streamsStreaming virtual compute nodesIs multiplied by the amount of data ofEqualing data streamsEgress virtual compute nodeThe amount of data of (a), wherein,computing functions for receivedThe data stream of (2);
(7) under the four set constraints, the problem of joint management of communication, storage and computing resources is converted into a data flow problem in a function expansion diagram, namely the function expansion diagram is used for uniformly representing dynamic network communication, storage and computing resources changing along with time.
Compared with the prior art, the invention has the following advantages:
1) the invention uniformly represents communication, storage and computing capabilities through the function expansion diagram framework, thereby solving the problem that a plurality of computing functions in one node cannot be represented in the traditional time expansion diagram. Specifically, each node with a computing function is virtually decomposed into three virtual components based on a conventional time expansion graph: the system comprises a sub-virtual node, a virtual computing node and a virtual transmission link, wherein the sub-virtual node maintains the communication and storage capacity of the original node, the virtual computing node provides the computing capacity of the original node, and the virtual transmission link connects the sub-virtual node and the virtual computing node. Meanwhile, the function expansion diagram of the invention can represent a plurality of parallel or a plurality of continuous computing functions in one node.
2) In the invention, the problem that the data processing process of the computing function unit of the node cannot be represented in the traditional time expansion diagram is solved by introducing the virtual transmission link in the construction of the function expansion diagram. The invention uses the transmission link, the storage link and the virtual transmission link in the function expansion diagram to represent communication, storage and calculation resources in the time-varying network, provides a uniform representation for the joint communication, storage and calculation resources, and the position relationship among different links in the function expansion diagram represents the connection and transformation relationship among different resources.
3) The invention sets communication capacity constraint, storage capacity constraint, calculation capacity constraint and flow conservation constraint on data flow in the function expansion diagram, overcomes the problem that the flow conservation constraint is difficult to quantify caused by the possible change of the type of the data flow and the proportion of the input data flow and the output data flow for a node with a plurality of parallel calculation functions or a plurality of continuous functions, and can uniformly represent dynamic network communication, storage and calculation resources changing along with time through the function expansion diagram.
Drawings
FIG. 1 is a schematic view of a scenario in which the present invention is used;
FIG. 2 is a general flow chart of an implementation of the present invention;
FIG. 3 is a schematic diagram illustrating the connection relationship between nodes in a network during a planning period according to the present invention;
FIG. 4 is a schematic diagram of different nodes and virtual transmission links obtained by decomposing functional nodes in the present invention;
FIG. 5 is a blank directed graph initialized in the present invention;
FIG. 6 is a functional node diagram constructed in the present invention.
Fig. 7 is a function expansion diagram constructed in the present invention.
Detailed Description
The invention is described in detail below with reference to the drawings and examples, which are only for the purpose of illustrating the invention and do not constitute any limitation to the invention.
Referring to fig. 1, the network scenario of the present example consists of 6 network nodesAre formed of network nodesAndthe three nodes can not provide task computing function and only play roles of communication and storage, butAndthe three network nodes can not only provide communication and storage functions, but also can respectively provide M1,M2And M3A computing function. The planning period of the network isConnectivity between nodes in a planning period is shown in fig. 3, and each abscissa and ordinate in fig. 3 corresponds to a connection relationship between a pair of nodes, where the abscissa represents time, the ordinate represents connectivity, state 1 represents connection, and state 0 represents disconnection.
Referring to fig. 2, the specific implementation steps of the present example under the above scenario conditions are as follows:
The number of the initialized network nodes is N, N is 6, and the initialized network nodes are aggregated intoI.e. the set is composed ofSix network nodes, wherein the network nodesAndthe three nodes can not provide task computing function and only play roles of communication and storage, butAndthe three network nodes can not only provide communication and storage functions, but also can respectively provide M1,M2And M3A computing function.
Dividing nodes which can not provide task computing function and only play communication and storage roles in the network into non-functional nodes, and dividing nodes which can not only provide communication and storage functions but also provide computing function in the network into functional nodes;
aggregating network nodes according to the partitioningExpressed as:whereinIs a collection of non-functional nodes that,in the form of a collection of functional nodes,represents the jth nonfunctional node, j ∈ [1, N)1],N 13 is the number of non-functional nodes,represents the ith functional node, i ∈ [1, N2],N 23 is the number of functional nodes, N is N1+N2。
And 2, decomposing the network nodes according to the functions of the network nodes.
If function nodeCan provide MiA computing function, then decomposing the node into a virtual child node viAnd MiA virtual computing nodeAnd two virtual transmission linksAndas shown in FIG. 4, wherein MiAs a functional nodeThe total number of computing functions can be provided,functional node of a representationThe mth virtual compute node of the decomposition,representing slave virtual child node viTo virtual computing nodeThe directional line segment of (a) is,representing slave virtual computing nodesTo virtual child node viIs directed line segment of (M ∈ [1, M)i]Virtual child node viVirtual computing node implementing communication and storage functionsImplementing a computing functionAnd data flowStreaming virtual compute nodesWill be processed and converted into a new type of data streamFrom the virtual computing nodeAnd the water flows out, wherein,to calculate the function for the upcoming receptionThe data stream of (a) is transmitted,computing functions for receivedThe data stream of (2).
And 3, dividing a network planning period into T continuous unequal time intervals according to the connectivity of the network nodes.
As shown in fig. 3, it comprises three time intervals, i.e. a first time interval τ1Second time interval τ2Third time interval τ3Wherein:
first time interval τ1Inner, 1 st non-functional nodeCan give the 1 st functional nodeTransmitting data, 1 st functional nodeCan give the 2 nd functional nodeTransmitting data, 2 nd non-functional nodeCan give the 3 rd non-functional nodeTransmitting data;
second time interval tau2Inner, 1 st non-functional nodeCan give the 1 st functional nodeTransmitting data, 2 ndFunction nodeCan give the 3 rd function nodeTransmitting data, 3 rd function nodeCan give the 2 nd non-functional nodeTransmitting data;
third time interval τ3Inner, 1 st function nodeCan give the 2 nd functional nodeTransmitting data, 2 nd functional nodeCan give the 3 rd function nodeTransmitting data, 3 rd function nodeCan give the 2 nd non-functional nodeTransmitting data, 2 nd non-functional nodeCan give the 3 rd non-functional nodeAnd transmitting the data.
Planning the network period according to the connectivity of the 6 network nodesDivided into 3 time intervals { tau1,τ2,τ3Where T is 3, τq=[tq-1,tq) And at a time interval tauqThe internal network topology is kept unchanged, and q belongs to [1, T ]]。
And 4, constructing a function expansion diagram.
4.1) initializing a blank T-3 layer directed graph, wherein the time interval of the q-th layer directed graph is tauqQ is more than or equal to 1 and less than or equal to 3, as shown in figure 5;
4.2) at each time interval τ of the directed graphqAll non-functional nodes, all virtual sub-nodes decomposed by functional nodes, and all virtual computing nodes decomposed by functional nodes in the network are added to form a functional node diagram, as shown in fig. 6, where:
set of non-functional nodes of a functional node map asI.e. the set is composed ofNine non-functional nodes, as shown by the pentagonal nodes of fig. 6, wherein,indicating the jth network non-functional nodeAt the qth time interval τqInner copy, j is more than or equal to 1 and less than or equal to N1,1≤q≤T,N1=3,T=3;
Set of virtual child nodes of functional node graph asI.e. the set is composed ofNine virtual child nodes, as shown by the circle node in fig. 6, wherein,representing the ith network virtual child node viAt the qth time interval τqInner copy, i is more than or equal to 1 and less than or equal to N2,1≤q≤T,N2=3,T=3;
The virtual computing node set of the functional node graph isThe set consists of virtual compute nodes in three time intervals, as shown by the rectangle of nodes in fig. 6, i.e.,
at a first time interval tau1Therein is provided with In total of M1+M2+M3A plurality of virtual compute nodes;
at a second time interval tau2Therein is provided with In total of M1+M2+M3A plurality of virtual compute nodes;
at a third time interval tau3Therein is provided with In total of M1+M2+M3A plurality of virtual compute nodes; wherein,representing the ith network virtual compute nodeAt the qth time interval τqInner copy, i is more than or equal to 1 and less than or equal to N2,1≤m≤Mi,1≤q≤T,N2=3,T=3;
4.3) adding links in the functional node graph, as shown in FIG. 7:
4.3.1) adding transmission links according to connectivity of the node, as shown by the solid line in fig. 7:
if at time interval τqThe jth non-functional node in the internal and external networkCan give the kth non-functional nodeTransmitting data, and then transmitting the data in the jth non-functional node in the functional node diagramAnd the kth non-functional nodeAdding a directed line segment between
If at time interval τqInner, j th non-functional nodeCan give the ith functional nodeTransmitting data, and then transmitting the data in the jth non-functional node in the functional node diagramAnd the ith virtual child nodeAdding a directed line segment between
If at time interval τqInner, i-th function nodeCan give the kth function nodeWhen data is transmitted, the ith virtual child node in the functional node diagramAnd the kth virtual child nodeAdding a directed line segment between
If at time interval τqInner, i-th function nodeCan give the jth non-functional nodeWhen data is transmitted, the ith virtual child node in the functional node diagramAnd the jth non-functional nodeAdding a directed line betweenSegment of
4.3.2) add storage links as shown by the dashed lines in FIG. 7:
adding a node from the q-th time interval between adjacent time intervals of each non-functional node of the functional node mapNode to the q +1 time intervalDirected line segment of
Adding a node from the q-th time interval between adjacent time intervals of each virtual child node of the functional node graphNode to qth time intervalDirected line segment of
4.3.3) adding virtual transmission links: at each virtual child node of the functional node mapVirtual sub-computing node corresponding theretoAdding two directed line segments in betweenAndas shown by the dotted line of fig. 7, the function expansion diagram shown in fig. 7 is obtained.
And 5, setting a communication capacity constraint, a storage capacity constraint, a calculation capacity constraint and a flow conservation constraint.
5.1) setting a communication capacity constraint, namely limiting that the sum of the data amount transmitted by all data streams on a transmission link or a virtual transmission link cannot exceed the communication capacity of the transmission link or the virtual transmission link:
5.1.1) for a transmission link, the sum of the data amounts transmitted on the transmission link by all data streams defined by the communication capacity constraint of the transmission link cannot exceed the communication capacity of the transmission link, and the formula is as follows:
wherein,representing a node from the jth non-functionTo the kth non-functional nodeIs transmitted over a networkThe number of types of data streams on the network,representing a node from the jth non-functionTo the ith virtual child nodeIs transmitted over a networkThe number of types of data streams on the network,representing the from the ith virtual child nodeTo the k-th virtual child nodeIs transmitted over a networkThe number of types of data streams on the network,representing the from the ith virtual child nodeTo the jth non-functional nodeIs transmitted over a networkThe number of types of data streams on the network, andrespectively representing data streams xinIn a transmission link Andthe amount of data to be transmitted over the network,andare respectively transmission linksAndthe communication capacity of the mobile communication terminal (c),a set of transmission links in the function expansion diagram;
5.1.2) for a virtual transmission link, the sum of the data amount transmitted on the virtual transmission link by the data flow defined in the restriction of the communication capacity of the virtual transmission link cannot exceed the communication capacity of the virtual transmission link, and the formula is as follows:
data stream xi of new type converted for received computing functionn' in a virtual transmission linkThe amount of data transmitted;
as a stream xinIn virtual transmission linksA transmission capacity of (a), which represents the maximum amount of data that can be transmitted;
data stream xi of new type converted for received computing functionn' in a virtual transmission linkThe transmission capacity of (a);
from virtual child nodes to virtual compute nodes in a function expansion diagramA set of virtual transmission links;
is a set of virtual transmission links from a virtual child compute node to a virtual child node in a functional expansion graph.
5.2) setting a storage capacity constraint, namely limiting the sum of the data quantity stored on the storage link of all data streams to be not more than the storage capacity of the storage link, wherein the formula is as follows:
wherein,andrespectively representing data streams xinIn a memory linkAndthe amount of data stored on the memory device,andrespectively representing memory linksAndthe storage capacity of (a) of (b),representing a flow through a transmission link and a storage link into a jth non-functional node in a functional expansion graphThe number of types of different data streams of (2),representing the flow through the transport link, storage link and virtual transport link into the ith virtual child node in the function expansion graphThe number of types of different data streams of (2),the set of links is stored for the function expansion map.
5.3) setting the computation capacity constraint, i.e. the defined function of the incoming computationOf a data streamStreaming virtual compute nodesThe consumed computing capacity cannot exceed the virtual computing nodeThe computational power provided is formulated as follows:
wherein,to calculate the function for the upcoming receptionThe data stream of (a) is transmitted,computing functions for receivedThe data stream of (a) is transmitted,as a stream of dataIn virtual transmission linksThe amount of data to be transmitted over the network,for calculating the factor, what is indicated is that the data stream per unit is processedAnd converted into a data streamThe amount of computing power that needs to be consumed,representing virtual computing nodesThe computing power of the device.
5.4) setting flow conservation constraint:
the constraint comprises three aspects of non-functional nodes, virtual child nodes and virtual computing nodes, and is specifically realized as follows:
5.4.1) for non-functional nodes, define data stream xinThe amount of data flowing into a non-functional node is equal to the amount of data it flows out of the non-functional node, and the formula is as follows:
wherein,andrespectively representing data streams xinIn a transmission linkAndthe amount of data to be transmitted over the network,andrespectively representing data streams xinIn a memory linkAndthe amount of data stored on the memory device,for streaming to non-functional nodes via transmission and storage linksDifferent data ofThe set of the streams is then selected,is a collection of transmission links in the function expansion diagram.
5.4.2) for a virtual child node, define a data stream ξnThe amount of data flowing into a virtual child is equal to the amount of data it flows out of the virtual child, and the formula is as follows:
wherein,andrespectively representing data streams xinIn a transmission linkAndthe amount of data to be transmitted over the network,andrespectively representing data streams xinIn virtual transmission linksAndthe amount of data to be transmitted over the network,andrespectively representing data streams xinIn a memory linkAndthe amount of data stored on the memory device,for streaming into the ith virtual sub-node over a transport link, a storage link and a virtual transport linkOf the different data streams of (a) to (b),a set of transmission links in the function expansion diagram;
a set of virtual transmission links from the virtual child nodes to the virtual compute nodes in the function expansion graph;
is a set of virtual transmission links from a virtual child compute node to a virtual child node in a functional expansion graph.
5.4.3) for virtual compute nodes, defining data flowsStreaming virtual compute nodesIs multiplied by the amount of data ofEqual to another type of data streamEgress virtual compute nodeThe formula is as follows:
wherein,computing functions for imminent receptionThe data stream of (a) is transmitted,computing functions for receivedThe data stream of (a);
to calculate the function for the upcoming receptionOf a data streamIn virtual transmission linksThe amount of data transmitted;
to be alreadyReceive computing functionalityOf a data streamIn a virtual transmission linkThe amount of data transmitted;
indicating that a computing function is about to be receivedData stream ofAnd received computing functionOf a data streamA scaling factor in between.
And 6, uniformly representing communication, storage and computing resources by using the function expansion diagram.
Because the transmission link, the storage link and the virtual transmission link in the function expansion diagram respectively represent communication, storage and calculation resources in the time-varying network, and the position relationship between different links in the function expansion diagram represents the bearing conversion relationship between different resources, and simultaneously, under the four constraints set in the step 5, the data stream in the function expansion diagram can meet the communication resource constraint, the storage resource constraint, the calculation constraint and the conversion relationship between various data streams in the network, so that the problem of joint management of the communication, storage and calculation resources can be converted into the problem of the data stream in the function expansion diagram, namely the problem of unified representation of dynamic network communication, storage and calculation resources changing along with time by using the function expansion diagram. The communication, storage and computing resources of the time-varying network can be uniformly analyzed and managed by using the function expansion diagram.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.
Claims (9)
1. A unified characterization method for communication, storage and computing resources based on a function expansion diagram is characterized by comprising the following steps:
(2) dividing network nodes, namely dividing nodes which can not provide a task computing function and only play a role in communication and storage in a network into non-functional nodes, and dividing nodes which can not only provide the communication and storage functions but also provide the computing function in the network into functional nodes; aggregating network nodes according to the partitioningExpressed as:whereinIs a collection of non-functional nodes that,in the form of a collection of functional nodes,denotes the jth non-functional node, N1The number of non-functional nodes is,denotes the ith functional node, N2Is the number of functional nodes, N ═ N1+N2;
(3) Decomposing the network nodes according to the functions of the network nodes:
if function nodeCan provide MiA computing function, then decomposing the node into a virtual child node viAnd MiA virtual computing nodeAnd two virtual transmission linksAndwherein M isiAs a functional nodeThe total number of computing functions can be provided,functional node of a representationThe mth virtual compute node of the decomposition,representing a slave virtual child node viTo virtual computing nodeThe directional line segment of (a) is,representing slave virtual computing nodesTo virtual child node viIs directed line segment of (M ∈ [1, M)i];
(4) Planning the network period according to the connectivity of the network nodeDivided into T time intervalsWherein tau isq=[tq-1,tq) And at a time interval tauqThe internal network topology is kept unchanged, and q belongs to [1, T ]];
(5) Constructing a function expansion diagram:
(5a) initializing a blank T-layer directed graph, wherein the time interval of the q-th layer directed graph is tauq;
(5b) At each time interval τ of the directed graphqAdding all non-functional nodes, virtual sub-nodes decomposed by all functional nodes and virtual computing nodes decomposed by all functional nodes in the network respectively to form a functional node graph and obtain three types of node sets of the functional node graph:
wherein,is a set of non-functional nodes of a functional node map,is a set of virtual child nodes of the functional node map,is a virtual compute node set of a functional node map,is expressed in time interval tauqNon-functional node of internal networkA copy of (a) is made of,for a virtual sub-node viA copy of (a) is made of,is expressed in time interval tauqIntra-network virtual compute nodeA copy of (1);
(5c) adding links in the functional node graph:
(5c1) adding a transmission link according to the connectivity of the node:
if at time interval τqThe jth non-functional node in the internal and external networkCan give the kth non-functional nodeTransmitting data, and then transmitting the data in the jth non-functional node in the functional node diagramAnd the kth non-functional nodeAdding a directed line segment between
If at time interval τqInner, j th non-functional nodeCan give the ith functional nodeTransmitting data, and then transmitting the data in the jth non-functional node in the functional node diagramAnd the ith virtual child nodeAdding a directed line segment between
If at time interval τqInner, i-th function nodeCan give the kth function nodeWhen data is transmitted, the ith virtual child node in the functional node diagramAnd the kth virtual child nodeAdding a directed line segment between
If at time interval τqInner, i-th function nodeCan give the jth non-functional nodeWhen data is transmitted, the ith virtual child node in the functional node diagramAnd the jth non-functional nodeAdding a directed line segment between
(5c2) Adding a storage link:
adding a slave node between adjacent time intervals of each non-functional node of the functional node mapTo the nodeDirected line segment of
Adding a slave node between adjacent time intervals of each virtual child node of the functional node graphTo the nodeDirected line segment of
(5c3) Adding a virtual transmission link: at each virtual child node of the functional node mapVirtual sub-computing node corresponding theretoAdding two directed line segments in betweenAndthus obtaining a function expansion diagram;
(6) setting communication capacity constraint, storage capacity constraint, calculation capacity constraint and flow conservation constraint:
the communication capacity constraint is to limit the sum of the data amount transmitted by all the data streams on the transmission link or the virtual transmission link not to exceed the communication capacity of the transmission link or the virtual transmission link;
the storage capacity constraint is to limit the sum of the data amount stored on the storage link of all the data streams to be not more than the storage capacity of the storage link;
the computing capacity constraint is to restrict the data flowStreaming virtual compute nodesThe consumed computing capacity cannot exceed the virtual computing nodeProvided computing power, whereinTo calculate the function for the upcoming receptionThe data stream of (a) is transmitted,m∈[1,Mi],showing the flow through the transport and storage links into virtual child nodes in a functional expansion graphThe number of types of different data streams;
the flow conservation constraint includes: defining the amount of data flowing into the non-functional node of each data stream to be equal to the amount of data flowing out of the non-functional node; defining the amount of data flowing into the virtual child node for each data stream to be equal to the amount of data flowing out of the virtual child node; constrained data streamStreaming virtual compute nodesIs multiplied by the amount of data ofEqualing data streamsEgress virtual compute nodeThe amount of data of (a), wherein,computing functions for receivedThe data stream of (2);
(7) under the four set constraints, the problem of joint management of communication, storage and computing resources is converted into a data flow problem in a function expansion diagram, namely the function expansion diagram is used for uniformly representing dynamic network communication, storage and computing resources changing along with time.
2. The method of claim 1, wherein the virtual child nodes and virtual compute nodes in (3), each implement different functions, namely:
virtual child node viCommunication and storage functions are realized;
virtual computing nodeImplementing a computing functionAnd data flowStreaming virtual compute nodesWill be processed and converted into a new type of data streamFrom the virtual computing nodeAnd the water flows out, wherein,to calculate the function for the upcoming receptionThe data stream of (a) is transmitted,to calculate functions for receivedThe data stream of (2).
3. The method of claim 1, wherein the sum of the data amounts transmitted on the transmission link for all the data streams defined by the communication capacity constraint in (6) cannot exceed the communication capacity of the transmission link, and the formula is as follows:
wherein,representing slave non-functional nodesTo non-functional nodesIs transmitted over a networkThe number of types of data streams on the network,representing slave non-functional nodesTo the virtual child nodeIs transmitted over a networkThe number of types of data streams on the network,representing slave virtual child nodesTo the virtual child nodeIs transmitted over a networkThe number of types of data streams on the network,representing slave virtual child nodesTo non-functional nodesIs transmitted toThe number of types of data streams on the network,andrespectively representing data streams xinIn a transmission link Andthe amount of data to be transmitted over the network,andare respectively transmission linksAndthe communication capacity of the mobile communication terminal (c),is a collection of transmission links in the function expansion diagram.
4. The method according to claim 1, wherein the sum of the data amount transmitted on the virtual transmission link for the data streams defined in the communication capacity constraint in (6) cannot exceed the communication capacity of the virtual transmission link, and the formula is as follows:
wherein,as a stream xinIn virtual transmission linksThe amount of data to be transmitted over the network,
data stream xi of new type converted for received computing functionn' in a virtual transmission linkThe amount of data to be transmitted over the network,
as a stream xinIn virtual transmission linksA transmission capacity of (a), which represents the maximum amount of data that can be transmitted;
data stream xi of new type converted for received computing functionn' in a virtual transmission linkThe transmission capacity of (a);
a set of virtual transmission links from the virtual child nodes to the virtual compute nodes in the function expansion graph;
5. The method of claim 1, wherein the sum of the amounts of data stored on the storage links for all data streams defined in (6) in the storage capacity constraint cannot exceed the storage capacity of its storage link, and the formula is expressed as follows:
wherein,andrespectively representing data streams xinIn a memory linkAndthe amount of data stored on the memory device,andrespectively representing memory linksAndthe storage capacity of (a) of (b),representing non-functional nodes flowing into a functional expansion graph over transport links and storage linksThe number of types of different data streams of (2),representing flow into a virtual child node in a function extension graph through transport links, storage links, and virtual transport linksThe number of types of different data streams of (2),the set of links is stored for the function expansion map.
6. The method of claim 1 wherein the receive-soon-to-compute function defined in (6) in the constraint on computing capacityOf a data streamStreaming virtual compute nodesThe consumed computing capacity cannot exceed the virtual computing nodeThe computational power provided, the formula is as follows:
wherein,to calculate the function for the upcoming receptionOf a data stream,Computing functions for receivedThe data stream of (a) is transmitted,as a stream of dataIn virtual transmission linksThe amount of data to be transmitted over the network,for calculating the factor, what is indicated is that the data stream per unit is processedAnd converted into a data streamThe amount of computing power that needs to be consumed,representing virtual computing nodesThe computing power of the device.
7. The method of claim 1, wherein the data stream ξ as defined in the conservation of flow constraint in (6)nThe amount of data flowing into a non-functional node is equal to the amount of data it flows out of the non-functional node, and the formula is as follows:
wherein,andrespectively representing data streams xinIn a transmission linkAndthe amount of data to be transmitted over the network,andrespectively representing data streams xinIn a memory linkAndthe amount of data stored on the memory device,for streaming to non-functional nodes via transmission and storage linksOf the different data streams of (a) to (b),is a collection of transmission links in the function expansion diagram.
8. The method of claim 1, wherein the data stream ξ as defined in the conservation of flow constraint in (6)nThe amount of data flowing into a virtual child is equal to the amount of data it flows out of the virtual child, and the formula is as follows:
wherein,andrespectively representing data streams xinIn a transmission linkAndthe amount of data to be transmitted over the network,andrespectively representing data streams xinIn virtual transmission linksAndthe amount of data to be transmitted over the network,andrespectively representing data streams xinIn a memory linkAndthe amount of data stored on the memory device,for streaming to virtual sub-nodes via transport links, storage links and virtual transport linksOf the different data streams of (a) to (b),a set of transmission links in the function expansion diagram;
a set of virtual transmission links from the virtual child nodes to the virtual compute nodes in the function expansion graph;
9. The method of claim 1The method of (1), wherein the data stream defined in the conservation of flow constraint in (6)Streaming virtual compute nodesIs multiplied by the amount of data ofEqual to another type of data streamEgress virtual compute nodeThe formula is as follows:
wherein,to calculate the function for the upcoming receptionOf a data streamIn virtual transmission linksThe amount of data transmitted;
for counting receivedComputing functionOf a data streamIn virtual transmission linksThe amount of data transmitted;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210087697.5A CN114500296B (en) | 2022-01-25 | 2022-01-25 | Communication, storage and computing resource unified characterization method based on function expansion diagram |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210087697.5A CN114500296B (en) | 2022-01-25 | 2022-01-25 | Communication, storage and computing resource unified characterization method based on function expansion diagram |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114500296A true CN114500296A (en) | 2022-05-13 |
CN114500296B CN114500296B (en) | 2023-02-07 |
Family
ID=81475438
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210087697.5A Active CN114500296B (en) | 2022-01-25 | 2022-01-25 | Communication, storage and computing resource unified characterization method based on function expansion diagram |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114500296B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114641029A (en) * | 2022-05-17 | 2022-06-17 | 亚太卫星宽带通信(深圳)有限公司 | Transmitting method and system based on satellite communication signal device |
CN115833899A (en) * | 2022-09-30 | 2023-03-21 | 西安电子科技大学 | Virtual network function deployment and routing joint optimization method in spatial information network |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160149993A1 (en) * | 2013-06-25 | 2016-05-26 | Orange | Cooperative network architecture |
CN106877921A (en) * | 2017-03-14 | 2017-06-20 | 西安电子科技大学 | Information Network multi dimensional resource integration characterizing method |
CN108718251A (en) * | 2018-05-10 | 2018-10-30 | 西安电子科技大学 | Information Network connectivity analysis methods based on resource time-varying figure |
CN110636002A (en) * | 2019-10-15 | 2019-12-31 | 西安电子科技大学 | Time division based maximum flow routing method for limited space information network of transceiver |
US20210194907A1 (en) * | 2019-12-20 | 2021-06-24 | Microsoft Technology Licensing, Llc | Detecting anomalous network activity |
CN113365355A (en) * | 2021-04-29 | 2021-09-07 | 西安交通大学 | Multidimensional resource management method and system under air-space-ground integrated network |
CN113438156A (en) * | 2021-06-29 | 2021-09-24 | 西安电子科技大学 | Time deterministic multi-path routing method based on time expansion diagram |
US20210382754A1 (en) * | 2021-06-12 | 2021-12-09 | Intel Corporation | Serverless computing architecture for artificial intelligence workloads on edge for dynamic reconfiguration of workloads and enhanced resource utilization |
-
2022
- 2022-01-25 CN CN202210087697.5A patent/CN114500296B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160149993A1 (en) * | 2013-06-25 | 2016-05-26 | Orange | Cooperative network architecture |
CN106877921A (en) * | 2017-03-14 | 2017-06-20 | 西安电子科技大学 | Information Network multi dimensional resource integration characterizing method |
CN108718251A (en) * | 2018-05-10 | 2018-10-30 | 西安电子科技大学 | Information Network connectivity analysis methods based on resource time-varying figure |
CN110636002A (en) * | 2019-10-15 | 2019-12-31 | 西安电子科技大学 | Time division based maximum flow routing method for limited space information network of transceiver |
US20210194907A1 (en) * | 2019-12-20 | 2021-06-24 | Microsoft Technology Licensing, Llc | Detecting anomalous network activity |
CN113365355A (en) * | 2021-04-29 | 2021-09-07 | 西安交通大学 | Multidimensional resource management method and system under air-space-ground integrated network |
US20210382754A1 (en) * | 2021-06-12 | 2021-12-09 | Intel Corporation | Serverless computing architecture for artificial intelligence workloads on edge for dynamic reconfiguration of workloads and enhanced resource utilization |
CN113438156A (en) * | 2021-06-29 | 2021-09-24 | 西安电子科技大学 | Time deterministic multi-path routing method based on time expansion diagram |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114641029A (en) * | 2022-05-17 | 2022-06-17 | 亚太卫星宽带通信(深圳)有限公司 | Transmitting method and system based on satellite communication signal device |
CN114641029B (en) * | 2022-05-17 | 2022-08-02 | 亚太卫星宽带通信(深圳)有限公司 | Transmitting method and system based on satellite communication signal device |
CN115833899A (en) * | 2022-09-30 | 2023-03-21 | 西安电子科技大学 | Virtual network function deployment and routing joint optimization method in spatial information network |
Also Published As
Publication number | Publication date |
---|---|
CN114500296B (en) | 2023-02-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN114500296B (en) | Communication, storage and computing resource unified characterization method based on function expansion diagram | |
CN106934497B (en) | Intelligent community power consumption real-time prediction method and device based on deep learning | |
Leinonen et al. | Distributed joint resource and routing optimization in wireless sensor networks via alternating direction method of multipliers | |
CN109359385A (en) | A kind of training method and device of service quality assessment model | |
WO2023179010A1 (en) | User packet and resource allocation method and apparatus in noma-mec system | |
CN103596191B (en) | A kind of wireless sensor network intelligent configuration system and method thereof | |
CN102546220B (en) | Key quality indicator (KQI) composition method based on service characteristics | |
CN112650581A (en) | Cloud-side cooperative task scheduling method for intelligent building | |
CN110928694B (en) | Computer system | |
CN113518007B (en) | Multi-internet-of-things equipment heterogeneous model efficient mutual learning method based on federal learning | |
CN106304191A (en) | A kind of data receiver method based on cluster structured radio sensor network and device | |
CN115665227B (en) | Universal heterogeneous integrated computing network resource intelligent adaptation network architecture and method | |
CN114745383A (en) | Mobile edge calculation assisted multilayer federal learning method | |
CN107454009B (en) | Data center-oriented offline scene low-bandwidth overhead traffic scheduling scheme | |
CN115392481A (en) | Federal learning efficient communication method based on real-time response time balancing | |
CN105491329A (en) | Large-scale monitoring video stream converging method based on stream-oriented computation | |
CN108376099B (en) | Mobile terminal calculation migration method for optimizing time delay and energy efficiency | |
CN102946443B (en) | Multitask scheduling method for realizing large-scale data transmission | |
CN114357676A (en) | Aggregation frequency control method for hierarchical model training framework | |
CN112506656B (en) | Distribution method based on calculation tasks of power distribution Internet of things | |
CN107528731B (en) | Network segmentation optimization algorithm applied to NS3 parallel simulation | |
Zhou et al. | TSEngine: Enable efficient communication overlay in distributed machine learning in WANs | |
CN104168603B (en) | The building method of the data acquisition tree with high energy efficiency based on compressed sensing technology | |
CN104065580A (en) | M2M network topology control method and system | |
CN107820293A (en) | A kind of wireless relay nodes system of selection, system, equipment and computer media |
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 |