CN108845886A - Cloud computing energy consumption optimization method and system based on phase space - Google Patents
Cloud computing energy consumption optimization method and system based on phase space Download PDFInfo
- Publication number
- CN108845886A CN108845886A CN201810845198.1A CN201810845198A CN108845886A CN 108845886 A CN108845886 A CN 108845886A CN 201810845198 A CN201810845198 A CN 201810845198A CN 108845886 A CN108845886 A CN 108845886A
- Authority
- CN
- China
- Prior art keywords
- energy consumption
- phase space
- task
- cloud computing
- node
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/5044—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering hardware capabilities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
- G06F1/3234—Power saving characterised by the action undertaken
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- 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
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to field of cloud calculation, disclose a kind of cloud computing energy consumption optimization method and system based on phase space, to optimize the energy consumption in cloud computing, reduce task and distribute the time, improve task allocative efficiency；The method of the present invention includes the context informations for obtaining node and node all in cloud computing system, and establish energy consumption model；The total energy consumption of all being carrying out for tasks of measuring node establishes static energy consumption phase space according to the total energy consumption；The energy consumption of newly-increased required by task is predicted using energy consumption model, and establishes dynamic energy consumption phase space；Optimal phase subspace is calculated according to static energy consumption phase space；The energy consumption for predicting newly-increased task each node in optimal phase subspace according to static energy consumption phase space and dynamic energy consumption phase space design objective dispatching algorithm, and determines the optimal scheduling scheme of newly-increased task.
Description
Technical field
The present invention relates to field of cloud calculation, more particularly to cloud computing energy consumption optimization method and system based on phase space.
Background technique
The appearance of cloud computing allow enterprise obtain online it is a large amount of calculate and storage resource, only need to be by using resource
Duration is paid, and the investment to expensive IT infrastructure is avoided.With the fast development of cloud computing technology, global cloud computing
The quantity of data center is being continuously increased, and the ratio that the energy that data center consumes every year accounts for worldwide energy consumption increases year by year
It is long.While cloud computation data center consumes mass energy every year, also causing certain pressure to environment, (such as data center is generated
Greenhouse gases), cloud computing energy consumption problem be increasingly becoming researcher concern emphasis.Cloud computation data center is an association
The entirety for adjusting work, there are a large amount of computation migration, storages to migrate contour coupling operation in cloud computing system, so cloud computing system
System is a kind of high coupled system with magnanimity node.To study cloud computing energy consumption problem in system level, it is necessary to will
All nodes in cloud computing system are treated and are analyzed as a whole.
Currently, the energy optimization management of cloud computing is one of the important content of energy consumption correlative study, mainly around dynamic electric
Voltage-frequency rate adjustment technology, closing dormant technology, several big main flow directions of virtualization technology conduct a research.Accurate energy consumption measurement and pre-
The basis that model is energy optimization management is surveyed, current existing energy consumption assessment model includes the energy consumption of system main component utilization rate
Model and energy consumption prediction model based on performance counter.To the energy consumption evaluation method based on system utilization rate, the research of early stage
It is the energy consumption model based on processor utilization mostly, it is believed that CPU is consumed energy in Cloud Server (cloud computing system) working node
The highest component of ratio, and the occupancy of other assemblies improves and frequently can lead to the soaring of cpu busy percentage.Therefore with CPU benefit
With rate come evaluation system load, further estimating system energy consumption is feasible.The frequency and use that some researchs pass through concern CPU
Rate, and the power consumption state of server other assemblies is set as constant to estimate server node energy consumption.Based on performance counter
(PMC) energy consumption model, the performance monitoring system mainly provided by major hardware vendor, such as to instructed in processor, it is slow
Deposit, the monitoring of page table cache, then filter out important impact factor and establish model such methods measurement expense it is extremely low and
As a result relatively accurate, the truthful data of " bottom " can be obtained and also energy consumption can be assessed on the whole, but this side
The deployment difficulty of method is relatively large.Above method has respective applicable scene, can provide more accurate energy consumption prediction and degree
Amount.But the above energy consumption model estimates the energy consumption of High Performance Computing Center more concerned with the power consumption state of individual server node
Also only simply the energy consumption of server node is added up.Therefore the above method is more suitably applied to the energy consumption between server
State couples unconspicuous situation, is not suitable for the cloud computing system of high coupling.
Summary of the invention
It is an object of that present invention to provide a kind of cloud computing energy consumption optimization method and system based on phase space, to optimize in terms of cloud
Energy consumption in calculation reduces task and distributes the time, improves task allocative efficiency.
To achieve the above object, the cloud computing energy consumption optimization method based on phase space that the present invention provides a kind of, including with
Lower step：
S1：The context information of node and the node all in cloud computing system is obtained, and establishes energy consumption
Model；
S2：The total energy consumption for measuring all being carrying out for tasks of the node establishes static energy consumption phase according to the total energy consumption
Space；
S3：The energy consumption of newly-increased required by task is predicted using the energy consumption model, and establishes dynamic energy consumption phase space；
S4：Optimal phase subspace is calculated according to the static energy consumption phase space；
S5：The energy consumption for predicting newly-increased task each node in the optimal phase subspace, it is mutually empty according to the static energy consumption
Between and the dynamic energy consumption phase space design objective dispatching algorithm, and determine the optimal scheduling scheme of newly-increased task.
Preferably, in the S2, the static energy consumption phase space of the foundation specifically includes following steps：
S21：Obtain the institute that all nodes current in the cloud computing system are carrying out the energy consumption parameter composition of task
Directed quantity；
S22：By all DUAL PROBLEMS OF VECTOR MAPPINGs into phase space, the number of the energy consumption parameter is considered as to the dimension of phase space
The sum of all described vector field homoemorphisms are considered as the current load total energy consumption of cloud computing system, are built according to the load total energy consumption by degree
Vertical static state energy consumption phase space.
Preferably, in the S3, the dynamic energy consumption phase space of establishing specifically includes following steps：
S31：All energy of the energy consumption parameter composition of the newly-increased task in the cloud computing system are predicted using energy consumption model
Consumption increases vector；
S32：The energy consumption is increased into DUAL PROBLEMS OF VECTOR MAPPING into phase space, the number of parameters of the energy consumption of the newly-increased task is regarded
For the dimension of phase space, all energy consumptions are increased into the sum of vector field homoemorphism and are considered as newly-increased total energy consumption, according to described newly-increased
Total energy consumption establishes dynamic energy consumption phase space.
Preferably, the S4 specifically includes following steps：
S41：The reference axis for reacting the node energy consumption state in static energy consumption phase space is established, and determines the static energy
Consume the projection point set of phase space；
S42：In the static energy consumption phase space, the center of gravity of the excessively described projection point set is obtained to the reference axis as vertical line
To the subitem space of respective numbers；
S43：If coordinate value of any point in each reference axis is respectively less than the coordinate value at center in some subitem space,
The subitem space is considered as optimal subitem space.
Preferably, in the S5, the optimal scheduling scheme that the determination increases task newly specifically includes following steps：
S51：Newly-increased energy consumption of the task in the optimal subitem space on each node is predicted using the energy consumption model, and
Obtain all possible allocation plan；
S52：The all possible allocation plan is mapped to by the dynamic energy consumption phase space using task scheduling algorithm
In, the optimal task assignment that phase space center of gravity will be made to be considered as compared to the smallest scheme of origin moving distance the newly-increased task
Scheme.
Preferably, in the S1, the context information includes node type, network condition and cluster position.
Preferably, the vector includes CPU energy consumption, energy consumption of memory and disk energy consumption.
Preferably, the increase vector includes that CPU energy consumption value added, energy consumption of memory value added and disk energy consumption increase
Value.
As a general technical idea, the cloud computing energy optimization system based on phase space that the present invention also provides a kind of,
Including memory, processor and store the computer program that can be run on a memory and on a processor, the processor
The step of realizing the above method when executing the computer program.
The invention has the advantages that：
The present invention provides a kind of cloud computing energy consumption optimization method and system based on phase space, first acquisition cloud computing system
In all node and node context information, and establish energy consumption model；And measuring node is all is carrying out
The total energy consumption of task establishes static energy consumption phase space according to the total energy consumption；Then newly-increased required by task is predicted using energy consumption model
Energy consumption, and establish dynamic energy consumption phase space；Optimal phase subspace is calculated further according to static energy consumption phase space and predicts new
The energy consumption of increasing task each node in optimal phase subspace, according to static energy consumption phase space and dynamic energy consumption phase space design objective
Dispatching algorithm, and determine the optimal scheduling scheme of newly-increased task；It considers upper and lower between each node in cloud computing system
Literary environmental factor can make the prediction to energy consumption more accurate, to optimize the energy consumption in cloud computing, reduce task and distribute the time, mention
High task allocative efficiency.
Below with reference to accompanying drawings, the present invention is described in further detail.
Detailed description of the invention
The attached drawing constituted part of this application is used to provide further understanding of the present invention, schematic reality of the invention
It applies example and its explanation is used to explain the present invention, do not constitute improper limitations of the present invention.In the accompanying drawings：
Fig. 1 is the cloud computing energy consumption optimization method flow chart based on phase space of the preferred embodiment of the present invention；
Fig. 2 is the effect picture of the cloud computing static state energy consumption phase space of the preferred embodiment of the present invention；
Fig. 3 is the effect picture of the cloud computing dynamic energy consumption phase space of the preferred embodiment of the present invention；
Fig. 4 is the threshold value facilities figure in the optimal subitem space of the preferred embodiment of the present invention；
Fig. 5 is that the distribution of four kinds of dispatching algorithm subpoints in dynamic energy consumption phase space of the preferred embodiment of the present invention is special
Sign；
Fig. 6 is the performance comparable situation schematic diagram of four based on the phase space kind algorithm of the preferred embodiment of the present invention；
Fig. 7 is the subpoint center of gravity that is in static phase space of the result of four kinds of algorithms of the preferred embodiment of the present invention to former
The distance versus situation schematic diagram of point.
Specific embodiment
The embodiment of the present invention is described in detail below in conjunction with attached drawing, but the present invention can be defined by the claims
Implement with the multitude of different ways of covering.
Embodiment 1
Referring to Fig. 1, the cloud computing energy consumption optimization method based on phase space that the present embodiment provides a kind of includes the following steps：
S1：The context information of node and node all in cloud computing system is obtained, and establishes energy consumption model；
S2：The total energy consumption of all being carrying out for tasks of measuring node establishes static energy consumption phase space according to the total energy consumption；
S3：The energy consumption of newly-increased required by task is predicted using energy consumption model, and establishes dynamic energy consumption phase space；
S4：Optimal phase subspace is calculated according to static energy consumption phase space；
S5：The energy consumption for predicting newly-increased task each node in optimal phase subspace, according to static energy consumption phase space and dynamic
Energy consumption phase space design objective dispatching algorithm, and determine the optimal scheduling scheme of newly-increased task.
In above step, it is contemplated that the context environmental factor in cloud computing system between each node establishes static energy
Phase space and dynamic energy consumption phase space are consumed, the prediction to energy consumption can be made more accurate, optimizes the energy consumption in cloud computing, reduces task
The time is distributed, task allocative efficiency is improved.
Optimized as follows in practical application, the embodiment of the present invention can also increase step：
S1：The context information of node and node all in cloud computing system is obtained, and establishes energy consumption model.
Specifically, it is able to achieve by optimizing task allocation plan to cloud computing system progress energy optimization, therefore, it is necessary to pre-
Survey the energy consumption that newly-increased task executes on the node that may receive the task.If target cloud computing system includes M node, each
Node has N number of component, CiIndicate i-th of node, CijIndicate j-th of component of i-th of node.Remember that component real time expense is Respectively indicate reality of the current task on processor, memory, disk
Time overhead.Using the history average power consumption of node component come each component energy consumption of calculate node, calculate as follows：
Wherein, PijIndicate component CijHistory average power consumption,Expression task is in component CijOn real time expense.
By real time expense and power decision, calculation formula is node energy consumption：
Wherein,For node CiActual consumption, obtained by the power consumption values of the N number of component of the node are cumulative,For node Ci
J-th of component CijActual consumption.
Since the ideal time expense of task is to obtain after excluding idle waiting situation as far as possible, and the real time of task opens
Pin can be greater than ideal time expense, therefore the energy consumption obtained using ideal time overhead computational is less than actual execution time expense and obtained
Energy consumption.Remember ideal time expense of the current task on each component
Respectively indicate ideal time expense of the current task on processor, memory, disk.
In large-scale cloud cluster, the whole energy consumption of system will be closely related with context environmental locating for clustered node, should
Context environmental includes cluster position, network condition, system architecture, node type etc..Such as component Configuration identical two
A large-scale cluster, the difference of the context environmental as locating for it will also result in certain journey when system task request is consistent
The energy consumption difference of degree.This context environmental causes consistency to influence cluster energy consumption within a certain period of time, i.e., for some
Node, context environmental parameter reflect integrality locating for system, remain unchanged within certain time.
It should be noted that the energy in optimization cloud computing system is time-consuming, context environmental locating for node is fully considered
Calculated result can be made more accurate.For a node, current context environmental parameter is θ=Tm/Tr.Further, it uses
Weighted mean method in time series analysis predicts the upper of present node according to the preceding K of node batches of history measured time expenses
Hereafter environmental parameter calculates as follows：
Wherein, θK+1Indicate the current discreet value of context environmental parameter, θiIndicate that i-th group of history from the distant to the near surveys number
According to environmental parameter, wiIndicate the weight of i-th group of history measured data from the distant to the near, the closer historical data away from current time
Weight is bigger,
In the situation known to the ideal time expense of current task request, need to go to obtain task by some means
Real time expense.In conjunction with above-mentioned context environmental parameter, the real time expense of calculating task is as follows：
Expression task is in component CijOn ideal time expense, θiFor the environmental parameter of node, newly arriving for task exists
When running on the node, the environmental parameter feature of node context is also obeyed.Thus it is possible to establish energy consumption model and be：
S2：The total energy consumption for measuring all being carrying out for tasks of the node establishes static energy consumption phase according to the total energy consumption
Space.It specifically includes：
S21：The institute for obtaining the energy consumption parameter composition that all nodes current in cloud computing system are carrying out task is oriented
Amount；
S22：By all DUAL PROBLEMS OF VECTOR MAPPINGs into phase space, the number of energy consumption parameter is considered as to the dimension of phase space, will be owned
The sum of vector field homoemorphism is considered as the current load total energy consumption of cloud computing system, establishes static energy consumption phase space according to load total energy consumption.
Specifically, if being currently included N number of real time energy consumption in cloud computing system, and by cloud computing system it is current it is N number of in real time
For the DUAL PROBLEMS OF VECTOR MAPPING of energy consumption parameter composition into phase space, the vector at this includes CPU energy consumption, energy consumption of memory and disk energy consumption
Deng the dimension of phase space being considered as real time energy consumption number of parameters, vector field homoemorphism indicates the current load total energy consumption of node, claims in this way
Phase space be cloud computing static state energy consumption phase space.
Relationship between node and vector in cloud computing system in order to facilitate understanding, it should be noted that assuming that cloud meter
Calculation system includes n node, then should form n vector, and i-th of vector field homoemorphism indicates the total energy consumption of i-th of node, and institute is oriented
The present load total energy consumption of the sum of mould of amount expression cloud computing system.The case where following dynamic energy consumption phase space, is similar to this,
All energy consumptions, which increase the sum of vector field homoemorphism, indicates that the increased total energy consumption of system, the mould of single vector only indicate the energy consumption of individual node
Value added.
It is all on the foundation consideration cloud computing system node of the static energy consumption phase space of cloud computing system to be carrying out task
Total energy consumption.Node CiPosition coordinates A in static energy consumption phase spaceiIt is expressed as follows：
Ai=(Xi1,Xi2,…Xij…XiN),1≤j≤N；
Wherein,
In formula, XijIndicate component CijThe normalized value of upper current total energy consumption,Indicate node CiUpper current general assignment load
In component CijThe power consumption values of upper generation.λiIndicate node CiEnergy consumption accounts for the percentage of cloud cluster total energy consumption, i.e., It indicates
CiHistory average energy consumption；Indicate the history average energy consumption of cloud data center.wijIndicate component CijEnergy consumption accounts for node CiTotal energy
The percentage of consumption, i.e., Indicate CijHistory average energy consumption.
It should be noted that by node CiAfter the energy consumption under present load is normalized to the generalized coordinates in phase space,
Just unique subpoint A has been determined in phase spacei, the current power consumption state of cloud computing system interior joint can thus be projected
Into energy consumption phase space, the position at static energy consumption phase space midpoint indicates the whole power consumption state of node.Cloud computing static energy
Consume the effect of phase space as shown in Fig. 2, it is worth noting that, it is mutually empty to only considered CPU, memory, 3 dimensions of disk in the present embodiment
Between, multidimensional situation is similar, and this will not be repeated here.
Node and group are already have accounted for when establishing mapping of the cloud computing system interior joint power consumption state to energy consumption phase space
The weight of part, the center of gravity that power consumption state projects point set in energy consumption phase space is G (X1,X2,…Xj…XN), in phase
Position in space indicates the current whole power consumption state of cloud computing system, wherein XjIt calculates as follows：
In formula, 1≤j≤N.
S3：The energy consumption of newly-increased required by task is predicted using energy consumption model, and establishes dynamic energy consumption phase space.It specifically includes：
S31：All energy consumptions using the energy consumption parameter composition of the newly-increased task in energy consumption model prediction cloud computing system increase
Add vector；
S32：Energy consumption is increased into DUAL PROBLEMS OF VECTOR MAPPING into phase space, the number of parameters of the energy consumption of newly-increased task is considered as phase space
Dimension, all energy consumptions are increased into the sum of vector field homoemorphism and are considered as newly-increased total energy consumption, dynamic energy is established according to newly-increased total energy consumption
Consume phase space.
Then, it obtains in cloud computing system and vector is increased by the energy consumption that N number of energy consumption parameter forms under newly-increased task, and will
The increase DUAL PROBLEMS OF VECTOR MAPPING into phase space, the increase vector at this include CPU energy consumption value added, energy consumption of memory value added and
Phase space dimension is considered as number of parameters by disk energy consumption value added etc., and when being mapped in phase space, energy consumption increases the dimension of vector
The as dimension of phase space, vector field homoemorphism represents the energy consumption variation degree of node in phase space, and such phase space is referred to as cloud meter
Dynamic energy consumption phase space is calculated, as shown in Figure 3.
It is worth noting that the foundation of cloud computing system dynamic energy consumption phase space considers that node receives the energy consumption that task generates
Increase situation.By node CiThe normalized value of bring energy consumption value added is as C due to receiving taskiIn dynamic energy consumption phase space
The generalized coordinates of middle subpoint, node CiEnergy consumption changing value in dynamic energy consumption phase space generalized coordinates it is as follows：
A′i=(X 'i1,X′i2…X′ij…X′iN),1≤j≤N；
Wherein, X 'ijIndicate component CijThe normalized value of upper current energy consumption variation,Indicate node CiReceive task load
Component C afterwardsijEnergy consumption changing value.
Similarly, the center of gravity that point set is projected in dynamic energy consumption phase space is G ' (X '1,X′1,…X′j…X′N), in phase
Position indicates the increase situation of energy consumption after cloud computing system reception task in space, wherein X 'jIt calculates as follows：
In formula, 1≤j≤N.
In the present embodiment, establishes static energy consumption phase space and dynamic energy consumption phase space respectively using the above method, pass through
Phase space describes the whole power consumption state in cloud computing system, easy to use, and can assist realizing to energy in cloud computing system
The Accurate Analysis of consumption state.
As those skilled in the art, it should be apparent that, in static energy consumption phase space, when task execution is complete on node
When finishing release resource, subpoint is reduced at a distance from origin towards phase space origin direction motion projection point, i.e. cloud computing system
The total energy consumption of upper current task load reduces；When the node in cloud computing system receives task, the energy consumption that node generates increases,
Subpoint is that subpoint increases at a distance from origin away from the direction movement of phase space origin, indicates cloud cluster present load
The total energy consumption of generation increases.Equally, subpoint corresponding node reception at a distance from initial point distance is negative in dynamic energy consumption phase space
The value added of energy consumption after load, subpoint is bigger from initial point distance, and the energy consumption that the received task load of node generates on node is got over
Greatly.
It should be noted that move process unrelated for the state change at either statically or dynamically energy consumption phase space midpoint, only
It is related with external load requests.And in thermodynamic system, the variable of state behavior is only related with state always, with approach without
It closes.Cloud computing system power consumption state is put into thermodynamic system research model, state change feature and thermodynamic system state
Variation characteristic is consistent, only related with external request therefore work as cloud computing system so the state change unrelated procedures at phase space midpoint
Power consumption state when changing, receive load and release load often exist simultaneously, if the load energy of system external world transmitting
It is D that consumption state, which increases, and system energy consumption state changes to S2 from S1, and the load power consumption state of system release at the same time is reduced to R, that
D=S2-S1+R.
When the node release or increase load in cloud computing system, subpoint position is in static energy consumption phase space with such as
Under type variation：
If original projection point coordinate of the node of cloud computing system in energy consumption phase space is C_node (x, y, z), node
Increased load indicates T_add=(a, b, c) with a three-dimensional vector, receives after load subpoint in static energy consumption phase space
Coordinate is：
In formula,Respectively indicate receive the task node energy consumption changing value be mapped to dynamic energy consumption
The generalized coordinates value of each dimension after in phase space.
S4：Optimal phase subspace is calculated according to static energy consumption phase space.It specifically includes：
S41：The reference axis for reacting node energy consumption state in static energy consumption phase space is established, and determines static energy consumption phase space
Projection point set；
S42：In static energy consumption phase space, the center of gravity for crossing projection point set obtains respective numbers as vertical line to reference axis
Subitem space；
S43：If coordinate value of any point in each reference axis is respectively less than the coordinate value at center in some subitem space,
The subitem space is considered as optimal subitem space.
In practical applications, in static energy consumption phase space, the center of gravity G for crossing projection point set makees vertical line to N number of reference axis,
N-dimensional phase space is resolved into 2 by these vertical linesNA phase subspace.Coordinate value of any point in each reference axis in phase subspace
When coordinate value respectively less than in the reference axis of center of gravity G, such phase subspace is referred to as optimal phase subspace.
Node in optimal phase subspace is roughly divided into two classes, and one kind is that node efficiency itself is preferable, this kind of node
In the task of execution, than other nodes, low energy consumption, therefore can pay the utmost attention to this kind of nodes in task distribution；Second class
Node is to cause energy consumption lower since resource utilization is lower, and existing research shows cloud computing system electric energy when utilization rate is very low
Consumption is more than the peak value energy consumption of the node, therefore just needs under the premise of not closing this kind of node to improve efficiency sometimes
The utilization rate for improving this kind of node, energy consumption can be obviously increased by assigning the task to this kind of node, be reduced entire when execution task
The energy consumption of system.Under the purpose of energy optimization, whole Energy Efficiency Ratio is in it after the node reception task in optimal phase subspace
Efficiency is more excellent after the node reception task of his phase subspace.
Node CiGeneralized coordinates A in static energy consumption phase spaceiWhether meet following formula judges subpoint in optimal son
In phase space.
Such as Fig. 2, in cloud computing three-dimensional static energy consumption phase space, center of gravity, which makees vertical line to reference axis, excessively can divide phase space
It is segmented into P1~P8 totally 8 phase subspaces.The coordinate value of each dimension in arbitrary point is respectively less than the coordinate value of center of gravity in phase subspace P1, presses
According to the definition of optimal phase subspace, P1 is optimal phase subspace in the cloud computing three-dimensional static energy consumption phase space of Fig. 2.It is preferred that
Ground can make X to simplify the searching of optimal phase subspace1….XNThe unified value equal to threshold value a S, S is according to the actual situation
Setting.
S5：The energy consumption for predicting newly-increased task each node in optimal phase subspace, according to static energy consumption phase space and dynamic
Energy consumption phase space design objective dispatching algorithm, and determine the optimal scheduling scheme of newly-increased task.It specifically includes：
S51：Newly-increased energy consumption of the task in optimal subitem space on each node is predicted using energy consumption model, and is owned
Possible allocation plan；
S52：All possible allocation plan is mapped in dynamic energy consumption phase space using task scheduling algorithm, will so that
Phase space center of gravity is considered as the optimal scheduling scheme of newly-increased task compared to the smallest scheme of origin moving distance.
In practical applications, the weight of first task scheduling algorithm, most of method for allocating tasks are all using random
Then iteration finds the smallest total energy consumption to selection task, can increase the complexity of algorithm in this way.The present embodiment is appointed by calculating
The dispatching priority weight rank value of business, wherein the average value of parameter in phase space determines rank value in each subpoint by task.
Carrying out task schedule according to the non-decreasing of rank value sequence is that can guarantee that task schedule meets priority scheduling energy consumption lesser
Business, then selects optimal node to be scheduled according to dispatching sequence.When not using optimal phase subspace, node to be matched
Number is M, therefore the time complexity of algorithm is O (N*M), is considering preferentially to use subpoint in optimal phase subspace corresponding
When node is matched, number of nodes to be matched is less than M, and therefore time complexity is less than O (N*M).
Therefore, when there is newly-increased task to reach in cloud computing system, task scheduling algorithm is in the target based on least energy consumption
In the case of task is quickly distributed.Wherein, which can be single, or batch, it should be noted that
The method of how many pairs of the present embodiment of newly-increased task does not impact, and therefore, does not do specific point to newly-increased task quantity
Analysis.Power consumption model predicts energy consumption of the task in optimal phase subspace on each node, then will be various by task scheduling algorithm
Possible allocation plan is mapped in dynamic energy consumption phase space, is found so that phase space gravity motion is carried out apart from the smallest scheme
The task of this round is distributed.Herein, phase space gravity motion distance minimum specifically, in dynamic power consumption phase space, compare by center of gravity
It is minimum in origin moving distance.
Experimental verification：
In a practical situation, it to be predicted by the historical energy consumption data of node when next subtask arrives in the node
Energy consumption is executed, needs to know the parameter of node, in this experiment, the parameter of node is as shown in table 1, and component mean power is divided by size
For 1~5 totally 5 ranks.
1 node parameter of table
Since the present embodiment is directed to the application of mixed type, therefore multiple practical applications are selected to represent various inhomogeneities
The application of type, including meteorological pre-, image procossing, molecule modeling, data visualization and mail service.It is all in this experiment to use
Using as shown in table 2 below.
By the application of five seed types more than upper execution, the data that monitoring is obtained are as the emulation data set of system.
The application of mixed load type is generated using them, and it is divided into 3 to each component resource requirement degree according to the characteristics of application
A rank：It is basic, normal, high.It is indicated respectively with 1,2,3 in table 2.
2 multiple types application parameter of table
This experiment is assessed the cloud computing phase space in extensive cloud computation data center by establishing analogue system and is dispatched
The performance of algorithm, in the analogue system, by above-mentioned authentic task and the simulation of parameter situation is gone out on missions and cloud computing system, when
When batch artificial tasks reach analogue system, analogue system generates total after the execution energy consumption using energy consumption prediction model prediction task
The smallest task allocation plan of energy consumption carries out epicycle task schedule according to the scheme generated.Task requests are formed using timeslice
Vector simulate, timeslice representative simulation task occupies the unit time length of node component, such as cpu request time, memory
Request time, disk requests time are an artificial tasks.Task is every after being assigned to pass through a unit time timeslice just
Reduce by a unit, all 0 when then indicates to be released after task execution.The simulation parameter of node is made of multiple parameters
Vector, such as (node environment parameter θ, the rated disspation P of energy consumption component, the energy consumption weight λ of node and the weights omega of component) most
The performance of phase space energy optimization frame is assessed according to the data that statistics obtains afterwards.
In this experiment, it is Inter (R) Core (TM) i5-6500 that the physical platform for carrying out emulation experiment, which is processor,
3.20GHZ；4GB RAM；The DELL host of 64 bit manipulation system of Windows7.Each node is according to above-mentioned true server
Parametric distribution virtual cpu, memory and disk space parameter.The emulation server cluster an of isomery is thus constituted, it is in office
Business is assigned to after cluster and calculates the energy consumption after distribution, and the result with other methods of salary distribution according to the data of virtual server feedback
It compares.
In order to analyze the effect of optimal phase subspace in phase space dispatching algorithm, Fig. 4 has recorded selected optimal of more wheels
Number of nodes in phase subspace accounts for the specific gravity of total node and exceeds use to initial point distance with center of gravity when not having to optimal phase subspace
The specific gravity of optimal phase subspace.
According to Fig. 4 it is found that the throwing being selected in optimal phase subspace when threshold value is respectively set to 0.2,0.4,0.6 and 0.8
Shadow point accounting and center of gravity to initial point distance be more than without using optimal phase subspace when ratio.The section selected when threshold value is 0.2
Point is less, and distance difference is larger, and difference when selection threshold value is between 0.4-0.6 beyond apart from accounting and selection 0.8 is little,
So selecting threshold value can reduce to greatest extent complexity and not increasing excessive additional energy near 0.4-0.6.
Further, by task scheduling algorithm (PSA) proposed by the invention and currently used Min-Min, RR and
Tri- kinds of algorithms of CPSO compare assessment to scheduling result in phase space.Fig. 5 (a)~(d) is 1000 nodes in unit
Between piece task requests it is several 10000 when, the distribution characteristics of four kinds of dispatching algorithm subpoints in dynamic energy consumption phase space, subpoint sit
Mark the average value of multiple time slice scheduling results under the stabilization sub stage after taking emulation to start.Comparison diagram 5 (a)~(d) can be seen that
After RR algorithmic dispatching, node subpoint in dynamic energy consumption phase space in a kind of state dissipated at random, center of gravity from
Origin is also farthest.The reason is that because RR algorithm is adjusted according to the arrival queue of task and the preparation queue of cloud computing system
Degree has biggish randomness, most suitable task execution node cannot be chosen, when task being caused to be distributed in dynamic phase space
Subpoint is also at relatively random distribution.Node after Min-Min algorithmic dispatching is in dynamic energy consumption phase space
Subpoint it is opposite assemble, and its subpoint center of gravity is also more closer apart from origin than the center of gravity after RR algorithmic dispatching, so it is dispatched
Performance is better than RR algorithm.Subpoint after CPSO scheduling is then more gathered near origin compared with Min-Min algorithm.But and PSA
Compare, before center of gravity after the scheduling of three kinds of dispatching algorithms be all larger than the center of gravity after PSA scheduling to initial point distance, root away from initial point distance
Increase smaller corresponding relationship away from the nearlyr cloud computing cluster energy consumption of origin according to subpoint center of gravity in dynamic phase space, PSA scheduling is calculated
The performance of method is better than other three kinds.Although the subpoint of small part node more dissipates after PSA is dispatched, by its tune
The projection dot density highest being gathered in dynamic energy consumption phase space near origin after degree, the reason of generating this feature are PSA
The case where energy optimization for being intended for whole level is dispatched, and Min-Min and CPSO has fallen into local optimum in scheduling.
The above analysis, in the case where the smallest task schedule of cloud computing entirety energy consumption requires, the performance of PSA better than other three
Kind dispatching algorithm.
Fig. 6 indicates the scheduling of each dispatching algorithm when unit timeslice number of tasks is from 5000~30000 under 1000 nodes
As a result distance PSA, NPSA (than the PSA fewer step of selection optimal phase subspace of the center of gravity away from origin in static energy consumption phase space
Suddenly, as other), the scheduling performance of RR, Min-Min and CPSO compares.Close to the center of gravity and origin of NPSA in PSA and NPSA
Distance is increased speed minimum.Although NPSA is slightly better than PSA, PSA algorithm considers optimal phase subspace, it is possible to reduce needs
The number of nodes of consideration can quickly distribute the task of arrival, and NPSA can ignore not relative to the advantage of PSA
Meter, therefore the integrated scheduling best performance of PSA.
Fig. 7 indicates that in unit time task requests number be 10000, when number of nodes is by 100~5000 variation, PSA, Min-
The result of Min, RR, CPSO dispatching algorithm is in the subpoint center of gravity in static phase space to the distance versus situation of origin.With
Being continuously increased for number of nodes, it can be seen that the throwing after each algorithmic dispatching, shadow point center of gravity in phase space arrive origin
Distance decreases.This is because occurring more alternative nodes with the increase of node and then occurring more
Optinal plan, cause whole power consumption state to become excellent.It can see simultaneously, the dispatching effect of CPSO algorithm when number of nodes is less
Better than other algorithms, but with the increase of number of nodes, the scheduling performance of PSA has obtained larger promotion, gradually be more than other three
Kind dispatching algorithm.Therefore it may be concluded that PSA is general for the task schedule effect of a small amount of node, for there is great deal of nodes
Cloud computing system data center's scheduling problem have more obvious advantage.
Embodiment 2
With above method embodiment correspondingly, the present embodiment provides a kind of cloud computing energy optimization system based on phase space
System including memory, processor and stores the computer program that can be run on a memory and on a processor, the processing
The step of device realizes the above method when executing the computer program.
As described above, the present invention provides a kind of cloud computing energy consumption optimization method and system based on phase space, obtain first
The context information of all nodes and node in cloud computing system, and establish energy consumption model；And measuring node is all
The total energy consumption of being carrying out for task establishes static energy consumption phase space according to the total energy consumption；Then new using energy consumption model prediction
Increase the energy consumption of required by task, and establishes dynamic energy consumption phase space；It is mutually empty that optimal son is calculated further according to static energy consumption phase space
Between and predict the energy consumption of newly-increased task each node in optimal phase subspace, according to static energy consumption phase space and dynamic energy consumption mutually sky
Between design objective dispatching algorithm, and determine the optimal scheduling scheme of newly-increased task；Consider each node in cloud computing system
Between context environmental factor, the prediction to energy consumption can be made more accurate, to optimize the energy consumption in cloud computing, reduce task point
With the time, task allocative efficiency is improved.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (9)
1. a kind of cloud computing energy consumption optimization method based on phase space, which is characterized in that include the following steps：
S1：The context information of node and the node all in cloud computing system is obtained, and establishes energy consumption model；
S2：The total energy consumption for measuring all being carrying out for tasks of the node establishes static energy consumption phase space according to the total energy consumption；
S3：The energy consumption of newly-increased required by task is predicted using the energy consumption model, and establishes dynamic energy consumption phase space；
S4：Optimal phase subspace is calculated according to the static energy consumption phase space；
S5：The energy consumption for predicting newly-increased task each node in the optimal phase subspace, according to the static energy consumption phase space and
The dynamic energy consumption phase space design objective dispatching algorithm, and determine the optimal scheduling scheme of newly-increased task.
2. the cloud computing energy consumption optimization method according to claim 1 based on phase space, which is characterized in that in the S2,
The static energy consumption phase space of the foundation specifically includes following steps：
S21：The institute for obtaining the energy consumption parameter composition that all nodes current in the cloud computing system are carrying out task is oriented
Amount；
S22：By all DUAL PROBLEMS OF VECTOR MAPPINGs into phase space, the number of the energy consumption parameter is considered as to the dimension of phase space, it will
The sum of all described vector field homoemorphisms are considered as the current load total energy consumption of cloud computing system, are established according to the load total energy consumption static
Energy consumption phase space.
3. the cloud computing energy consumption optimization method according to claim 1 based on phase space, which is characterized in that in the S3,
The dynamic energy consumption phase space of establishing specifically includes following steps：
S31：Predict that all energy consumptions of the energy consumption parameter composition of the newly-increased task in the cloud computing system increase using energy consumption model
Add vector；
S32：The energy consumption is increased into DUAL PROBLEMS OF VECTOR MAPPING into phase space, the number of parameters of the energy consumption of the newly-increased task is considered as phase
All energy consumptions are increased the sum of vector field homoemorphism and are considered as newly-increased total energy consumption, according to the newly-increased total energy by the dimension in space
Consumption establishes dynamic energy consumption phase space.
4. the cloud computing energy consumption optimization method according to claim 1 based on phase space, which is characterized in that the S4 is specific
Include the following steps：
S41：The reference axis for reacting the node energy consumption state in static energy consumption phase space is established, and determines the static energy consumption phase
The projection point set in space；
S42：In the static energy consumption phase space, the center of gravity of the excessively described projection point set obtains phase as vertical line to the reference axis
Answer the subitem space of quantity；
S43：If coordinate value of any point in each reference axis is respectively less than the coordinate value at center in some subitem space, should
Subitem space is considered as optimal subitem space.
5. the cloud computing energy consumption optimization method according to claim 1 based on phase space, which is characterized in that in the S5,
The optimal scheduling scheme that the determination increases task newly specifically includes following steps：
S51：Newly-increased energy consumption of the task in the optimal subitem space on each node is predicted using the energy consumption model, and is obtained
All possible allocation plan；
S52：The all possible allocation plan is mapped in the dynamic energy consumption phase space using task scheduling algorithm, it will
So that phase space center of gravity is considered as the optimal scheduling scheme of the newly-increased task compared to the smallest scheme of origin moving distance.
6. the cloud computing energy consumption optimization method according to claim 1 based on phase space, which is characterized in that in the S1,
The context information includes node type, network condition and cluster position.
7. the cloud computing energy consumption optimization method according to claim 2 based on phase space, which is characterized in that the vector packet
Include CPU energy consumption, energy consumption of memory and disk energy consumption.
8. the cloud computing energy consumption optimization method according to claim 3 based on phase space, which is characterized in that it is described increase to
Amount includes CPU energy consumption value added, energy consumption of memory value added and disk energy consumption value added.
9. a kind of cloud computing energy optimization system based on phase space, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, which is characterized in that the processor is realized when executing the computer program
The step of any the method for the claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810845198.1A CN108845886B (en) | 2018-07-27 | 2018-07-27 | Cloud computing energy consumption optimization method and system based on phase space |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810845198.1A CN108845886B (en) | 2018-07-27 | 2018-07-27 | Cloud computing energy consumption optimization method and system based on phase space |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108845886A true CN108845886A (en) | 2018-11-20 |
| CN108845886B CN108845886B (en) | 2022-03-08 |
Family
ID=64195686
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810845198.1A Active CN108845886B (en) | 2018-07-27 | 2018-07-27 | Cloud computing energy consumption optimization method and system based on phase space |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108845886B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109947551A (en) * | 2019-03-19 | 2019-06-28 | 中南大学 | A kind of more round method for allocating tasks, edge calculations system and its storage medium |
| CN110096349A (en) * | 2019-04-10 | 2019-08-06 | 山东科技大学 | A kind of job scheduling method based on the prediction of clustered node load condition |
| CN111338649A (en) * | 2020-02-14 | 2020-06-26 | 浪潮商用机器有限公司 | Heterogeneous system, acceleration method and device thereof and readable storage medium |
| CN113205695A (en) * | 2021-04-13 | 2021-08-03 | 东南大学 | Multi-period length bidirectional trunk line green wave control method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103365727A (en) * | 2013-07-09 | 2013-10-23 | 南京大学 | Host load forecasting method in cloud computing environment |
| US20140032185A1 (en) * | 2010-09-17 | 2014-01-30 | The University Of Texas M.D. Anderson Cancer Center | Gpu-based fast dose calculator for cancer therapy |
| CN105389212A (en) * | 2015-10-21 | 2016-03-09 | 浪潮(北京)电子信息产业有限公司 | Job assigning method and apparatus |
| CN105706076A (en) * | 2013-07-31 | 2016-06-22 | 埃克萨公司 | Temperature coupling algorithm for hybrid thermal lattice boltzmann method |
| CN106648890A (en) * | 2016-12-06 | 2017-05-10 | 中国科学院重庆绿色智能技术研究院 | Cloud computing server resource on-line management method and system with energy consumption sensing function |
| CN106951059A (en) * | 2017-03-28 | 2017-07-14 | 中国石油大学(华东) | Based on DVS and the cloud data center power-economizing method for improving ant group algorithm |
| CN107292061A (en) * | 2017-07-28 | 2017-10-24 | 西安交通大学 | A kind of process industry complex electromechanical systems information modelling approach of data-driven |
-
2018
- 2018-07-27 CN CN201810845198.1A patent/CN108845886B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140032185A1 (en) * | 2010-09-17 | 2014-01-30 | The University Of Texas M.D. Anderson Cancer Center | Gpu-based fast dose calculator for cancer therapy |
| CN103365727A (en) * | 2013-07-09 | 2013-10-23 | 南京大学 | Host load forecasting method in cloud computing environment |
| CN105706076A (en) * | 2013-07-31 | 2016-06-22 | 埃克萨公司 | Temperature coupling algorithm for hybrid thermal lattice boltzmann method |
| CN105389212A (en) * | 2015-10-21 | 2016-03-09 | 浪潮(北京)电子信息产业有限公司 | Job assigning method and apparatus |
| CN106648890A (en) * | 2016-12-06 | 2017-05-10 | 中国科学院重庆绿色智能技术研究院 | Cloud computing server resource on-line management method and system with energy consumption sensing function |
| CN106951059A (en) * | 2017-03-28 | 2017-07-14 | 中国石油大学(华东) | Based on DVS and the cloud data center power-economizing method for improving ant group algorithm |
| CN107292061A (en) * | 2017-07-28 | 2017-10-24 | 西安交通大学 | A kind of process industry complex electromechanical systems information modelling approach of data-driven |
Non-Patent Citations (1)
| Title |
|---|
| 王鹏: ""云计算系统相空间分析模型及仿真研究"", 《计算机学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109947551A (en) * | 2019-03-19 | 2019-06-28 | 中南大学 | A kind of more round method for allocating tasks, edge calculations system and its storage medium |
| CN109947551B (en) * | 2019-03-19 | 2021-04-23 | 中南大学 | Multi-turn task allocation method, edge computing system and storage medium thereof |
| CN110096349A (en) * | 2019-04-10 | 2019-08-06 | 山东科技大学 | A kind of job scheduling method based on the prediction of clustered node load condition |
| CN111338649A (en) * | 2020-02-14 | 2020-06-26 | 浪潮商用机器有限公司 | Heterogeneous system, acceleration method and device thereof and readable storage medium |
| CN113205695A (en) * | 2021-04-13 | 2021-08-03 | 东南大学 | Multi-period length bidirectional trunk line green wave control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108845886B (en) | 2022-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mapetu et al. | Low-time complexity and low-cost binary particle swarm optimization algorithm for task scheduling and load balancing in cloud computing | |
| CN110737529B (en) | Short-time multi-variable-size data job cluster scheduling adaptive configuration method | |
| CN110399222B (en) | GPU cluster deep learning task parallelization method and device and electronic equipment | |
| Wang et al. | Load balancing task scheduling based on genetic algorithm in cloud computing | |
| He et al. | Improving resource utilisation in the cloud environment using multivariate probabilistic models | |
| Rekha et al. | Efficient task allocation approach using genetic algorithm for cloud environment | |
| CN110365753A (en) | Internet of Things service low time delay load allocation method and device based on edge calculations | |
| CN108845886A (en) | Cloud computing energy consumption optimization method and system based on phase space | |
| Liu et al. | Resource preprocessing and optimal task scheduling in cloud computing environments | |
| CN109617826A (en) | A kind of storm dynamic load balancing method based on cuckoo search | |
| CN113627871A (en) | Workflow scheduling method, system and storage medium based on multi-target particle swarm algorithm | |
| CN105744006A (en) | Particle swarm optimization user request dispatching method facing multi-type service | |
| CN109947558A (en) | Host resource utilization rate calculation method and resource regulating method | |
| CN109710372A (en) | A kind of computation-intensive cloud workflow schedule method based on cat owl searching algorithm | |
| CN111309472A (en) | Online virtual resource allocation method based on virtual machine pre-deployment | |
| Ghafouri et al. | Mobile-kube: Mobility-aware and energy-efficient service orchestration on kubernetes edge servers | |
| CN111131447A (en) | Load balancing method based on intermediate node task allocation | |
| Chen et al. | Research on workflow scheduling algorithms in the cloud | |
| Jiang et al. | Hierarchical deployment of deep neural networks based on fog computing inferred acceleration model | |
| Zhu et al. | A multi-resource scheduling scheme of Kubernetes for IIoT | |
| CN105426247A (en) | HLA federate planning and scheduling method | |
| Long et al. | A reinforcement learning-based virtual machine placement strategy in cloud data centers | |
| Shuang et al. | Task scheduling based on Grey Wolf optimizer algorithm for smart meter embedded operating system | |
| Zhang et al. | Scheduling strategy for computational-intensive data flow in generalized cluster environments | |
| Kain et al. | Worker resource characterization under dynamic usage in multi-access edge computing |
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 |