CN111158612B - Edge storage acceleration method, device and equipment for cooperative mobile equipment - Google Patents

Edge storage acceleration method, device and equipment for cooperative mobile equipment Download PDF

Info

Publication number
CN111158612B
CN111158612B CN202010252903.4A CN202010252903A CN111158612B CN 111158612 B CN111158612 B CN 111158612B CN 202010252903 A CN202010252903 A CN 202010252903A CN 111158612 B CN111158612 B CN 111158612B
Authority
CN
China
Prior art keywords
mobile device
data
storage
edge
representing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010252903.4A
Other languages
Chinese (zh)
Other versions
CN111158612A (en
Inventor
包卫东
朱晓敏
高雄
王吉
吴冠霖
闫辉
张耀鸿
周文
张雄涛
马力
张亮
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National University of Defense Technology
Original Assignee
National University of Defense Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by National University of Defense Technology filed Critical National University of Defense Technology
Priority to CN202010252903.4A priority Critical patent/CN111158612B/en
Publication of CN111158612A publication Critical patent/CN111158612A/en
Application granted granted Critical
Publication of CN111158612B publication Critical patent/CN111158612B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/061—Improving I/O performance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/0614—Improving the reliability of storage systems
    • G06F3/0619—Improving the reliability of storage systems in relation to data integrity, e.g. data losses, bit errors
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
    • G06F3/0671—In-line storage system
    • G06F3/0683—Plurality of storage devices

Abstract

One or more embodiments of the present invention provide a method, an apparatus, and a device for edge storage acceleration in cooperation with a mobile device, where the method includes: establishing an edge collaborative storage task for transmitting data between mobile devices; edge-based collaborative storage tasksEstablishing an edge collaborative storage model, wherein the edge collaborative storage model comprises the following steps: mobile device collection
Figure 100004_DEST_PATH_IMAGE001
Sending mobile device
Figure 991390DEST_PATH_IMAGE002
Receiving mobile device
Figure 100004_DEST_PATH_IMAGE003
And network topology
Figure 705269DEST_PATH_IMAGE004
Set of mobile devices
Figure 78481DEST_PATH_IMAGE001
Comprises at least two mobile devices; calculating to obtain an A2CS acceleration algorithm based on the edge collaborative storage model; based on an edge collaborative storage model and an A2CS acceleration algorithm, an edge collaborative storage strategy is provided; and guiding the edge cooperative storage task by an edge cooperative storage strategy. The acceleration method provided by one or more embodiments of the present invention considers the unique characteristics of the mobile device and the dynamic network topology of the plurality of mobile devices, ensures flexible data backup, has low latency, and improves storage reliability.

Description

Edge storage acceleration method, device and equipment for cooperative mobile equipment
Technical Field
One or more embodiments of the present invention relate to the field of edge storage technologies, and in particular, to a method, an apparatus, and a device for accelerating edge storage of a cooperative mobile device.
Background
More and more mobile devices are used to collect and process large amounts of data, on one hand, in the military field, soldiers upload and share data through mobile devices such as tactical handheld terminals, UAVs and UGVs (unmanned ground vehicles). Personal data needs to be backed up to avoid data loss due to possible personal injuries and deaths. On the other hand, in the civilian field, mobile devices may be used to store data of disasters or emergencies to perform real-time disaster monitoring. And (4) a data storage mode. However, there are still some obstacles to solving the cooperative storage problem. First, edge mobile devices have unique characteristics compared to conventional storage devices. Unlike the relatively stable network architecture of servers, computers, and other conventional storage devices, the network topology of the edge formed by mobile devices is highly dynamic. For conventional storage problems, in order to ensure availability of data storage, a distributed data backup method is generally used. However, as mobile device resources are gradually consumed, incorrect copy number selection and unreasonable backup data allocation may result in failed backups of data. The selection of the number of copies and the allocation of backup data should be adjusted as the resources of the edge mobile device change. The time to complete the collaborative storage is a key indicator. On the one hand, as the cooperative storage time increases, it is likely to cause a storage failure. On the other hand, as the cost increases over time, more energy is consumed. Therefore, the requirement of low delay time is crucial and of concern. The designed algorithm should converge as quickly as possible to reduce the time overhead and power consumption of the mobile device. The prior art cannot consider a highly dynamic network topology, and has the advantages of low storage reliability, high data backup failure rate, long cooperative storage time and high delay.
Disclosure of Invention
In view of this, one or more embodiments of the present invention provide a method, an apparatus, and a device for accelerating edge storage of a cooperative mobile device, so as to solve the problems that the prior art cannot consider a highly dynamic network topology, and is low in storage reliability, high in data backup failure rate, long in cooperative storage time, and high in latency.
In view of the above, one or more embodiments of the present invention provide an edge storage acceleration method for a cooperative mobile device, including:
establishing an edge collaborative storage task for transmitting data between mobile devices;
establishing an edge collaborative storage model based on the edge collaborative storage task, wherein the edge collaborative storage model comprises: mobile device collection
Figure DEST_PATH_IMAGE001
Sending mobile device
Figure DEST_PATH_IMAGE002
Receiving mobile device
Figure DEST_PATH_IMAGE003
And network topology
Figure DEST_PATH_IMAGE004
The set of mobile devices
Figure 440731DEST_PATH_IMAGE001
Comprises at least two mobile devices;
calculating to obtain an A2CS acceleration algorithm based on the edge collaborative storage model;
based on the edge collaborative storage model and an A2CS acceleration algorithm, an edge collaborative storage strategy is proposed;
and guiding the edge cooperative storage task by the edge cooperative storage strategy.
Optionally, the establishing an edge collaborative storage task for transmitting data between mobile devices includes:
the mobile device collecting data;
the mobile device is based on data loss probability
Figure DEST_PATH_IMAGE005
And energy consumption
Figure DEST_PATH_IMAGE006
Sending the copy of the data to adjacent mobile devices connected with the mobile device, wherein at least one of the adjacent mobile devices is provided;
the set of mobile devices
Figure 762733DEST_PATH_IMAGE001
Determining the number of copies based on the storage capacity and battery level of the neighboring mobile device
Figure DEST_PATH_IMAGE007
And the neighboring mobile device receiving the copy.
Optionally, the set of mobile devices
Figure 135946DEST_PATH_IMAGE001
Is modeled as
Figure DEST_PATH_IMAGE008
Wherein
Figure DEST_PATH_IMAGE009
Is the first
Figure DEST_PATH_IMAGE010
-a plurality of said mobile devices to be controlled,
Figure DEST_PATH_IMAGE011
is the total number of said mobile devices in the edge.
Optionally, the first
Figure DEST_PATH_IMAGE012
A mobile device
Figure 235882DEST_PATH_IMAGE009
Is modeled as
Figure DEST_PATH_IMAGE013
Wherein
Figure DEST_PATH_IMAGE014
To represent
Figure 440467DEST_PATH_IMAGE009
The storage capacity of (a) of (b),
Figure DEST_PATH_IMAGE015
to represent
Figure 387564DEST_PATH_IMAGE009
The amount of battery charge of (a) is,
Figure DEST_PATH_IMAGE016
to represent
Figure 44810DEST_PATH_IMAGE009
The amount of data that is collected is large and small,
Figure DEST_PATH_IMAGE017
to represent
Figure 574536DEST_PATH_IMAGE009
The number of data copies of the collected data,
Figure DEST_PATH_IMAGE018
to represent
Figure 633628DEST_PATH_IMAGE009
The data transmission rate of (a) is,
Figure DEST_PATH_IMAGE019
to represent
Figure 830254DEST_PATH_IMAGE009
The data reception rate of.
Optionally, the network topology
Figure 37113DEST_PATH_IMAGE004
Is modeled as
Figure DEST_PATH_IMAGE020
，
Figure 102021DEST_PATH_IMAGE001
A set of mobile devices is represented as a set of mobile devices,
Figure DEST_PATH_IMAGE021
a contiguous matrix is represented that is,
Figure DEST_PATH_IMAGE022
representing a distance matrix, said adjacency matrix
Figure 284128DEST_PATH_IMAGE021
The method comprises the following steps:
for any of the mobile devices
Figure DEST_PATH_IMAGE023
With other said mobile devices
Figure DEST_PATH_IMAGE024
Is represented by an element in the adjacency matrix,
Figure DEST_PATH_IMAGE025
representing the adjacency matrix
Figure 697661DEST_PATH_IMAGE021
To (1) a
Figure DEST_PATH_IMAGE026
Line and first
Figure DEST_PATH_IMAGE027
The elements in the column, which are determined by the following expression:
Figure DEST_PATH_IMAGE028
the distance matrix
Figure 736024DEST_PATH_IMAGE022
The method comprises the following steps:
for any of the mobile devices
Figure 454888DEST_PATH_IMAGE023
To other said mobile devices
Figure DEST_PATH_IMAGE029
Is the distance matrix
Figure 567201DEST_PATH_IMAGE022
The value of the medium element(s),
Figure DEST_PATH_IMAGE030
representing said distance matrix
Figure 495843DEST_PATH_IMAGE022
To (1) a
Figure 411715DEST_PATH_IMAGE026
Line and first
Figure 756109DEST_PATH_IMAGE027
The elements in the column, which are determined by the following expression:
Figure DEST_PATH_IMAGE031
wherein
Figure DEST_PATH_IMAGE032
To represent
Figure DEST_PATH_IMAGE033
And
Figure DEST_PATH_IMAGE034
the distance between them.
Optionally, the sending mobile device
Figure 771862DEST_PATH_IMAGE002
And receiving mobile device
Figure 684455DEST_PATH_IMAGE003
Form a transceiving mobile device group
Figure DEST_PATH_IMAGE035
The sending mobile device
Figure 25306DEST_PATH_IMAGE002
And receiving mobile device
Figure 173391DEST_PATH_IMAGE003
Data storage and data transmission are carried out between the two, and the data storage generates data storage energy consumption
Figure DEST_PATH_IMAGE036
Said
Figure 119350DEST_PATH_IMAGE036
Is expressed as
Figure DEST_PATH_IMAGE037
Wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE038
is shown as
Figure DEST_PATH_IMAGE039
From the sending mobile device
Figure 376413DEST_PATH_IMAGE002
Transmission to receiving mobile device
Figure 17609DEST_PATH_IMAGE003
The data size of (2), and
Figure DEST_PATH_IMAGE040
the sending mobile device
Figure 156336DEST_PATH_IMAGE002
The amount of data that is stored is of a size,
Figure DEST_PATH_IMAGE041
represents the above-mentioned group
Figure DEST_PATH_IMAGE042
Energy consumption of data storage;
the data transmission generates data transmission energy consumption
Figure DEST_PATH_IMAGE043
Said
Figure 691222DEST_PATH_IMAGE043
Is expressed as
Figure 460464DEST_PATH_IMAGE044
Wherein
Figure DEST_PATH_IMAGE045
Representing the sending mobile device
Figure 854536DEST_PATH_IMAGE002
The transmission power of the antenna is set to be,
Figure DEST_PATH_IMAGE046
representing a receiving mobile device
Figure 675249DEST_PATH_IMAGE003
The received power of the antenna,
Figure DEST_PATH_IMAGE047
representing data from the transmitting mobile device
Figure 454855DEST_PATH_IMAGE002
Transmitting to the receiving mobile device
Figure 880151DEST_PATH_IMAGE003
The time taken for the process to be carried out,
Figure DEST_PATH_IMAGE048
representing the receiving mobile device
Figure 10787DEST_PATH_IMAGE003
Receiving the transmitting mobile device
Figure 242049DEST_PATH_IMAGE002
The time taken for the transmitted data to be transmitted,
Figure DEST_PATH_IMAGE049
which represents the gain of the transmit antenna,
Figure DEST_PATH_IMAGE050
which represents the gain of the receiving antenna,
Figure DEST_PATH_IMAGE051
which represents the wavelength of the light emitted by the light source,
Figure DEST_PATH_IMAGE052
representing the sending mobile device
Figure DEST_PATH_IMAGE053
And the receiving mobile device
Figure 77760DEST_PATH_IMAGE003
The distance between the two or more of the two or more,
Figure DEST_PATH_IMAGE054
representing a system loss factor that is independent of propagation,
Figure DEST_PATH_IMAGE055
representing the sending mobile device
Figure 392067DEST_PATH_IMAGE002
The data transmission rate of (a) is,
Figure DEST_PATH_IMAGE056
represents the connectionReceive mobile device
Figure 760731DEST_PATH_IMAGE003
The data reception rate of (d);
the data storage energy consumption
Figure 779372DEST_PATH_IMAGE036
And said data transmission energy consumption
Figure DEST_PATH_IMAGE057
Summing total energy consumption of edge collaborative storage
Figure DEST_PATH_IMAGE058
Said total energy consumption
Figure 943024DEST_PATH_IMAGE058
Is expressed as
Figure DEST_PATH_IMAGE059
。
Optionally, the A2CS acceleration algorithm includes:
initial value
Figure DEST_PATH_IMAGE060
、
Figure DEST_PATH_IMAGE061
And
Figure DEST_PATH_IMAGE062
the initial value represents the value of the current mobile device
Figure DEST_PATH_IMAGE063
Deploying an initial value of the A2CS acceleration algorithm, the initial value expressed as
Figure DEST_PATH_IMAGE064
Wherein
Figure DEST_PATH_IMAGE065
Is shown and described
Figure 146341DEST_PATH_IMAGE063
The number of connected mobile devices is such that,
Figure DEST_PATH_IMAGE066
to represent
Figure 267881DEST_PATH_IMAGE065
A dimension vector;
a stopping criterion that brings the A2CS acceleration algorithm to a viable solution and ensures that the A2CS acceleration algorithm converges quickly, the stopping criterion expressed as
Figure DEST_PATH_IMAGE067
And
Figure DEST_PATH_IMAGE068
，
wherein
Figure DEST_PATH_IMAGE069
Is shown as
Figure DEST_PATH_IMAGE070
The original residual at the time of the sub-iteration,
Figure DEST_PATH_IMAGE071
is shown as
Figure 624301DEST_PATH_IMAGE070
The dual residual at the time of the sub-iteration,
Figure DEST_PATH_IMAGE072
it is shown that the absolute tolerance is,
Figure DEST_PATH_IMAGE073
indicating a relative tolerance.
Optionally, the edge collaborative storage policy includes the following steps: data collection, request distribution, decision making, decision feedback, multiple interactions and data transmission.
Based on the same inventive concept, one or more embodiments of the present invention further provide an edge storage acceleration apparatus for a cooperative mobile device, including:
the mobile device comprises a first establishing module, a second establishing module and a third establishing module, wherein the first establishing module is configured to establish an edge collaborative storage task for transmitting data between mobile devices;
a second building module configured to build an edge collaborative storage model based on the edge collaborative storage task, the edge collaborative storage model including: mobile device collection
Figure 905109DEST_PATH_IMAGE001
Sending mobile device
Figure 92377DEST_PATH_IMAGE002
Receiving mobile device
Figure 232371DEST_PATH_IMAGE003
And network topology
Figure 609126DEST_PATH_IMAGE004
The set of mobile devices
Figure 278529DEST_PATH_IMAGE001
Comprises at least two mobile devices;
a calculation module configured to calculate an A2CS acceleration algorithm based on the edge collaborative storage model;
a policy module configured to propose an edge collaborative storage policy based on the edge collaborative storage model and an A2CS acceleration algorithm;
an execution module configured to direct the edge collaborative storage task with the edge collaborative storage policy.
Based on the same inventive concept, one or more embodiments of the present invention further provide an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the above-mentioned methods when executing the program.
As can be seen from the foregoing description, according to one or more embodiments of the present invention, an edge storage acceleration method, an edge storage acceleration apparatus, and an edge storage acceleration apparatus for a collaborative mobile device are provided, which particularly pay attention to unique characteristics of a mobile device and a dynamic network topology of a plurality of mobile devices, and convert a collaborative storage problem into a solvable optimization problem, and by studying an acceleration policy, an A2CS acceleration algorithm is proposed to efficiently solve the collaborative storage optimization problem, in an A2CS acceleration algorithm, a convergence speed can be theoretically improved, and a collaborative storage policy is proposed at the same time, the policy includes six steps and may represent an entire process of collaborative storage, and the A2CS acceleration algorithm applies all steps throughout the policy, and guides an entire cycle period of collaborative storage with the policy. The A2CS acceleration algorithm provides better convergence performance under different step rules than the two prior methods ADMM baseline and ADMM-OR (ADMM with excessive relaxation), with an acceleration percentage of at least 25.33% and up to 64.01%. In addition, by performing utility performance comparative analysis using the existing average allocation policy (ADS) and the existing distance-first allocation policy (DPDS), the result shows that the A2CS acceleration algorithm is superior to the ADS policy and the DPDS policy in terms of total utility and energy consumption.
Drawings
In order to more clearly illustrate one or more embodiments or prior art solutions of the present invention, the drawings used in the description of the embodiments or prior art will be briefly described below, and it is obvious that the drawings in the description below are only one or more embodiments of the present invention, and that other drawings can be obtained by those skilled in the art without inventive effort from these drawings.
FIG. 1 is a schematic flow diagram of an acceleration method in accordance with one or more embodiments of the invention;
FIG. 2 is a block diagram of an acceleration algorithm A2CS according to one or more embodiments of the invention;
FIG. 3 is a schematic diagram of a collaborative storage policy in accordance with one or more embodiments of the invention;
FIG. 4 is a schematic illustration of an acceleration device in accordance with one or more embodiments of the present disclosure;
FIG. 5 is a schematic diagram of an electronic device in accordance with one or more embodiments of the invention;
FIG. 6 is a graph of experimental data for normalized values of energy consumption and data loss for a fixed amount of data in one or more embodiments of the invention;
FIG. 7(a) is a diagram illustrating the relationship between the number of replicas and the number of converges and the use of a step rule in one or more embodiments of the present invention
Figure DEST_PATH_IMAGE074
A graph of the total utility of A2CS versus the number of copies;
FIG. 7(b) is a diagram illustrating the relationship between the number of replicas and the number of converges and the use of step size rules in one or more embodiments of the present invention
Figure DEST_PATH_IMAGE075
A graph of the total utility of A2CS versus the number of copies;
FIG. 7(c) is a diagram illustrating the relationship between the number of replicas and the number of converges and the use of a step rule in one or more embodiments of the invention
Figure DEST_PATH_IMAGE076
A graph of the total utility of A2CS versus the number of copies;
FIG. 8(a) is a graph illustrating the effect of the A2CS algorithm using different step size rules on convergence times when the data size is fixed in one or more embodiments of the present invention;
FIG. 8(b) is a graph illustrating the effect of different step size rules on convergence times by the ADMM algorithm when the data size is fixed in one or more embodiments of the invention;
FIG. 8(c) is a graph illustrating the effect of different step size rules on convergence times for the ADMM-OR algorithm with a fixed data size in one OR more embodiments of the invention;
FIG. 9(a) illustrates the use of different algorithms (A2 CS, ADMM and ADMM-OR) in conjunction with step size rules in one OR more embodiments of the invention
Figure DEST_PATH_IMAGE077
A graph of the relationship between the data amount size and the number of convergence times in the case of (1);
FIG. 9(b) illustrates the use of different algorithms (A2 CS, ADMM and ADMM-OR) in conjunction with step size rules in one OR more embodiments of the invention
Figure 839961DEST_PATH_IMAGE076
A graph of the relationship between the data amount size and the number of convergence times in the case of (1); (ii) a
FIG. 10(a) is a graph illustrating the relationship between total utility and percentage of dominance of the A2CS acceleration algorithm over different time ranges relative to the ADS algorithm and the DPDS algorithm using the A2CS acceleration algorithm, respectively, in one or more embodiments of the present invention;
FIG. 10(b) is a graph illustrating the relationship between energy consumption and percentage of dominance of the A2CS acceleration algorithm over the ADS algorithm and the DPDS algorithm over different time ranges using the A2CS acceleration algorithm, the ADS algorithm and the DPDS algorithm, respectively, in one or more embodiments of the present invention;
FIG. 11(a) is a graph of the overall utility of the acceleration algorithm using A2CS in accordance with one or more embodiments of the present invention and a graph of the magnitude of data for different time ranges;
FIG. 11(b) is a graph of energy consumption using the A2CS acceleration algorithm and a graph of data size over different time ranges in accordance with one or more embodiments of the present invention.
Detailed Description
For the purpose of promoting a better understanding of the objects, aspects and advantages of the present disclosure, reference is made to the following detailed description taken in conjunction with the accompanying drawings.
It is to be understood that unless otherwise defined, technical or scientific terms used in one or more embodiments of the present invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this disclosure belongs. The use of "first," "second," and similar terms in one or more embodiments of the invention are not intended to indicate any order, quantity, or importance, but rather are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", and the like are used merely to indicate relative positional relationships, and when the absolute position of the object being described is changed, the relative positional relationships may also be changed accordingly.
One or more embodiments of the invention provide a method, a device and equipment for accelerating edge storage of a cooperative mobile device.
Referring to fig. 1, one or more embodiments of the invention provide a method comprising the steps of:
s101, establishing an edge collaborative storage task for transmitting data between mobile devices.
In this embodiment, on the basis that the edge cooperative storage task establishes the edge cooperative storage framework, the edge cooperative storage framework includes a plurality of mobile devices, and each mobile device includes: the storage unit, the data processor, the scheduler and the data receiver/transmitter, the current storage capacity of different mobile devices is different, and the edge cooperative storage task comprises the following steps:
the mobile device collects data;
the mobile device is based on data loss probability
Figure DEST_PATH_IMAGE078
And energy consumption
Figure DEST_PATH_IMAGE079
Sending the copy of the data to adjacent mobile devices connected with the mobile device, wherein at least one of the adjacent mobile devices is provided;
the set of mobile devices
Figure 795147DEST_PATH_IMAGE001
Based onThe storage capacity and battery level of the neighboring mobile device, determining the number of copies and
Figure DEST_PATH_IMAGE080
the neighboring mobile device receiving the copy.
In order to reduce the probability of data loss
Figure DEST_PATH_IMAGE081
The mobile device sends copies of the data to the connected mobile device while taking into account energy consumption, storage capacity and battery level, number of copies
Figure 896964DEST_PATH_IMAGE080
And the selection of the mobile device for receiving the data is determined by the plurality of mobile devices in the edge, the battery power of different mobile devices is different, if the battery power is low, the mobile device for sending the data does not send the copy to the mobile device with low power, but selects to send the copy of the data to other mobile devices with sufficient power connected with the mobile device for sending the data.
S102, establishing an edge collaborative storage model based on the edge collaborative storage task, wherein the edge collaborative storage model comprises: mobile device collection
Figure 143576DEST_PATH_IMAGE001
Sending mobile device
Figure 813592DEST_PATH_IMAGE002
Receiving mobile device
Figure 600282DEST_PATH_IMAGE003
And network topology
Figure DEST_PATH_IMAGE082
The set of mobile devices
Figure 771369DEST_PATH_IMAGE001
Comprising at least two mobile devices.
In this embodiment, the mobile devices are first assembled
Figure 959905DEST_PATH_IMAGE001
Is modeled as
Figure DEST_PATH_IMAGE083
Wherein
Figure DEST_PATH_IMAGE084
Is the first
Figure 394298DEST_PATH_IMAGE085
-a plurality of said mobile devices to be controlled,
Figure DEST_PATH_IMAGE086
is the total number of said mobile devices in the edge. For any edge mobile device
Figure DEST_PATH_IMAGE087
The first mentioned
Figure 792918DEST_PATH_IMAGE085
A mobile device
Figure DEST_PATH_IMAGE088
Is modeled as
Figure DEST_PATH_IMAGE089
Wherein
Figure DEST_PATH_IMAGE090
To represent
Figure 505047DEST_PATH_IMAGE087
The storage capacity of (a) of (b),
Figure DEST_PATH_IMAGE091
to represent
Figure 735040DEST_PATH_IMAGE088
The amount of battery charge of (a) is,
Figure DEST_PATH_IMAGE092
to represent
Figure 137071DEST_PATH_IMAGE088
The amount of data that is collected is large and small,
Figure 632774DEST_PATH_IMAGE017
to represent
Figure 880085DEST_PATH_IMAGE084
The number of data copies of the collected data,
Figure DEST_PATH_IMAGE093
to represent
Figure 308792DEST_PATH_IMAGE088
The data transmission rate of (a) is,
Figure DEST_PATH_IMAGE094
to represent
Figure 556759DEST_PATH_IMAGE088
The data reception rate of.
TABLE 1 symbolic description
Figure DEST_PATH_IMAGE095
Referring to table 1, dynamic network topology in the edge remains a great challenge for collaborative storage, considering mobile devices and relationships between them, including connectivity and distance, in order to model the network topology. Thus, dynamic network topology in the edge
Figure DEST_PATH_IMAGE096
Is modeled as
Figure DEST_PATH_IMAGE097
，
Figure 992288DEST_PATH_IMAGE001
A set of mobile devices is represented as a set of mobile devices,
Figure DEST_PATH_IMAGE098
a contiguous matrix is represented that is,
Figure DEST_PATH_IMAGE099
representing a distance matrix, said adjacency matrix
Figure 433502DEST_PATH_IMAGE098
The method comprises the following steps:
for any of the mobile devices
Figure DEST_PATH_IMAGE100
With other said mobile devices
Figure 654399DEST_PATH_IMAGE101
Is represented by an element in the adjacency matrix,
Figure DEST_PATH_IMAGE102
representing the adjacency matrix
Figure 729059DEST_PATH_IMAGE098
To (1) a
Figure 402617DEST_PATH_IMAGE103
Line and first
Figure DEST_PATH_IMAGE104
The elements in the column, which are determined by the following expression:
Figure 319626DEST_PATH_IMAGE105
the adjacency matrix defined above
Figure 129451DEST_PATH_IMAGE098
Representing the connection between mobile devices in an ideal state, which means that the connection is sufficiently robust. However, communication interruptions between mobile devices and the withdrawal of a mobile device may result in a loss of connection. For each mobile device, the loss of connection with the other mobile device is random. In the present embodimentIt follows approximately a 0-1 distribution, which can be expressed as:
Figure DEST_PATH_IMAGE106
wherein
Figure 106503DEST_PATH_IMAGE103
Is the probability of connection loss, k represents the order,
Figure 267357DEST_PATH_IMAGE107
a time indicates a connection loss. The distance matrix
Figure 269948DEST_PATH_IMAGE099
The method comprises the following steps:
for any of the mobile devices
Figure 245863DEST_PATH_IMAGE100
To other said mobile devices
Figure 816653DEST_PATH_IMAGE101
Is the distance matrix
Figure 259877DEST_PATH_IMAGE099
The value of the medium element(s),
Figure DEST_PATH_IMAGE108
representing said distance matrix
Figure 472684DEST_PATH_IMAGE099
To (1) a
Figure 568685DEST_PATH_IMAGE103
Line and first
Figure 435010DEST_PATH_IMAGE104
The elements in the column, which are determined by the following expression:
Figure DEST_PATH_IMAGE109
wherein
Figure DEST_PATH_IMAGE110
To represent
Figure DEST_PATH_IMAGE111
And
Figure DEST_PATH_IMAGE112
the distance between them. According to the above definition, the adjacency matrix
Figure 882040DEST_PATH_IMAGE098
And distance matrix
Figure 226434DEST_PATH_IMAGE099
Are all symmetric matrices. In addition to this, the present invention is,
Figure 130936DEST_PATH_IMAGE098
and
Figure 92463DEST_PATH_IMAGE099
the dimension of (a) represents the number of edge mobile devices. With mobile device aggregation
Figure DEST_PATH_IMAGE113
Of a contiguous matrix
Figure 777523DEST_PATH_IMAGE098
And distance matrix
Figure 315820DEST_PATH_IMAGE099
Of the network topology
Figure DEST_PATH_IMAGE114
Changes occur and thus can represent dynamic changes in the network topology in the edge. Since complete and correct data is important in the field of collaborative storage, the present embodiment focuses on the reliability of edge collaborative storage. In addition, the reliability of the collaborative storage is closely related to the redundancy mechanism, the probability of copy loss and the failure rate of the data object, and therefore, the aim is to provideHigh storage reliability, the present embodiment allows for data backup. Will be provided with
Figure DEST_PATH_IMAGE115
Defined as the failure rate of the data storage,
Figure DEST_PATH_IMAGE116
indicating the probability of recovery after a data storage failure. The storage reliability model in the collaborative storage is similar to the Markov process, so any one mobile device can
Figure DEST_PATH_IMAGE117
For example, a mobile device
Figure 917572DEST_PATH_IMAGE117
The probability of data loss can be expressed as:
Figure DEST_PATH_IMAGE118
wherein
Figure DEST_PATH_IMAGE119
Is the number of copies that are to be made,
Figure DEST_PATH_IMAGE120
the order is indicated. Assuming that the data transmitted and stored between different mobile devices are independent and the probability of loss of different data in different mobile devices is the same, the amount of data loss is usedThe storage reliability is described. Obviously, the more data is lost, the lower the storage reliability. The formula can be described as:
Figure 925367DEST_PATH_IMAGE122
wherein
Figure DEST_PATH_IMAGE123
Is a mobile device
Figure DEST_PATH_IMAGE124
The size of the data amount above. When the number of copies is about 5, a higher reliability requirement can be satisfied. When the number of copies is more than 5, improvement in reliability is meaningless, and data maintenance cost is increased. However, the fixed number of copies may be too large resulting in excessive energy costs, or too small resulting in reduced reliability. Furthermore, when stored in a distributed manner, the energy consumption of transmitting data to different mobile devices is dynamically changing. To balance the relationship between storage reliability and energy consumption, the present embodiment considers a flexible data backup strategy consisting of copy number selection and allocation strategy of data copies, which focuses on reliably storing data with as low energy consumption as possible.
For collaborative storage, the duration of a mobile device is a crucial factor. Longer endurance means that more data can be collected, transmitted and stored. Furthermore, the endurance time can be described in terms of energy consumption, which increases when the energy consumption decreases. The energy consumption of edge collaborative storage mainly consists of two parts: energy consumption for data storage and energy consumption for data transmission. In this embodiment, it is assumed that data transmission and reception do not affect each other. In addition, all mobile devices are classified as transmitting devices
Figure DEST_PATH_IMAGE125
And a receiving apparatus
Figure DEST_PATH_IMAGE126
Group of said transmitting mobile device
Figure 409306DEST_PATH_IMAGE125
And receiving mobile device
Figure 361082DEST_PATH_IMAGE127
Form a transceiving mobile device group
Figure 36914DEST_PATH_IMAGE128
Said sending is shiftedMobile equipment
Figure 274997DEST_PATH_IMAGE125
And receiving mobile device
Figure 403490DEST_PATH_IMAGE127
For data storage and data transmission, the transceiving mobile device groups can be formulated as
Figure DEST_PATH_IMAGE129
. Each mobile device may be simultaneously grouped into different groups. The data storage generates data storage energy consumption
Figure 348837DEST_PATH_IMAGE130
Said
Figure 82438DEST_PATH_IMAGE130
Is expressed as
Figure DEST_PATH_IMAGE131
Wherein the content of the first and second substances,
Figure 84898DEST_PATH_IMAGE132
is shown as
Figure DEST_PATH_IMAGE133
From the sending mobile device
Figure 90900DEST_PATH_IMAGE125
Transmission to receiving mobile device
Figure 791003DEST_PATH_IMAGE126
The data size of (2), and
Figure 238164DEST_PATH_IMAGE134
the sending mobile device
Figure 145946DEST_PATH_IMAGE125
The amount of data that is stored is of a size,
Figure DEST_PATH_IMAGE135
represents the above-mentioned group
Figure 717873DEST_PATH_IMAGE136
Energy consumption of data storage. Generating data transfer energy consumption for said data transfer
Figure DEST_PATH_IMAGE137
The present embodiment uses a free space propagation model to describe data transmission in multiple mobile devices. For the sake of analysis, assume
Figure 868670DEST_PATH_IMAGE125
Transmit data to
Figure 435918DEST_PATH_IMAGE126
。
Figure DEST_PATH_IMAGE138
Representing the sending mobile device
Figure 452284DEST_PATH_IMAGE125
The transmission power of the antenna is set to be,
Figure DEST_PATH_IMAGE139
representing a receiving mobile device
Figure 760774DEST_PATH_IMAGE126
The received power of. Based on the Friis transfer formula, will
Figure 68259DEST_PATH_IMAGE138
And
Figure 880226DEST_PATH_IMAGE139
the relationship between them is modeled as:
Figure DEST_PATH_IMAGE140
wherein
Figure DEST_PATH_IMAGE141
Which represents the gain of the transmit antenna,
Figure DEST_PATH_IMAGE142
which represents the gain of the receiving antenna,
Figure DEST_PATH_IMAGE143
which represents the wavelength of the light emitted by the light source,
Figure DEST_PATH_IMAGE144
representing the sending mobile device
Figure 929478DEST_PATH_IMAGE125
And the receiving mobile device
Figure 803893DEST_PATH_IMAGE126
The distance between the two or more of the two or more,
Figure DEST_PATH_IMAGE145
representing a system loss factor independent of propagation. In a collaborative storage framework, the transmit power, wavelength, and system loss factor of each mobile device may be considered constant. In addition, data is sent from the sending mobile device
Figure 836440DEST_PATH_IMAGE125
Transmitting to the receiving mobile device
Figure 112701DEST_PATH_IMAGE126
Time spent and receiving mobile device
Figure 283919DEST_PATH_IMAGE127
Receiving the transmitting mobile device
Figure 35843DEST_PATH_IMAGE125
The time taken for the transmitted data may be expressed as:
Figure DEST_PATH_IMAGE146
wherein
Figure DEST_PATH_IMAGE147
Is shown as
Figure DEST_PATH_IMAGE148
From the sending mobile device
Figure 875011DEST_PATH_IMAGE149
Transmission to receiving mobile device
Figure DEST_PATH_IMAGE150
The size of the data amount of (a), in addition,
Figure DEST_PATH_IMAGE151
representing the sending mobile device
Figure 395991DEST_PATH_IMAGE125
The data transmission rate of (a) is,
Figure DEST_PATH_IMAGE152
representing the receiving mobile device
Figure 206952DEST_PATH_IMAGE126
The data reception rate of. The data transmission rate may be set to a constant since there is no signal attenuation due to propagation, while the data reception rate is related to the Received Signal Strength Indicator (RSSI), which may be described as:
Figure DEST_PATH_IMAGE153
however, the relationship between the data reception rate and the RSSI cannot be described simply by a linear model. Referring to table 2, a correspondence relationship between RSSI and data transmission rate can be obtained,
table 2 correspondence between RSSI and data transmission rate
Figure DEST_PATH_IMAGE154
According to (7), (8) and (9), energy consumption of data transmission
Figure DEST_PATH_IMAGE155
Can be expressed as:
Figure DEST_PATH_IMAGE156
wherein
Figure DEST_PATH_IMAGE157
Representing the above-mentioned group of transceiving mobile devices
Figure DEST_PATH_IMAGE158
Energy consumption of data transmission. Based on (6) and (11), the total energy consumption of edge collaborative storage can be expressed as:
Figure DEST_PATH_IMAGE159
s103, calculating to obtain an A2CS acceleration algorithm based on the edge collaborative storage model.
In this embodiment, the optimization function selected by the A2CS acceleration algorithm is the optimization function F, which is an optimization function
Figure DEST_PATH_IMAGE160
The overall utility of the collaborative storage in the above steps is described, including the reliability and endurance of the storage. The optimization goal is to minimize data loss and power consumption in cooperative storage. In order to meet the requirements of high storage reliability and long endurance time of the collaborative storage, on the basis of the storage reliability and endurance time model, the optimization function of the collaborative storage system is defined as:
Figure 432790DEST_PATH_IMAGE161
wherein
Figure DEST_PATH_IMAGE162
And
Figure DEST_PATH_IMAGE163
is the weight coefficient of the weight of the image,
Figure 807140DEST_PATH_IMAGE164
indicating by a sending mobile device
Figure DEST_PATH_IMAGE165
The size of the amount of data collected,
Figure 713785DEST_PATH_IMAGE164
and
Figure 226806DEST_PATH_IMAGE166
can be described as
Figure DEST_PATH_IMAGE167
。
The cooperative memory problem has turned into an optimization problem, which is very close to the application field of ADMM. Furthermore, the optimization function consists of two independent functions, which is suitable for the resolvability of the ADMM. However, the ADMM also has some disadvantages, such as the convergence speed in reality is very slow, and the cooperative storage problem urgently needs a fast convergence speed, so that the slow convergence speed is not acceptable, but the prior art shows that most of the current research is not suitable for the cooperative storage problem, so that the acceleration strategy should be considered and modified, and the inventor needs to make some adjustments in order to make the algorithm more compatible with the optimization problem, and proposes to standardize and decompose the optimization function. The data storage process is always the same for each mobile device in the edge, so the A2CS acceleration algorithm can be deployed in a distributed manner. To make the optimization function clearly visible, in a basic form satisfying A2CS, a mobile device
Figure 891005DEST_PATH_IMAGE168
For example, the amount of data collected is of the size
Figure DEST_PATH_IMAGE169
The number of the copies is
Figure DEST_PATH_IMAGE170
. Order to
Figure DEST_PATH_IMAGE171
Wherein
Figure 614853DEST_PATH_IMAGE065
Is shown and
Figure 110425DEST_PATH_IMAGE172
the number of connected mobile devices is such that,
Figure DEST_PATH_IMAGE173
. In addition to this, the present invention is,
Figure DEST_PATH_IMAGE174
is represented at each and
Figure 715719DEST_PATH_IMAGE172
the amount of data transferred in the connected mobile device,
Figure DEST_PATH_IMAGE175
is represented at each and
Figure 477001DEST_PATH_IMAGE172
the amount of data stored in the connected mobile device. Corresponds to (13)
Figure DEST_PATH_IMAGE176
And
Figure DEST_PATH_IMAGE177
in this embodiment, there are
Figure 193153DEST_PATH_IMAGE178
And
Figure DEST_PATH_IMAGE179
for movingDevice
Figure 483845DEST_PATH_IMAGE172
. In particular, it is possible to use, for example,
Figure 587936DEST_PATH_IMAGE178
is meant to be moved by the mobile device
Figure 898832DEST_PATH_IMAGE172
The amount of data collected, and
Figure 497303DEST_PATH_IMAGE180
is a vector representing a slave mobile device
Figure 233047DEST_PATH_IMAGE172
Data sent to other mobile devices.
Figure 462034DEST_PATH_IMAGE178
，
Figure DEST_PATH_IMAGE181
，
Figure 447177DEST_PATH_IMAGE182
And
Figure DEST_PATH_IMAGE183
the relationship between can be expressed as:
Figure DEST_PATH_IMAGE184
the original residual can be described as
Figure DEST_PATH_IMAGE185
Can be prepared by
Figure 914586DEST_PATH_IMAGE172
Is normalized to:
Figure 35994DEST_PATH_IMAGE186
wherein
Figure DEST_PATH_IMAGE187
Is and
Figure DEST_PATH_IMAGE188
the associated column vector is then used to determine,
Figure DEST_PATH_IMAGE189
and
Figure 357254DEST_PATH_IMAGE190
in connection with
Figure 111584DEST_PATH_IMAGE191
，
Figure 566705DEST_PATH_IMAGE192
The overall optimization function for all mobile devices can be described as
Figure 496614DEST_PATH_IMAGE193
. After normalization, the
Figure 192038DEST_PATH_IMAGE194
Decomposition into two sub-functions
Figure DEST_PATH_IMAGE195
And
Figure 30068DEST_PATH_IMAGE196
as follows:
Figure 101930DEST_PATH_IMAGE198
the initial values may be set as:
Figure DEST_PATH_IMAGE200
wherein
Figure DEST_PATH_IMAGE201
Is and
Figure DEST_PATH_IMAGE202
number of connected mobile devices, subscript denoting mobile device
Figure 994668DEST_PATH_IMAGE202
Iteration 0.
Figure DEST_PATH_IMAGE203
And
Figure DEST_PATH_IMAGE204
is set to zero, which means that no data is initially stored.
One of the limitations of A2CS is that the sub-function must be a convex function. Therefore, it is necessary to prove that the optimization function is convex and A2CS is applicable. For function
Figure 516785DEST_PATH_IMAGE205
Having only one variable
Figure 386652DEST_PATH_IMAGE206
And other parameters can be assigned values based on background knowledge, which means that
Figure 996625DEST_PATH_IMAGE205
Is a linear function. For function
Figure DEST_PATH_IMAGE207
The variables include
Figure 344868DEST_PATH_IMAGE208
And the distance between two nodes
Figure DEST_PATH_IMAGE209
. Once a distance is reached
Figure 772307DEST_PATH_IMAGE209
Fixed and regarded as constant, then there is only one variable, the function
Figure DEST_PATH_IMAGE210
Can be simplified to a linear function. Based on the collaborative storage framework and the collaborative storage model, each mobile device can obtain distance information from the connected mobile device, which means that the distance can be set to be constant, and
Figure DEST_PATH_IMAGE211
it can be further simplified to a linear function. Thus, the convexity of the sub-function is demonstrated as follows.
And (3) proving that:
as defined in (17) and (18), the derivative of sum can be described as:
Figure DEST_PATH_IMAGE213
Figure 644317DEST_PATH_IMAGE214
and
Figure DEST_PATH_IMAGE215
are all positive numbers, and thus
Figure DEST_PATH_IMAGE216
，
Figure DEST_PATH_IMAGE217
. Order to
Figure 57981DEST_PATH_IMAGE218
，
Figure DEST_PATH_IMAGE219
It is possible to obtain:
Figure DEST_PATH_IMAGE220
therefore, the temperature of the molten metal is controlled,
Figure 803608DEST_PATH_IMAGE221
and
Figure DEST_PATH_IMAGE222
all satisfy subfunction convexity proving formula in background knowledge
Figure 401948DEST_PATH_IMAGE223
I.e. both subfunctions are convex.
After the syndrome is confirmed.
Thus, subfunctions
Figure 371041DEST_PATH_IMAGE221
And
Figure 260500DEST_PATH_IMAGE222
may prove convex and A2CS is applicable.
According to the ADMM theory, a reasonable stopping criterion is set for the A2CS acceleration algorithm in this embodiment to obtain a satisfactory feasible solution and ensure fast convergence. When the original residual and the dual residual are small, the target suboptimum must also be small. In particular for mobile devices
Figure 123282DEST_PATH_IMAGE202
Using the original residual error
Figure DEST_PATH_IMAGE224
Sum and dual residual
Figure 908836DEST_PATH_IMAGE225
And may be at the second
Figure DEST_PATH_IMAGE226
The sub-iteration is formulated as:
Figure DEST_PATH_IMAGE227
thus, a reasonable stopping criterion is proposed:
Figure DEST_PATH_IMAGE228
wherein
Figure DEST_PATH_IMAGE229
Representing the original residual at the k-th iteration,
Figure DEST_PATH_IMAGE230
representing the dual residual at the kth iteration,
Figure 351843DEST_PATH_IMAGE231
it is shown that the absolute tolerance is,
Figure DEST_PATH_IMAGE232
indicating a relative tolerance. Furthermore, the method of tolerance selection is to use absolute and relative tolerances, which can be expressed as:
Figure 294260DEST_PATH_IMAGE233
wherein
Figure DEST_PATH_IMAGE234
Is a tolerance which is an absolute tolerance and,
Figure 214812DEST_PATH_IMAGE235
is a relative tolerance that is a function of,
Figure 30321DEST_PATH_IMAGE201
is a variable quantity
Figure DEST_PATH_IMAGE236
Dimension (d) of (a).
Referring to fig. 2, the optimization function, the initial value and the stopping criterion in the present solution are integrated, and based on the acceleration gradient descent scheme NA and the ADMM algorithm, the scaling form and NA of the ADMM algorithm are combined, so as to obtain the A2CS acceleration algorithm in the present embodiment.
S104, based on the edge collaborative storage model and the A2CS acceleration algorithm, an edge collaborative storage strategy is proposed.
To show in detail and intuitivelyShowing the process of cooperative storage, the embodiment proposes an edge cooperative storage strategy based on an edge cooperative storage model and an A2CS acceleration algorithm, and referring to fig. 3, in the strategy, a name is added
Figure 849372DEST_PATH_IMAGE026
To distinguish roles of each mobile device at the middle edge, wherein
Figure 329901DEST_PATH_IMAGE237
. When in use
Figure DEST_PATH_IMAGE238
A mobile device is a requesting node that transmits data to other connected mobile devices. When in use
Figure DEST_PATH_IMAGE239
A mobile device is a storage node that receives data from other connected mobile devices. It is obvious that the mobile device may be both a requesting node and a storing node. For a certain mobile device
Figure DEST_PATH_IMAGE240
It collects data
Figure DEST_PATH_IMAGE241
And then sends the message to the connected mobile device. It then determines the number of copies, selects the mobile devices to store, and determines the data volume allocation for each selected mobile device. At the same time it can also receive requests from other connected mobile devices and decide whether to receive data.
In this embodiment, the edge collaborative storage policy includes six steps: data collection, request distribution, decision making, decision feedback, multiple interactions and data transmission. Data collection step referring to (a) of fig. 3, different types of mobile devices, which are represented by rectangles and circles, have different roles. In addition, the mobile device is in different states at the same time. The shaded rectangles and circles represent requesting nodes, while the others represent free storage nodes. The solid lines in the picture represent the connections between different mobile devices. The requesting node is collecting data and preparing to send data storage messages, while the storage node is ready to receive data storage messages from the requesting node. Request distribution step referring to (b) in fig. 3, the requesting node transmits a data storage message to the connected mobile device and waits for feedback. The solid arrows represent data storage messages from the requesting node to the storage nodes. Decision making step referring to (c) of fig. 3, a free storage node receives a data storage message and becomes busy, represented by the shaded rectangle and the dashed circle. The busy storage node then decides how to respond to the request message, taking into account storage capacity and battery power. Decision feedback step referring to (d) in fig. 3, the busy storage node returns feedback to the requesting node in (a) according to the decision in (c). The decision includes whether and how much to store. (d) The dashed arrow in (d) represents decision feedback. Multiple interaction step referring to (e) in fig. 3, the requesting node interacts with the storage node multiple times to obtain a feasible solution. The two-way dashed arrows represent interactions between different mobile devices. Data transmission step referring to (f) in fig. 3, the requesting node in (a) receives and aggregates the feedback. From the summary, the requesting node first calculates and determines the number of copies. They then select a copy of the data and transmit it to some storage node. In (f), data transmission is represented by solid arrows and data icons. The entire process of cooperative storage is repeated from (a) to (f), and the mobile device may be in more than one of the above states at the same time.
S105, guiding the edge cooperative storage task by the edge cooperative storage strategy.
The edge cooperative storage policy covers the whole process of cooperative storage, and the edge cooperative storage policy guides the edge cooperative storage task established in step S101 so as to apply the edge cooperative storage task to the task execution scenario grouped by the mobile devices in the edge.
As can be seen from the foregoing description, according to one or more embodiments of the present invention, an edge storage acceleration method, an edge storage acceleration apparatus, and an edge storage acceleration apparatus for a collaborative mobile device are provided, which particularly pay attention to unique characteristics of a mobile device and a dynamic network topology of a plurality of mobile devices, and convert a collaborative storage problem into a solvable optimization problem, and by studying an acceleration policy, an A2CS acceleration algorithm is proposed to efficiently solve the collaborative storage optimization problem, in an A2CS acceleration algorithm, a convergence speed can be theoretically improved, and a collaborative storage policy is proposed at the same time, the policy includes six steps and may represent an entire process of collaborative storage, and the A2CS acceleration algorithm applies all steps throughout the policy, and guides an entire cycle period of collaborative storage with the policy. The A2CS acceleration algorithm provides better convergence performance under different step rules than the two prior methods ADMM baseline and ADMM-OR (ADMM with excessive relaxation), with an acceleration percentage of at least 25.33% and up to 64.01%. In addition, by performing utility performance comparison analysis using the existing average allocation policy (ADS) and the existing distance-first allocation policy (DPDS), the result shows that the A2CS acceleration algorithm is superior to ADS and DPDS in terms of total utility and power consumption.
It should be noted that the method of one or more embodiments of the present invention may be performed by a single device, such as a computer or server. The method of the embodiment can also be applied to a distributed scene and completed by the mutual cooperation of a plurality of devices. In such a distributed scenario, one of the devices may perform only one or more steps of the method of one or more embodiments of the present invention, and the devices may interact with each other to complete the method.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
Referring to fig. 4, based on the same inventive concept, one or more embodiments of the present invention further provide an edge storage acceleration apparatus for a cooperative mobile device, including: the device comprises a first establishing module, a second establishing module, a calculating module, a strategy module and an executing module.
The first establishing module is configured to establish an edge collaborative storage task for transmitting data between the mobile devices;
a second building module configured to build an edge collaborative storage model based on the edge collaborative storage task, the edge collaborative storage model including: mobile device collection
Figure DEST_PATH_IMAGE242
Sending mobile device
Figure 232522DEST_PATH_IMAGE243
Receiving mobile device
Figure 953354DEST_PATH_IMAGE003
And network topology
Figure DEST_PATH_IMAGE244
The set of mobile devices
Figure 712231DEST_PATH_IMAGE242
Comprises at least two mobile devices;
a calculation module configured to calculate an A2CS acceleration algorithm based on the edge collaborative storage model;
a policy module configured to propose an edge collaborative storage policy based on the edge collaborative storage model and an A2CS acceleration algorithm;
an execution module configured to direct the edge collaborative storage task with the edge collaborative storage policy.
For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. Of course, the functionality of the various modules may be implemented in the same one or more software and/or hardware implementations of one or more embodiments of the invention.
The apparatus of the foregoing embodiment is used to implement the corresponding method in the foregoing embodiment, and has the beneficial effects of the corresponding method embodiment, which are not described herein again.
Based on the same inventive concept, one or more embodiments of the present invention further provide an electronic device, which includes a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the program, the method according to any of the above embodiments is implemented.
Fig. 5 is a schematic diagram illustrating a more specific hardware structure of an electronic device according to this embodiment, where the electronic device may include: a processor 501, a memory 502, an input/output interface 503, a communication interface 504, and a bus 505. Wherein the processor 501, the memory 502, the input/output interface 503 and the communication interface 504 are communicatively connected to each other within the device via a bus 505.
The processor 501 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits, and is configured to execute related programs to implement the technical solutions provided in the embodiments of the present specification.
The Memory 502 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random access Memory), a static storage device, a dynamic storage device, or the like. The memory 502 may store an operating system and other application programs, and when the technical solution provided by the embodiments of the present specification is implemented by software or firmware, the relevant program codes are stored in the memory 5020 and called by the processor 501 for execution.
The input/output interface 503 is used for connecting an input/output module to realize information input and output. The i/o module may be configured as a component in a device (not shown) or may be external to the device to provide a corresponding function. The input devices may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output devices may include a display, a speaker, a vibrator, an indicator light, etc.
The communication interface 504 is used for connecting a communication module (not shown in the figure) to realize communication interaction between the device and other devices. The communication module can realize communication in a wired mode (such as USB, network cable and the like) and also can realize communication in a wireless mode (such as mobile network, WIFI, Bluetooth and the like).
Bus 505 comprises a path that transfers information between the various components of the device, such as processor 501, memory 502, input/output interface 503, and communication interface 504.
It should be noted that although the above-mentioned device only shows the processor 501, the memory 502, the input/output interface 503, the communication interface 504 and the bus 505, in a specific implementation, the device may also include other components necessary for normal operation. In addition, those skilled in the art will appreciate that the above-described apparatus may also include only those components necessary to implement the embodiments of the present description, and not necessarily all of the components shown in the figures.
To verify the performance of the A2CS acceleration algorithm, the inventors designed and carried out a series of experiments, in particular, taking an Unmanned Aerial Vehicle (UAV) as an example, and used the parameters of DJIUAV in the following. The experiments were performed using the step rule.
First, assume that the edge contains 20 drones. The network topology of the edge includes connectivity and distance between drones at different times. To simulate the dynamics of the edge network topology in a cooperative storage scheme, the connectivity and distance between drones is represented using random numbers within a given range. Specifically, a random number between 1 and 19 is first generated to represent the number of UAVs that can be connected to the requesting UAV to allow data transmission. This means that the requesting drone can connect up to 1 drone and up to 19 drones, consistent with the actual situation. Furthermore, since the loss of connection with another mobile device occurs approximately following a 0-1 distribution, a random number is used to generate a 0-1 Boolean variable to represent the connection state. Then, the distance between drones is also randomly generated within the range of [5,20 ]. The upper and lower bounds are derived from typical values of distances between drones in a cluster of small scale drones.
Furthermore, to verify the validity of the collaborative storage model and reduce the time cost, a simple pre-experiment was performed on the number of copies. Referring to fig. 6, normalized values of energy consumption and data loss for a fixed data size are shown. As the number of copies increases, energy consumption increases and data loss decreases, which means that endurance and marginal storage reliability of the mobile device increases. Therefore, the number of copies should be selected to take into account both energy consumption and data loss, consistent with the collaborative storage model described above. In addition, more than 8 copies do not reduce data loss but rather increase energy consumption, which is considered herein as an unacceptable method of cooperative storage. Therefore, the number of copies is set to 1 to 8 to avoid unnecessary experiments and reduce time cost.
After multiple tests, the values of the parameters are finally set, and the augmented Lagrange parameter is set as
Figure 809500DEST_PATH_IMAGE245
Coefficient of weight
Figure 986535DEST_PATH_IMAGE162
And
Figure DEST_PATH_IMAGE246
is a hyper-parameter, which is set to
Figure DEST_PATH_IMAGE247
，
Figure DEST_PATH_IMAGE248
，μ=0.001，θ=0.002，
Figure DEST_PATH_IMAGE249
，PT=100mW，GT=2dbi，GR=2dbi，λ=0.125m，L=1，c=[5,30]MB, setting up
Figure DEST_PATH_IMAGE250
And
Figure 470080DEST_PATH_IMAGE251
satisfied after multiple testsAnd (5) carrying out feasible solution. The performance of the comparison algorithm is mainly evaluated from three indexes, including: convergence times, overall utility and energy consumption, wherein the convergence times represent the convergence speed of the algorithm; the overall utility represents the value of the optimization function using different algorithms; energy consumption describes the duration of a mobile device.
ADMM-OR algorithm is derived from ADMM algorithm, and pairs are added
Figure DEST_PATH_IMAGE252
Can be expressed as:
Figure 450674DEST_PATH_IMAGE253
wherein
Figure DEST_PATH_IMAGE254
Is the relaxation parameter. Particularly when
Figure 492579DEST_PATH_IMAGE255
This mode is called over-relaxation. After excessive relaxation, is used in the next step
Figure DEST_PATH_IMAGE256
Updating variables
Figure 632442DEST_PATH_IMAGE257
And dual variables
Figure DEST_PATH_IMAGE258
. When in use
Figure 101601DEST_PATH_IMAGE259
In the middle, it was shown that the astringency was improved. Thus, in the experiment, set up
Figure 366229DEST_PATH_IMAGE260
. For better convergence performance, the inventors added the step rule when conducting experiments. Step size rule refers to evaluating
Figure DEST_PATH_IMAGE261
By different methods of (2), which means
Figure 742984DEST_PATH_IMAGE261
Not fixed in each iteration. This operation is advantageous in that it is possible to improve the convergence speed in practice and reduce the dependence of the performance on the initial value. When in use
Figure 412387DEST_PATH_IMAGE261
It is difficult to prove convergence when changing in each iteration, but if it is changed, it is difficult to verify convergence
Figure 583605DEST_PATH_IMAGE261
The above theory still applies if it can become some fixed value after a limited number of iterations. Therefore, two other step size rules are proposed, among which
Figure 945317DEST_PATH_IMAGE262
Can be fixed within the accuracy range required by the cooperative storage problem. They can be represented as:
Figure DEST_PATH_IMAGE263
first, 20 experiments were performed and the results were each evaluated on average based on the same experimental setup using the three algorithms and the three step rules described above. Specifically, the data size is set to
Figure 437347DEST_PATH_IMAGE264
. Referring to fig. 7(a), 7(b) and 7(c), the histogram indicates the relationship between the number of duplicates and the number of convergence times, while the line graph shows the relationship between the total utility of the acceleration algorithm using A2CS with different step size rules and the number of duplicates. As the number of copies increases, when
Figure DEST_PATH_IMAGE265
，
Figure 302535DEST_PATH_IMAGE266
And
Figure 97184DEST_PATH_IMAGE267
the number of convergence times also increases. For each step size rule, the proposed A2CS acceleration algorithm converges the least number of times, exhibiting the best convergence performance. Furthermore, the overall utility of using the A2CS acceleration algorithm first decreases and then increases. The minimum is reached when the number of copies equals 4, which means that the best choice for the edge requesting node is to backup 4 copies and send them to the connected storage node. It can also be concluded that the step size rule greatly affects the convergence speed, but has little impact on the overall utility.
Referring to fig. 8(a), 8(b) and 8(c), when the data size is fixed, the influence of different step rules on the convergence number is used. In general, step size rules are used
Figure 946191DEST_PATH_IMAGE265
And
Figure 602432DEST_PATH_IMAGE267
than using a fixed step rule
Figure 118864DEST_PATH_IMAGE266
The results are better. Although there are some exceptions, the difference between those exceptional results using the three step rule is modest, which is acceptable. Furthermore, it was found that certain step-size rules do not always have better performance than other step-size rules when used in conjunction with other algorithms. In particular, the results with the fastest convergence rates were selected for each algorithm and are listed in table 4. The results of these three combinations were compared. The best performance is underlined in table 3 and the percent acceleration is calculated.
TABLE 3 influence of the number of copies on the convergence of the average number of different combinations
Figure DEST_PATH_IMAGE269
Reference is made to FIG. 9(a)And fig. 9(b), in order to analyze the influence of the data size on the convergence performance, another 20 experiments were performed and the results were averaged, and the number of copies was set to 4 according to the results in fig. 7. Since the results in FIG. 8 show the use
Figure 25027DEST_PATH_IMAGE266
Relatively slow convergence rate of using only
Figure 361331DEST_PATH_IMAGE265
And
Figure 70530DEST_PATH_IMAGE267
the results are analyzed. The histogram specifies the relationship between the data size and the number of convergence times. Generally, the number of convergence increases as the size of the data amount increases. It is noteworthy that as the size of the data volume increases, A2CS performs much better than other algorithms, which means that the size of the data volume has little impact on the performance of A2CS, but much impact on ADMM and ADMM-OR. In particular, for each algorithm, a combination of the algorithm and the step size rule with the fastest convergence rate is selected. Selected results are shown in table 4, with the best performance underlined and the percent acceleration calculated. As the size of the data volume increases, the acceleration percentage also increases, which means that the proposed algorithm A2CS performs better when processing large amounts of data.
TABLE 4 influence of data size on mean convergence for different combinations
Figure 113572DEST_PATH_IMAGE271
The assignment of data copies is closely related to the utility of the optimization function and is analyzed in the proposed algorithm A2CS, so the utility performance analysis is defined as a comparison between A2CS and the algorithm that also performs the assignment of data. The algorithm compared to the A2CS acceleration algorithm includes:
ADS (average allocation strategy): uniformly distributing the data copies to the connected storage nodes;
DPDS (distance priority allocation policy): the DPDS prioritizes the distance between the requesting node and the connected storage nodes. The shorter the distance, the higher the priority.
Based on the above acceleration performance analysis, when the data amount sizes were randomly selected within the ranges of [5,30] MB, respectively, the number of data copies was set to 4, and 20 experiments were performed.
Referring to fig. 10(a), a histogram depicts the total utility of using different algorithms, and referring to fig. 10(b), a histogram depicts the energy consumption of using different algorithms, and accordingly, a line graph represents the percentage of dominance of A2CS over ADS and DPDS, respectively, over different time ranges. Overall, the proposed algorithm A2CS shows better performance than ADS and DPDS in both total utility and energy consumption. The results show that the data copy distribution strategy of A2CS can reduce the total utility and energy consumption to the maximum extent, thereby meeting the optimization target of high storage reliability of mobile device edge and long endurance time.
Referring to FIG. 11(a), the overall utility of the A2CS acceleration algorithm, referring to FIG. 11(b), the energy consumption of the A2CS acceleration algorithm, to analyze the effect of the step size rule on utility, another 20 experiments were conducted using A2CS, in which
Figure 125390DEST_PATH_IMAGE272
，
Figure DEST_PATH_IMAGE273
And
Figure DEST_PATH_IMAGE274
. Discovery use
Figure 11307DEST_PATH_IMAGE275
A2CS of (a) performed almost identically to the other two combinations in most experiments, and only in some cases better. Note that, when the number of copies is set to 4,
Figure DEST_PATH_IMAGE276
a2CS of (1) has the best acceleration performance. Although it is not limited to
Figure DEST_PATH_IMAGE277
A2CS of (a) may accelerate convergence even better. In some experiments, the results of utility performance indicate that the step size rule has little impact on utility performance.
Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure, including the claims, is limited to these examples; features from the above embodiments or from different embodiments may also be combined, steps may be implemented in any order, and there are many other variations of different aspects of one or more embodiments of the invention as described above, which are not provided in detail for the sake of brevity, within the spirit of this disclosure.
In addition, well-known power/ground connections to Integrated Circuit (IC) chips and other components may or may not be shown within the provided figures, for simplicity of illustration and discussion, and so as not to obscure one or more embodiments of the invention. Furthermore, devices may be shown in block diagram form in order to avoid obscuring the one or more embodiments of the present invention, and this also takes into account the fact that specifics with respect to implementation of such block diagram devices are highly dependent upon the platform within which the one or more embodiments of the present invention are to be implemented (i.e., specifics should be well within purview of one skilled in the art). Where specific details (e.g., circuits) are set forth in order to describe example embodiments of the disclosure, it should be apparent to one skilled in the art that one or more embodiments of the disclosure can be practiced without, or with variation of, these specific details. Accordingly, the description is to be regarded as illustrative instead of restrictive.
It is intended that the one or more embodiments of the present invention embrace all such alternatives, modifications and variations as fall within the broad scope of the appended claims. Therefore, any omissions, modifications, substitutions, improvements, and the like that may be made without departing from the spirit and principles of one or more embodiments of the present invention are intended to be included within the scope of the present disclosure.

Claims (9)

1. An edge storage acceleration method for a cooperative mobile device, comprising:
establishing an edge collaborative storage task for transmitting data between mobile devices;
establishing an edge collaborative storage model based on the edge collaborative storage task, wherein the edge collaborative storage model comprises: mobile device collection
Figure 766896DEST_PATH_IMAGE001
Sending mobile device
Figure 898800DEST_PATH_IMAGE002
Receiving mobile device
Figure 249228DEST_PATH_IMAGE003
And network topology
Figure 875512DEST_PATH_IMAGE004
The set of mobile devices
Figure 966965DEST_PATH_IMAGE005
Comprises at least two mobile devices;
calculating an A2CS acceleration algorithm based on the edge collaborative storage model, wherein the A2CS acceleration algorithm comprises:
initial value
Figure 82820DEST_PATH_IMAGE006
、
Figure 643245DEST_PATH_IMAGE007
And
Figure 322488DEST_PATH_IMAGE008
the initial value represents the value of the current mobile device
Figure 19180DEST_PATH_IMAGE009
Deploying an initial value of the A2CS acceleration algorithm, the initial value expressed as
Figure 305936DEST_PATH_IMAGE010
Wherein
Figure 353658DEST_PATH_IMAGE011
Is shown and described
Figure 367750DEST_PATH_IMAGE012
The number of connected mobile devices is such that,
Figure 450107DEST_PATH_IMAGE013
to represent
Figure 642185DEST_PATH_IMAGE011
A dimension vector;
a stopping criterion that brings the A2CS acceleration algorithm to a viable solution and ensures that the A2CS acceleration algorithm converges quickly, the stopping criterion expressed as
Figure 442782DEST_PATH_IMAGE014
，
Wherein
Figure 339193DEST_PATH_IMAGE015
Is shown as
Figure 744898DEST_PATH_IMAGE016
The original residual at the time of the sub-iteration,
Figure 373457DEST_PATH_IMAGE017
is shown as
Figure 130191DEST_PATH_IMAGE016
The dual residual at the time of the sub-iteration,
Figure 486086DEST_PATH_IMAGE018
it is shown that the absolute tolerance is,
Figure 277456DEST_PATH_IMAGE019
indicating a relative tolerance;
optimization function
Figure 280178DEST_PATH_IMAGE020
For describing the total utility of the edge collaborative storage task, the optimization function is expressed as
Figure 39055DEST_PATH_IMAGE021
Wherein
Figure 937655DEST_PATH_IMAGE022
And
Figure 521214DEST_PATH_IMAGE023
is the weight coefficient of the weight of the image,
Figure 6422DEST_PATH_IMAGE024
indicating by a sending mobile device
Figure 941011DEST_PATH_IMAGE025
The size of the amount of data collected,
Figure 389441DEST_PATH_IMAGE024
and
Figure 139091DEST_PATH_IMAGE026
can be described as
Figure 545933DEST_PATH_IMAGE027
，
Figure 498977DEST_PATH_IMAGE028
Representing the sending mobile device
Figure 469207DEST_PATH_IMAGE029
The number of data copies of the collected data,
Figure 620834DEST_PATH_IMAGE030
which is indicative of the amount of data loss,
Figure 136260DEST_PATH_IMAGE031
represents the total energy consumption for the storage of the data,
Figure 842179DEST_PATH_IMAGE032
indicating transmitting mobile device
Figure 84941DEST_PATH_IMAGE033
The probability of data loss of (a) is,
Figure 887812DEST_PATH_IMAGE034
representing the sending mobile device
Figure 370877DEST_PATH_IMAGE029
The data transmission rate of (a) is,
Figure 298513DEST_PATH_IMAGE035
representing the receiving mobile device
Figure 344966DEST_PATH_IMAGE036
The rate of reception of the data of (c),
Figure 940027DEST_PATH_IMAGE037
representing adjacency matrices
Figure 593993DEST_PATH_IMAGE038
To (1) a
Figure 461455DEST_PATH_IMAGE039
Line and first
Figure 859070DEST_PATH_IMAGE040
The elements in the column are selected from the group,
Figure 308637DEST_PATH_IMAGE041
representing a distance matrix
Figure 586034DEST_PATH_IMAGE042
To (1) a
Figure 499981DEST_PATH_IMAGE039
Line and first
Figure 904549DEST_PATH_IMAGE040
The elements in the column are selected from the group,
Figure 457890DEST_PATH_IMAGE043
representing the power consumption for storing a unit of data,
Figure 656921DEST_PATH_IMAGE044
representing the sending mobile device
Figure 312024DEST_PATH_IMAGE029
The transmission power of the antenna is set to be,
Figure 300709DEST_PATH_IMAGE045
which represents the gain of the transmit antenna,
Figure 396972DEST_PATH_IMAGE046
which represents the gain of the receiving antenna,
Figure 94801DEST_PATH_IMAGE047
which represents the wavelength of the light emitted by the light source,
Figure 424151DEST_PATH_IMAGE048
representing the sending mobile device
Figure 701680DEST_PATH_IMAGE029
And the receiving mobile device
Figure 167296DEST_PATH_IMAGE036
The distance between the two or more of the two or more,
Figure 36026DEST_PATH_IMAGE049
representing a system loss factor independent of propagation;
based on the edge collaborative storage model and an A2CS acceleration algorithm, an edge collaborative storage strategy is proposed;
and guiding the edge cooperative storage task by the edge cooperative storage strategy.
2. The method of claim 1, wherein establishing an edge collaborative storage task for data transmission between mobile devices comprises:
the mobile device collecting data;
the mobile device is based on data loss probability
Figure 868984DEST_PATH_IMAGE050
And energy consumption
Figure 668313DEST_PATH_IMAGE051
Sending the copy of the data to adjacent mobile devices connected with the mobile device, wherein at least one of the adjacent mobile devices is provided;
the set of mobile devices
Figure 801485DEST_PATH_IMAGE052
Determining the number of copies based on the storage capacity and battery level of the neighboring mobile device
Figure 309958DEST_PATH_IMAGE053
And the neighboring mobile device receiving the copy.
3. The method of claim 1, wherein the set of mobile devices
Figure 348321DEST_PATH_IMAGE052
Is modeled as
Figure 498811DEST_PATH_IMAGE054
Wherein
Figure 220910DEST_PATH_IMAGE055
Is the first
Figure 165864DEST_PATH_IMAGE011
-a plurality of said mobile devices to be controlled,
Figure 222681DEST_PATH_IMAGE056
is the total number of said mobile devices in the edge.
4. The method of claim 3, wherein the first step is performed
Figure 114545DEST_PATH_IMAGE011
A mobile device
Figure 956730DEST_PATH_IMAGE055
Is modeled as
Figure 525115DEST_PATH_IMAGE057
Wherein
Figure 339401DEST_PATH_IMAGE058
To represent
Figure 18644DEST_PATH_IMAGE055
The storage capacity of (a) of (b),
Figure 980915DEST_PATH_IMAGE059
to represent
Figure 2092DEST_PATH_IMAGE055
The amount of battery charge of (a) is,
Figure 236764DEST_PATH_IMAGE060
to represent
Figure 736009DEST_PATH_IMAGE055
The amount of data that is collected is large and small,
Figure 287208DEST_PATH_IMAGE061
to represent
Figure 259712DEST_PATH_IMAGE055
The number of data copies of the collected data,
Figure 529150DEST_PATH_IMAGE062
to represent
Figure 300928DEST_PATH_IMAGE055
The data transmission rate of (a) is,
Figure 487059DEST_PATH_IMAGE063
to represent
Figure 318880DEST_PATH_IMAGE055
The data reception rate of.
5. The method of claim 1, wherein the network topology
Figure 528144DEST_PATH_IMAGE064
Is modeled as
Figure 900351DEST_PATH_IMAGE065
，
Figure 691720DEST_PATH_IMAGE052
A set of mobile devices is represented as a set of mobile devices,
Figure 943710DEST_PATH_IMAGE066
a contiguous matrix is represented that is,
Figure 656582DEST_PATH_IMAGE067
representing a distance matrix, said adjacency matrix
Figure 832480DEST_PATH_IMAGE068
The method comprises the following steps:
for any of the mobile devices
Figure 665307DEST_PATH_IMAGE069
With other said mobile devices
Figure 635668DEST_PATH_IMAGE070
Is represented by an element in the adjacency matrix,
Figure 366995DEST_PATH_IMAGE071
representing the adjacency matrix
Figure 64692DEST_PATH_IMAGE066
To (1) a
Figure 565075DEST_PATH_IMAGE072
Line and first
Figure 503075DEST_PATH_IMAGE073
The elements in the column, which are determined by the following expression:
Figure 671100DEST_PATH_IMAGE074
the distance matrix
Figure 641330DEST_PATH_IMAGE067
The method comprises the following steps:
for any of the mobile devices
Figure 261798DEST_PATH_IMAGE075
To other said mobile devices
Figure 26492DEST_PATH_IMAGE076
Is in the distance matrix
Figure 935673DEST_PATH_IMAGE077
The value of the element(s) is,
Figure 257064DEST_PATH_IMAGE078
representing said distance matrix
Figure 184569DEST_PATH_IMAGE077
To (1) a
Figure 933213DEST_PATH_IMAGE079
Line and first
Figure 126428DEST_PATH_IMAGE073
The elements in the column, which are determined by the following expression:
Figure 438461DEST_PATH_IMAGE080
wherein
Figure 33521DEST_PATH_IMAGE081
To represent
Figure 890750DEST_PATH_IMAGE082
And
Figure 758212DEST_PATH_IMAGE083
the distance between them.
6. The method of claim 1, wherein the method is performed in a batch processThe transmitting mobile device
Figure 155826DEST_PATH_IMAGE084
And receiving mobile device
Figure 339814DEST_PATH_IMAGE085
Form a transceiving mobile device group
Figure 679529DEST_PATH_IMAGE086
The sending mobile device
Figure 581757DEST_PATH_IMAGE084
And receiving mobile device
Figure 720745DEST_PATH_IMAGE085
Data storage and data transmission are carried out between the two, and the data storage generates data storage energy consumption
Figure 539665DEST_PATH_IMAGE087
Said
Figure 738697DEST_PATH_IMAGE087
Is expressed as
Figure 128221DEST_PATH_IMAGE088
Wherein the content of the first and second substances,
Figure 585747DEST_PATH_IMAGE089
is shown as
Figure 744327DEST_PATH_IMAGE090
From the sending mobile device
Figure 363527DEST_PATH_IMAGE084
Transmission to receiving mobile device
Figure 771506DEST_PATH_IMAGE085
The data size of (2), and
Figure 302895DEST_PATH_IMAGE091
the sending mobile device
Figure 519244DEST_PATH_IMAGE084
The amount of data that is stored is of a size,
Figure 371662DEST_PATH_IMAGE092
represents the above-mentioned group
Figure 673462DEST_PATH_IMAGE093
Energy consumption of data storage;
the data transmission generates data transmission energy consumption
Figure 285840DEST_PATH_IMAGE094
Said
Figure 402700DEST_PATH_IMAGE094
Is expressed as
Figure 114436DEST_PATH_IMAGE095
Wherein
Figure 231427DEST_PATH_IMAGE096
Representing a receiving mobile device
Figure 116338DEST_PATH_IMAGE085
The received power of the antenna,
Figure 353284DEST_PATH_IMAGE097
representing data from the transmitting mobile device
Figure 501500DEST_PATH_IMAGE084
Is transmitted to the receiverMobile device
Figure 27159DEST_PATH_IMAGE085
The time taken for the process to be carried out,
Figure 981340DEST_PATH_IMAGE098
representing the receiving mobile device
Figure 761208DEST_PATH_IMAGE085
Receiving the transmitting mobile device
Figure 126330DEST_PATH_IMAGE084
The time taken for the transmitted data to be transmitted,
the data storage energy consumption
Figure 890018DEST_PATH_IMAGE087
And said data transmission energy consumption
Figure 382310DEST_PATH_IMAGE094
Summing total energy consumption of edge collaborative storage
Figure 531532DEST_PATH_IMAGE099
Said total energy consumption
Figure 818288DEST_PATH_IMAGE099
Is expressed as
Figure 318539DEST_PATH_IMAGE100
。
7. The method according to claim 1, wherein the edge collaborative storage policy comprises the steps of: data collection, request distribution, decision making, decision feedback, multiple interactions and data transmission.
8. An edge storage acceleration apparatus for cooperating with a mobile device, comprising:
the mobile device comprises a first establishing module, a second establishing module and a third establishing module, wherein the first establishing module is configured to establish an edge collaborative storage task for transmitting data between mobile devices;
a second building module configured to build an edge collaborative storage model based on the edge collaborative storage task, the edge collaborative storage model including: mobile device collection
Figure 817785DEST_PATH_IMAGE101
Sending mobile device
Figure 634562DEST_PATH_IMAGE084
Receiving mobile device
Figure 544750DEST_PATH_IMAGE085
And network topology
Figure 814188DEST_PATH_IMAGE102
The set of mobile devices
Figure 100813DEST_PATH_IMAGE101
Comprises at least two mobile devices;
a calculation module configured to calculate an A2CS acceleration algorithm based on the edge collaborative storage model, the A2CS acceleration algorithm including:
initial value
Figure 65157DEST_PATH_IMAGE103
、
Figure 693716DEST_PATH_IMAGE104
And
Figure 902980DEST_PATH_IMAGE105
the initial value represents the value of the current mobile device
Figure 540766DEST_PATH_IMAGE106
Deploying the A2CS plusInitial value of the fast algorithm, said initial value being expressed as
Figure 784666DEST_PATH_IMAGE107
Wherein
Figure 787388DEST_PATH_IMAGE108
Is shown and described
Figure 31418DEST_PATH_IMAGE106
The number of connected mobile devices is such that,
Figure 191004DEST_PATH_IMAGE109
to represent
Figure 836880DEST_PATH_IMAGE108
A dimension vector;
a stopping criterion that brings the A2CS acceleration algorithm to a viable solution and ensures that the A2CS acceleration algorithm converges quickly, the stopping criterion expressed as
Figure 10504DEST_PATH_IMAGE110
Wherein
Figure 725519DEST_PATH_IMAGE111
Is shown as
Figure 908370DEST_PATH_IMAGE112
The original residual at the time of the sub-iteration,
Figure 408752DEST_PATH_IMAGE113
is shown as
Figure 64862DEST_PATH_IMAGE112
The dual residual at the time of the sub-iteration,
Figure 221168DEST_PATH_IMAGE114
it is shown that the absolute tolerance is,
Figure 738868DEST_PATH_IMAGE115
indicating a relative tolerance;
optimization function
Figure 811866DEST_PATH_IMAGE020
For describing the total utility of the edge collaborative storage task, the optimization function is expressed as
Figure 592871DEST_PATH_IMAGE116
Wherein
Figure 751320DEST_PATH_IMAGE022
And
Figure 807132DEST_PATH_IMAGE023
is the weight coefficient of the weight of the image,
Figure 734637DEST_PATH_IMAGE024
indicating by a sending mobile device
Figure 420964DEST_PATH_IMAGE117
The size of the amount of data collected,
Figure 348600DEST_PATH_IMAGE024
and
Figure 457370DEST_PATH_IMAGE026
can be described as
Figure 786852DEST_PATH_IMAGE027
，
Figure 429099DEST_PATH_IMAGE028
Represents the aboveTransmitting mobile device
Figure 296561DEST_PATH_IMAGE029
The number of data copies of the collected data,
Figure 163017DEST_PATH_IMAGE030
which is indicative of the amount of data loss,
Figure 65114DEST_PATH_IMAGE031
represents the total energy consumption for the storage of the data,
Figure 155561DEST_PATH_IMAGE032
indicating transmitting mobile device
Figure 261051DEST_PATH_IMAGE033
The probability of data loss of (a) is,
Figure 711624DEST_PATH_IMAGE034
representing the sending mobile device
Figure 953381DEST_PATH_IMAGE029
The data transmission rate of (a) is,
Figure 932838DEST_PATH_IMAGE035
representing the receiving mobile device
Figure 56783DEST_PATH_IMAGE036
The rate of reception of the data of (c),
Figure 61779DEST_PATH_IMAGE037
representing adjacency matrices
Figure 672889DEST_PATH_IMAGE038
To (1) a
Figure 449346DEST_PATH_IMAGE039
Line and first
Figure 513117DEST_PATH_IMAGE040
The elements in the column are selected from the group,
Figure 321805DEST_PATH_IMAGE041
representing a distance matrix
Figure 803733DEST_PATH_IMAGE042
To (1) a
Figure 593834DEST_PATH_IMAGE039
Line and first
Figure 692371DEST_PATH_IMAGE040
The elements in the column are selected from the group,
Figure 835908DEST_PATH_IMAGE043
representing the power consumption for storing a unit of data,
Figure 188654DEST_PATH_IMAGE044
representing the sending mobile device
Figure 211974DEST_PATH_IMAGE029
The transmission power of the antenna is set to be,
Figure 809526DEST_PATH_IMAGE045
which represents the gain of the transmit antenna,
Figure 507485DEST_PATH_IMAGE046
which represents the gain of the receiving antenna,
Figure 291902DEST_PATH_IMAGE047
which represents the wavelength of the light emitted by the light source,
Figure 423806DEST_PATH_IMAGE048
representing the sending mobile device
Figure 824831DEST_PATH_IMAGE029
And the receiving mobile device
Figure 247854DEST_PATH_IMAGE036
The distance between the two or more of the two or more,
Figure 355618DEST_PATH_IMAGE049
representing a system loss factor independent of propagation;
a policy module configured to propose an edge collaborative storage policy based on the edge collaborative storage model and an A2CS acceleration algorithm;
an execution module configured to direct the edge collaborative storage task with the edge collaborative storage policy.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1 to 7 when executing the program.
CN202010252903.4A 2020-04-02 2020-04-02 Edge storage acceleration method, device and equipment for cooperative mobile equipment Active CN111158612B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010252903.4A CN111158612B (en) 2020-04-02 2020-04-02 Edge storage acceleration method, device and equipment for cooperative mobile equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010252903.4A CN111158612B (en) 2020-04-02 2020-04-02 Edge storage acceleration method, device and equipment for cooperative mobile equipment

Publications (2)

Publication Number Publication Date
CN111158612A CN111158612A (en) 2020-05-15
CN111158612B true CN111158612B (en) 2020-07-24

Family

ID=70567659

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010252903.4A Active CN111158612B (en) 2020-04-02 2020-04-02 Edge storage acceleration method, device and equipment for cooperative mobile equipment

Country Status (1)

Country Link
CN (1) CN111158612B (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110830520A (en) * 2020-01-13 2020-02-21 中国人民解放军国防科技大学 Robust and reliable edge storage method and system for Internet of things

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005157712A (en) * 2003-11-26 2005-06-16 Hitachi Ltd Remote copy network
US8612681B2 (en) * 2011-11-22 2013-12-17 Hitachi, Ltd Storage system, storage apparatus and method of controlling storage system
CN104573119B (en) * 2015-02-05 2017-10-27 重庆大学 Towards the Hadoop distributed file system storage methods of energy-conservation in cloud computing
CN107370802A (en) * 2017-07-10 2017-11-21 中国人民解放军国防科学技术大学 A kind of collaboration storage dispatching method based on alternating direction multiplier method
CN108121511B (en) * 2017-12-21 2022-05-27 北京猫盘技术有限公司 Data processing method, device and equipment in distributed edge storage system
CN110471621B (en) * 2019-07-29 2020-12-29 东南大学 Edge collaborative storage method for real-time data processing application
CN110443298B (en) * 2019-07-31 2022-02-15 华中科技大学 Cloud-edge collaborative computing-based DDNN and construction method and application thereof
CN110430266B (en) * 2019-08-06 2021-07-13 腾讯科技（深圳）有限公司 Edge cloud cooperative data transmission method, device, equipment and storage medium

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110830520A (en) * 2020-01-13 2020-02-21 中国人民解放军国防科技大学 Robust and reliable edge storage method and system for Internet of things

Also Published As

Publication number Publication date
CN111158612A (en) 2020-05-15

Similar Documents

Publication Publication Date Title
Xu et al. A computation offloading method over big data for IoT-enabled cloud-edge computing
Luo et al. Stability of cloud-based UAV systems supporting big data acquisition and processing
CN107766135B (en) Task allocation method based on particle swarm optimization and simulated annealing optimization in moving cloud
CN112422644B (en) Method and system for unloading computing tasks, electronic device and storage medium
CN111260076B (en) Block chain-based edge node training method, block chain and storage medium
Li et al. Distributed coordinated tracking control of multiple Euler–Lagrange systems by state and output feedback
Wang et al. Guaranteed cost consensus problems for second‐order multi‐agent systems
US20230281513A1 (en) Data model training method and apparatus
Li et al. Computation offloading strategy for improved particle swarm optimization in mobile edge computing
CN111158612B (en) Edge storage acceleration method, device and equipment for cooperative mobile equipment
Hao et al. Topology optimised fixed‐time consensus for multi‐UAV system in a multipath fading channel
Yang et al. Drone formation control via belief-correlated imitation learning
US20220197355A1 (en) Method for implementing power and/or heat generation control and electronic device therefor
CN114007231B (en) Heterogeneous unmanned aerial vehicle data unloading method and device, electronic equipment and storage medium
Liang et al. An effective resource scheduling model for edge cloud oriented AIoT
Shi et al. Multi-UAV-assisted computation offloading in DT-based networks: A distributed deep reinforcement learning approach
Meng et al. Edge computing task offloading method for load balancing and delay optimization
CN111541781A (en) Distributed cooperative task scheduling method for mobile equipment in dense network
CN117729113B (en) Network switching method of smart power grid and related equipment
Deb et al. Loop-the-Loops: Fragmented Learning Over Networks for Constrained IoT Devices
Santi et al. A resource management survey for mission critical and time critical applications in multi access edge computing
CN113326112B (en) Multi-unmanned aerial vehicle task unloading and migration method based on block coordinate descent method
Khani et al. Deep reinforcement learning‐based resource allocation in multi‐access edge computing
Lin et al. MADDPG-based Task Offloading and Resource Management for Edge System
CN115226130B (en) Multi-unmanned aerial vehicle data unloading method based on fairness perception and related equipment

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