CN113114714A - Energy-saving method and system for unloading large-scale tasks to 5G edge server - Google Patents

Energy-saving method and system for unloading large-scale tasks to 5G edge server Download PDF

Info

Publication number
CN113114714A
CN113114714A CN202011212941.3A CN202011212941A CN113114714A CN 113114714 A CN113114714 A CN 113114714A CN 202011212941 A CN202011212941 A CN 202011212941A CN 113114714 A CN113114714 A CN 113114714A
Authority
CN
China
Prior art keywords
task
edge server
mobile device
unloading
helper
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011212941.3A
Other languages
Chinese (zh)
Other versions
CN113114714B (en
Inventor
葛佳琦
徐高潮
付晓东
刘紫奇
李龙
刘鹏
李阳
金镇君
吴泰琦
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.)
Jilin University
Original Assignee
Jilin University
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 Jilin University filed Critical Jilin University
Priority to CN202011212941.3A priority Critical patent/CN113114714B/en
Publication of CN113114714A publication Critical patent/CN113114714A/en
Application granted granted Critical
Publication of CN113114714B publication Critical patent/CN113114714B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/56—Provisioning of proxy services
    • H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
    • H04W52/02—Power saving arrangements
    • H04W52/0209—Power saving arrangements in terminal devices
    • H04W52/0212—Power saving arrangements in terminal devices managed by the network, e.g. network or access point is master and terminal is slave
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention relates to an energy-saving method and system for unloading large-scale tasks to a 5G edge server, and belongs to the field of Internet of things and communication computing. Is a cellular-free data mobile device requesting to perform a large-scale complex task; the mobile equipment without cellular data selects an idle mobile equipment with a network as an assistant (helper) equipment, judges whether task unloading is needed or not according to a given initial cache strategy, and executes the task according to a judgment result. The invention provides a better energy-saving unloading method for a mobile device without cellular data and a network mobile device to execute large-scale complex tasks, solves the problem that the computing resources, the storage space and the electric quantity of the mobile device are limited, effectively saves the energy consumption of the whole system and reduces the unloading delay to a certain extent.

Description

Energy-saving method and system for unloading large-scale tasks to 5G edge server
Technical Field
The invention relates to the field of Internet of things and communication computing, in particular to a method and a system for energy-saving unloading of large-scale complex tasks from a network-free mobile device to a 5G edge server.
Background
With the development of mobile terminal devices of the internet of things and the continuous complication of device application programs, the defects of long transmission delay of mobile cloud computing and heavy load of a central server cannot further meet the user requirements, and a mobile edge computing concept is proposed for researchers. The advent of mobile edge computing has solved to some extent the problem of limited computing resources, storage space and power of mobile devices, and the tasks performed by mobile devices are more offloaded to mobile edge servers. However, with the landing of the fifth generation wireless system (5G), the access of large-scale intelligent devices inevitably brings spectrum resource shortage, and although the edge server carried by the 5G base station has large calculation resources and storage resources, a longer distance causes a larger time delay and energy consumption in the task offloading process, and cannot meet the requirements of users on service quality in practical application. Meanwhile, the unloading speed is continuously accelerated along with the improvement of the spectrum efficiency, higher requirements are provided for the performance of the mobile edge equipment under the coverage of the 5G base station, and the mobile edge equipment cannot execute large-scale complex calculation tasks due to the shortage of calculation resources, storage resources and electric quantity. Therefore, for 5G base stations and mobile devices covered by the base stations, it is the most interesting research content to reasonably allocate computing resources, reasonably utilize storage resources and maximally save energy consumption.
Existing task offloading studies have focused on interaction between networkable mobile devices, or cooperation of mobile devices with edge servers, and have temporarily failed to consider situations where a task cannot be handled by a mobile device on a network due to poor signal quality. Meanwhile, although many existing technologies attempt to solve the problem of excessive energy consumption of the 5G base station and achieve certain effects, the existing technologies mainly optimize computing and communication resources in the offloading process, and also have many researchers research tasks to optimize task caching strategies in the edge server, but all do not consider the overall optimization relation between mobile devices and between devices and the edge server in the whole offloading process, and have a large improvement space. In order to better achieve the aim, a better research approach is provided for the joint optimization of the task unloading capacity, the calculation and communication resource allocation and the caching strategy.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a method for unloading to a 5G edge server to save energy for executing a large-scale complex task by a mobile device without cellular data, which uses a mobile device with a network as a relay node to better stimulate cooperation among devices, and simultaneously jointly optimizes the task unloading amount, calculation and communication resources and a cache strategy of the 5G edge server in the unloading process, thereby effectively solving the problems of large energy consumption of the existing application in a 5G environment and insufficient electricity quantity and storage space of the mobile device, and greatly improving the time for executing the task by the mobile device without cellular data on the aspect of calculation delay.
To achieve the above object, the present invention provides a method for energy-saving offloading of large-scale tasks from a cellular-free data mobile device to a 5G edge server, comprising:
a cellular-free data mobile device requesting to perform a large-scale complex task;
the mobile device without the cellular data selects an idle mobile device with a network as an assistant (helper) device, judges whether task unloading is needed or not according to a given initial cache strategy, and executes a task according to a judgment result.
Further, the determining, according to a given initial caching policy, whether task unloading is required, and performing task execution according to a determination result, further includes:
if the task is not cached in the 5G edge server, task unloading is required;
if the task is cached in the 5G edge server, task unloading is not needed, and the 5G edge server directly performs calculation of the task and returns the result to the helper device and then to the mobile device without cellular data.
Further, if the task is cached in the 5G edge server, task offloading is required, which specifically includes:
partially offloading the task from the cellular data free mobile device to the corresponding helper device, wherein execution of remaining tasks of the cellular data free mobile device and offloading of a portion of the tasks to the helper device are performed concurrently; after the unloading is finished, the helper device performs secondary unloading of the partial tasks, continues to partially unload the task data to the 5G edge server, and at this time, the tasks are simultaneously executed by the cellular data-free mobile device, the helper device and the 5G edge server, and gradually returns results. Further, before the cellular data-less mobile device requests to perform a large-scale complex task, a model is built with the goal of minimizing the overall energy consumption of the mobile device, namely:
Figure BDA0002758198160000031
where a denotes a cache policy of each task, and is an M-dimensional vector, which may be denoted as a ═ a1,a2,…,aM}，amA caching policy representing a task m; vi,tAnd
Figure BDA0002758198160000032
respectively representing the local execution data volume of the mobile device without cellular data i and the data volume, V, executed by the corresponding helper device k in the t-th time slotiAnd VkRepresents Vi,tAnd
Figure BDA0002758198160000033
a vector solution of (a);
Figure BDA0002758198160000034
and
Figure BDA0002758198160000035
respectively representing the transmission power, p, provided by the cellular data-free mobile device i to the helper device k and the helper device k to the 5G edge server e in the t-th time sloti，pkIs its vector solution;
Figure BDA0002758198160000036
representing the mapping relation from the device i to the device k, and the solution vector is represented by mu; vmIs the total data size of task m, CmIndicating the period for which 1bit is to be executed, pwaitWhich is indicative of the waiting power,
Figure BDA0002758198160000037
and
Figure BDA0002758198160000038
respectively representing the calculated frequencies of the cellular data free mobile device i, the helper device k and the 5G edge server e,
Figure BDA0002758198160000039
and
Figure BDA00027581981600000310
represents the channel gain, BkAnd BeRepresenting channel bandwidths of the cellular data free mobile device i to the helper device k and the helper device k to the 5G edge server e; kappa is an effective capacitor and is determined by a CPU core; sigma2Representing noise interference, obtained according to the gaussian theorem.
Further, the determining, according to a given initial caching policy, whether task unloading is required, and performing task execution according to a determination result specifically includes: the method comprises 5 steps of problem decomposition, variable initialization, variable relaxation and SLSFP algorithm solution, and discrete variable and block coordinate descending alternative solution, and specifically comprises the following steps:
step one, problem decomposition: decomposing the original energy consumption target problem into two sub-problems according to the optimized discrete variable and continuous variable, wherein the two sub-problems comprise a sub-problem 1: optimizing the task offload data volume, computing resource allocation, and selection strategy of the cellular data-free mobile device to the helper device to minimize offload energy consumption; sub-problem 2: optimizing caching policies in the 5G edge servers to minimize offload energy consumption;
step two, initializing variables: before each time no cellular data mobile equipment requests a task, a caching strategy of the task at a 5G edge server is initialized, namely a is a0；
Step three, solving the relaxation variables and the SLQP algorithm: binary selection policy of the non-cellular data mobile device to the helper device
Figure BDA0002758198160000041
And (3) performing relaxation, converting the subproblem 1 into a perspective convex function by a mathematical method, and solving the subproblem 1 by an SLQP algorithm, namely:
Figure BDA0002758198160000042
wherein the constraint conditions are as follows:
Figure BDA0002758198160000043
Figure BDA0002758198160000044
Figure BDA0002758198160000045
Figure BDA0002758198160000046
Figure BDA0002758198160000047
Figure BDA0002758198160000048
wherein the content of the first and second substances,
Figure BDA0002758198160000051
represents a collection of the network-less mobile devices,
Figure BDA0002758198160000052
indicating a set of network mobile devices are idle; c1, C2, C3 indicate that the selection of the mobile device without cellular data i to the helper device k is one-to-one, C4 indicates that the mobile device without cellular data i executes locally, the helper device k executes, and the 5G edge server e executes data volume from 0 to VmMeanwhile, C5 indicates that the transmission power cannot exceed the maximum power of the transmission power, C6 indicates that the task execution time cannot exceed T; wherein
Figure BDA0002758198160000053
Represents the time of offloading of the cache at the 5G edge server, and
Figure BDA0002758198160000054
indicating the unloading time that the task is not cached to the 5G edge server, wherein P represents the number of time slots;
the subproblem 1 represents an energy consumption representation mode after a cache strategy is given, the first half section is a cache condition, and the energy consumption is the execution energy consumption of the 5G edge server; the second half is non-cache energy consumption, including the energy consumption of the non-cellular data mobile device, the hellper device and the 5G edge server for executing tasks and the transmission energy consumption in the unloading process;
step four, dispersing variables: and (3) solving the subproblem 2 according to the optimal unloading data volume obtained in the step three, the selection strategy among the devices and the transmission efficiency in the unloading process, and solving the energy consumption under the optimal caching strategy:
Figure BDA0002758198160000055
the constraints are as follows:
Figure BDA0002758198160000056
Figure BDA0002758198160000057
C9：0≤am≤1
wherein in C7
Figure BDA0002758198160000061
All the data are solved by subproblem 1, C8 and C9 indicate that the caching strategy is a number between 0 and 1, and the cached data volume cannot exceed the total capacity D;
step five, alternately solving the block coordinate descending: and (4) repeatedly solving the subproblems by adopting a block coordinate reduction method until the problem is converged, and selecting a mode of minimizing power consumption in the unloading process and obtaining the minimum energy consumption.
Therefore, the invention provides a better energy-saving unloading method for the non-cellular data mobile equipment and the non-network mobile equipment to execute the large-scale complex task, solves the problem of limited computing resources, storage space and electric quantity of the mobile equipment through the excitation cooperation among the mobile equipment and the interaction between the mobile equipment and the 5G base station, and simultaneously performs the joint optimization on the selection relation from the non-cellular data mobile equipment to the network mobile equipment in the task unloading and executing process, the cache strategy of the base station server, the task unloading data quantity and the computing resource distribution in the unloading process, thereby effectively saving the energy consumption of the whole system and reducing the unloading delay to a certain degree.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic diagram of a communication structure of a 5G edge-D2D system according to an embodiment of the present invention.
Fig. 2 is a flow chart of a specific unloading process in the embodiment of the invention.
Fig. 3 is a calculation flow chart of the energy saving offloading method according to the embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1: take a mobile edge computing system with a 5G base station supporting cached edge servers as an example. The system comprises a 5G base station, an edge server, a network mobile device and a non-cellular data mobile device, wherein the 5G edge server is in communication connection with the network mobile device through the base station. Aiming at the large-scale complex task executed by the mobile edge device without cellular data or poor signals, the invention provides a method for jointly optimizing energy-saving unloading from the mobile device without cellular data to the 5G edge server, as shown in figure 2.
A cellular-free data mobile device requesting to perform a large-scale complex task;
the mobile device without the cellular data selects an idle mobile device with a network as an assistant (helper) device, judges whether task unloading is needed or not according to a given initial cache strategy, and executes a task according to a judgment result.
Further, the determining, according to a given initial caching policy, whether task unloading is required, and performing task execution according to a determination result, further includes:
if the task is not cached in the 5G edge server, task unloading is required;
if the task is cached in the 5G edge server, task unloading is not needed, and the 5G edge server directly performs calculation of the task and returns the result to the helper device and then to the mobile device without cellular data.
Further, if the task is cached in the 5G edge server, task offloading is required, which specifically includes:
partially offloading the task from the cellular data free mobile device to the corresponding helper device, wherein execution of remaining tasks of the cellular data free mobile device and offloading of a portion of the tasks to the helper device are performed concurrently; after the unloading is finished, the helper device performs secondary unloading of the partial tasks, continues to partially unload the task data to the 5G edge server, and at this time, the tasks are simultaneously executed by the cellular data-free mobile device, the helper device and the 5G edge server, and gradually returns results.
By the method, the problems that the edge server is too high in energy consumption and the mobile equipment is insufficient in energy under the 5G environment can be effectively solved.
As another embodiment of the present invention, the method of power saving offload from a cellular data-less mobile device to a 5G edge server may further comprise:
(1.1) consider a 5G base station overlaid with N devices, and a network environment where the base station edge server supports multi-user caching, where I devices need to perform large-scale complex tasks (virtual reality gaming), and the device set is set to
Figure BDA0002758198160000071
The remaining idle devices
Figure BDA0002758198160000072
5G base station interaction may be performed using a network. Assuming that time is divided into time slots of size T, each device i,
Figure BDA0002758198160000073
any one intensive task can be requested at any time and the maximum delay that a user can accept is T. When the collection
Figure BDA0002758198160000074
When the device(s) in (1) requests a task, since the independent local computation cannot be completed within a prescribed time, it requires cooperation of incentives between the devices and assistance from the edge server.
(1.2) defining a set of computationally intensive and delay sensitive tasks M,
Figure BDA0002758198160000081
each task is composed of two elements, i.e. M ═ Vm,Cm) In which V ismData size (kilobytes), C, representing task mmIndicating the number of cycles used to execute a 1 bit. We consider the case of P slots, i.e., t ═ 1,2, …, P, at the beginning of each slot, device i,
Figure BDA0002758198160000082
starting to randomly request any one task M, wherein M belongs to M, and
∑m∈MPi(m)＝1。
(2) and (3) establishing a complex and delay-sensitive task problem model in the 5G network.
Modeling with the goal of minimizing the overall energy consumption of the mobile device, namely:
Figure BDA0002758198160000083
the objective function is to minimize the total energy consumption of the mobile device, and the energy consumption of the result feedback is negligible and is not considered in the present invention. Wherein
Figure BDA0002758198160000084
Representing the execution energy consumption if the task is cached at the 5G edge server,
Figure BDA0002758198160000085
and
Figure BDA0002758198160000086
respectively representing the local computational energy consumption of the cellular data-free mobile device i and the assisting device k,
Figure BDA0002758198160000087
represents the offloading transmission energy consumption of the task from the non-cellular data mobile device i to the assisting device k,
Figure BDA0002758198160000088
representing the transmission energy consumption of the secondary offloading of the task from the helper device k to the 5G edge server portion,
Figure BDA0002758198160000089
and the calculation energy consumption of the task executed on the 5G edge server after the task is unloaded to the 5G edge server is represented if the task is not cached in the 5G edge server.
(2.1) CPU computing capability of Mobile device i
Figure BDA0002758198160000091
The local computation delay of the tth slot can be expressed as:
Figure BDA0002758198160000092
(2.2) the local computational energy consumption can be expressed as:
Figure BDA0002758198160000093
where κ is the effective capacitance associated with the CPU, Vi,tIndicating the size of the amount of data when the mobile device i performs task m locally at the t-th slot.
(2.3) the local computation delay when the helper device k performs device i to offload tasks can be expressed as:
Figure BDA0002758198160000094
(2.4) the helper device calculates the energy consumption as:
Figure BDA0002758198160000095
wherein
Figure BDA0002758198160000096
Represents the mapping relationship of device i to device k in time slot t,
Figure BDA0002758198160000097
for the CPU computing power of the helper device k,
Figure BDA0002758198160000098
indicating the amount of data for the t-th slot helper device k to perform the offload task m from device i.
(2.5) edge calculated delays are:
Figure BDA0002758198160000099
(2.6) the mobile device is idle during the calculation period of the edge server, but the energy consumption consumed by the waiting time is considered to be proportional to the time, and the waiting power consumed by the waiting time device is pwaitWhen the server executes, the device consumes power
Figure BDA00027581981600000910
Comprises the following steps:
Figure BDA00027581981600000911
(3) in addition to computing time and energy consumption, task offloading requires transport, and the present invention contemplates two-stage transport, transport from local devices to helper devices and transport of helper devices to edge servers.
(3.1) unloading process of task from device i to helper device: suppose that the t-th slot device i provides a transmission power of
Figure BDA0002758198160000101
Then the communication rate between mobile i and mobile k in the t-th slot can be expressed as:
Figure BDA0002758198160000102
wherein sigma2Representing the noise interference of the channel, BkIn order to be the bandwidth of the channel,
Figure BDA0002758198160000103
channel gain from mobile device to edge server for tth time slot
(3.2) the amount of data to be executed locally is set to Vi,tDelay of mobile device i to k
Figure BDA0002758198160000104
Can be expressed as:
Figure BDA0002758198160000105
required energy consumption
Figure BDA0002758198160000106
Comprises the following steps:
Figure BDA0002758198160000107
(3.3) offloading process of task from helper device to 5G base station: the offloading rate of the helper device k to the base station is:
Figure BDA0002758198160000108
wherein
Figure BDA0002758198160000109
The power allocated to the kth device for handling the off-load task of device i for the tth slot edge server,
Figure BDA00027581981600001010
the channel gain from the mobile to the edge server for the T-th slot varies every T time, but remains constant in each slot, σ2Representing the noise interference of the channel.
(3.4) the amount of data offloaded to edge calculation is
Figure BDA00027581981600001011
Then 5G edge server to helper mobile k transmission delay
Figure BDA00027581981600001012
Can be expressed as:
Figure BDA00027581981600001013
the power consumed during transmission can be expressed as:
Figure BDA00027581981600001014
(4) the strong storage space of the base station is utilized for data caching, so that the data transmission delay can be effectively reduced, and the energy consumption is reduced. Suppose the edge server has a cache capacity of D, amRepresenting caching policiesIf amIf the result is too small for the whole execution process, the result can be ignored, at this time, the task execution time is the execution time on the edge server, and the consumed energy consumption is the execution energy consumption of the edge server, that is, the task is cached in the server, that is, the returned result is too small to be considered, the task execution time is the execution time on the edge server, and the consumed energy consumption is the execution energy consumption of the edge
Figure BDA0002758198160000111
On the contrary, if amIf 0 indicates that the task is not cached currently, local computation needs to be performed, and the helper device and the edge server need to perform jointly. The computation delay in the case of a task without buffering is:
Figure BDA0002758198160000112
(5) according to the above definition, the present invention performs the joint optimization of resource allocation, task segmentation and caching strategy with the aim of minimizing the energy consumption of the mobile device, that is
Figure BDA0002758198160000113
The relevant constraints are as follows:
Figure BDA0002758198160000114
Figure BDA0002758198160000115
Figure BDA0002758198160000116
Figure BDA0002758198160000121
Figure BDA0002758198160000122
Figure BDA0002758198160000123
the objective function is to minimize the total energy consumption of the mobile device, where a, Vi,Vk,pk,piμ is a solution of the problem, a represents a caching strategy of each task, is a vector with M dimensions, and can be expressed as a ═ a1,a2,…,aM}；ViAnd VkThe local execution data volume and the corresponding helper data execution data volume of each device respectively, each device has certain preference for the requested task, so we assume ViAnd VkAssociated with each device only, represented by vectors of dimensions I and K, respectively, i.e.
Figure BDA0002758198160000124
piMeans that the transmission power provided by mobile I to mobile K in the t-th time slot is represented by a vector of dimensions I x K, and pkThe transmission power of the mobile equipment K for the t time slot when unloading the task of the equipment I to the edge server is represented as an I x K dimensional vector;
Figure BDA0002758198160000125
representing a one-to-one mapping of I to K, which is a 0,1 vector matrix of I x K.
Constraint C1 indicates that the mapping between mobile devices is a binary variable, and constraints C2 and C3 ensure that the mapping from mobile device i to mobile device k is one-to-one; c4 indicates that the caching policy is a binary variable; constraint C5 ensures that the total amount of data for the tasks cached in the server cannot exceed capacity D; c6 indicates that the task execution time of all devices cannot exceed the time constraint T, regardless of whether there is a cache in the edge server; constraint C7 indicates that the local, helper, and edge execution data volumes are all 0 to Vm(ii) a Constraint C8 indicates that the transmit power of mobile device i to device k and the transmit power of mobile device k to the edge server are at0 and maximum power p of the mobile devicemaxIn the meantime. Since the resources are reallocated beginning at each time segment, all constraints should be satisfied for P time segments.
As another embodiment of the present invention, whether task unloading is required is determined according to a given initial caching policy, and a task is executed according to a determination result, which may further include 5 steps of problem decomposition, initializing variables, relaxing variables and slslsrp algorithm solution, and discrete variable and block coordinate descent alternative solution, as shown in fig. 3:
step one, problem decomposition: decomposing the original energy consumption target problem into two sub-problems according to the optimized discrete variable and continuous variable, wherein the two sub-problems comprise a sub-problem 1: optimizing the task offload data volume, computing resource allocation, and selection strategy of the cellular data-free mobile device to the helper device to minimize offload energy consumption; sub-problem 2: optimizing caching policies in the 5G edge servers to minimize offload energy consumption;
step two, initializing variables: before each time no cellular data mobile equipment requests a task, a caching strategy of the task at a 5G edge server is initialized, namely a is a0；
Step three, solving the relaxation variables and the SLQP algorithm: binary selection policy of the non-cellular data mobile device to the helper device
Figure BDA0002758198160000131
And (3) performing relaxation, converting the subproblem 1 into a perspective convex function by a mathematical method, and solving the subproblem 1 by an SLQP algorithm, namely:
Figure BDA0002758198160000132
wherein the constraint conditions are as follows:
Figure BDA0002758198160000133
Figure BDA0002758198160000134
Figure BDA0002758198160000135
Figure BDA0002758198160000136
Figure BDA0002758198160000137
Figure BDA0002758198160000138
wherein the content of the first and second substances,
Figure BDA0002758198160000139
represents a collection of the network-less mobile devices,
Figure BDA00027581981600001310
indicating a set of network mobile devices are idle; c1, C2, C3 indicate that the selection of the mobile device without cellular data i to the helper device k is one-to-one, C4 indicates that the mobile device without cellular data i executes locally, the helper device k executes, and the 5G edge server e executes data volume from 0 to VmMeanwhile, C5 indicates that the transmission power cannot exceed the maximum power of the transmission power, C6 indicates that the task execution time cannot exceed T; wherein
Figure BDA00027581981600001311
Represents the time of offloading of the cache at the 5G edge server, and
Figure BDA0002758198160000141
indicating the unloading time that the task is not cached to the 5G edge server, wherein P represents the number of time slots;
the subproblem 1 represents an energy consumption representation mode after a cache strategy is given, the first half section is a cache condition, and the energy consumption is the execution energy consumption of the 5G edge server; the second half is non-cache energy consumption, including the energy consumption of the non-cellular data mobile device, the hellper device and the 5G edge server for executing tasks and the transmission energy consumption in the unloading process;
step four, dispersing variables: and (3) solving the subproblem 2 according to the optimal unloading data volume obtained in the step three, the selection strategy among the devices and the transmission efficiency in the unloading process, and solving the energy consumption under the optimal caching strategy:
Figure BDA0002758198160000142
the constraints are as follows:
Figure BDA0002758198160000143
Figure BDA0002758198160000144
C9：0≤am≤1
wherein in C7
Figure BDA0002758198160000145
All the data are solved by subproblem 1, C8 and C9 indicate that the caching strategy is a number between 0 and 1, and the cached data volume cannot exceed the total capacity D;
step five, alternately solving the block coordinate descending: and (4) repeatedly solving the subproblems by adopting a block coordinate reduction method until the problem is converged, and selecting a mode of minimizing power consumption in the unloading process and obtaining the minimum energy consumption.
It should be noted that the algorithm of the present invention is applicable to, but not limited to, the case of a single 5G base station, a single server and multiple mobile devices, and is also applicable to the case of multiple 5G base stations, multiple servers and multiple mobile devices, and besides the calculation offload between the base station server and the mobile device, the algorithm can also be applied to the incentive cooperation between the base station server and the mobile device, so as to encourage the communication between the devices and reduce the offload energy consumption to the utmost.
It will be appreciated that although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (6)

1. An energy-saving method for unloading large-scale tasks to a 5G edge server is characterized in that:
a cellular-free data mobile device requesting to perform a large-scale complex task;
the mobile device without the cellular data selects an idle mobile device with a network as an assistant (helper) device, judges whether task unloading is needed or not according to a given initial cache strategy, and executes a task according to a judgment result.
2. The method according to claim 1, wherein the determining whether task offloading is required according to a given initial caching policy and performing task execution according to a determination result further comprises:
if the task is not cached in the 5G edge server, task unloading is required;
if the task is cached in the 5G edge server, task unloading is not needed, and the 5G edge server directly performs calculation of the task and returns the result to the helper device and then to the mobile device without cellular data.
3. The method according to claim 2, wherein if the task is not cached in the 5G edge server, the task offloading is required, specifically comprising:
partially offloading the task from the cellular data free mobile device to the corresponding helper device, wherein execution of remaining tasks of the cellular data free mobile device and offloading of a portion of the tasks to the helper device are performed concurrently; after the unloading is finished, the helper device performs secondary unloading of the partial tasks, continues to partially unload the task data to the 5G edge server, and at this time, the tasks are simultaneously executed by the cellular data-free mobile device, the helper device and the 5G edge server, and gradually returns results.
4. The method of claim 1, wherein before the cellular-free data mobile device requests to perform a large-scale complexity task, the model is established with the goal of minimizing the overall power consumption of the mobile device, namely:
Figure FDA0002758198150000011
where a denotes a cache policy of each task, and is an M-dimensional vector, which may be denoted as a ═ a1，a2，...，aM}，amA caching policy representing a task m; vi，tAnd
Figure FDA0002758198150000012
respectively representing the local execution data volume of the mobile device without cellular data i and the data volume, V, executed by the corresponding helper device k in the t-th time slotiAnd VkRepresents Vi，tAnd
Figure FDA0002758198150000021
a vector solution of (a);
Figure FDA0002758198150000022
and
Figure FDA0002758198150000023
respectively indicating the i direction of the mobile equipment without cellular data in the t time slotThe helper device k, the transmission power p provided by the helper device k to the 5G edge server ei，pkIs its vector solution;
Figure FDA0002758198150000024
representing the mapping relation from the device i to the device k, and the solution vector is represented by mu; vmIs the total data size of task m, CmIndicating the period for which 1bit is to be executed, pwaitWhich is indicative of the waiting power,
Figure FDA0002758198150000025
and
Figure FDA0002758198150000026
respectively representing the calculated frequencies of the cellular data free mobile device i, the helper device k and the 5G edge server e,
Figure FDA0002758198150000027
and
Figure FDA0002758198150000028
represents the channel gain, BkAnd BeRepresenting channel bandwidths of the cellular data free mobile device i to the helper device k and the helper device k to the 5G edge server e; kappa is an effective capacitor and is determined by a CPU core; sigma2Representing noise interference, obtained according to the gaussian theorem.
5. The method according to claim 4, wherein the determining whether task offloading is required according to a given initial caching policy and performing task execution according to a determination result specifically includes: the method comprises 5 steps of problem decomposition, variable initialization, variable relaxation and SLSFP algorithm solution, and discrete variable and block coordinate descending alternative solution, and specifically comprises the following steps:
step one, problem decomposition: decomposing the original energy consumption target problem into two sub-problems according to the optimized discrete variable and continuous variable, wherein the two sub-problems comprise a sub-problem 1: optimizing the task offload data volume, computing resource allocation, and selection strategy of the cellular data-free mobile device to the helper device to minimize offload energy consumption; sub-problem 2: optimizing caching policies in the 5G edge servers to minimize offload energy consumption;
step two, before the mobile equipment without cellular data requests the task each time, initializing a cache strategy of the task in the 5G edge server, namely a is a0；
Step three, solving the relaxation variables and the SLQP algorithm: binary selection policy of the non-cellular data mobile device to the helper device
Figure FDA0002758198150000029
And (3) performing relaxation, converting the subproblem 1 into a perspective convex function by a mathematical method, and solving the subproblem 1 by an SLQP algorithm, namely:
Figure FDA00027581981500000210
wherein the constraint conditions are as follows:
Figure FDA0002758198150000031
Figure FDA0002758198150000032
Figure FDA0002758198150000033
Figure FDA0002758198150000034
Figure FDA0002758198150000035
Figure FDA0002758198150000036
wherein the content of the first and second substances,
Figure FDA0002758198150000037
represents a collection of the network-less mobile devices,
Figure FDA0002758198150000038
indicating a set of network mobile devices are idle; c1, C2, C3 indicate that the selection of the mobile device without cellular data i to the helper device k is one-to-one, C4 indicates that the mobile device without cellular data i executes locally, the helper device k executes, and the 5G edge server e executes data volume from 0 to VmMeanwhile, C5 indicates that the transmission power cannot exceed the maximum power of the transmission power, C6 indicates that the task execution time cannot exceed T; wherein
Figure FDA0002758198150000039
Represents the time of offloading of the cache at the 5G edge server, and
Figure FDA00027581981500000310
indicating the unloading time that the task is not cached to the 5G edge server, wherein P represents the number of time slots;
the sub-problem 1 represents a representation mode of energy consumption after a cache strategy is given, the first half section is a cache condition, and the energy consumption is that the second half section of the energy consumption executed by the 5G edge server is non-cache energy consumption, and comprises the energy consumption of the non-cellular data mobile device, the hellper device and the 5G edge server for executing tasks and the transmission energy consumption in an unloading process;
step four, dispersing variables: and (3) solving the subproblem 2 according to the optimal unloading data volume obtained in the step three, the selection strategy among the devices and the transmission efficiency in the unloading process, and solving the energy consumption under the optimal caching strategy:
Figure FDA00027581981500000311
the constraints are as follows:
Figure FDA0002758198150000041
Figure FDA0002758198150000042
C9：0≤am≤1
wherein in C7
Figure FDA0002758198150000043
All the data are solved by subproblem 1, C8 and C9 indicate that the caching strategy is a number between 0 and 1, and the cached data volume cannot exceed the total capacity D;
step five, alternately solving the block coordinate descending: and (4) repeatedly solving the subproblems by adopting a block coordinate reduction method until the problem is converged, and selecting a mode of minimizing power consumption in the unloading process and obtaining the minimum energy consumption.
6. An energy-saving system for unloading large-scale tasks to a 5G edge server is characterized in that
The method comprises the following steps: a base station, no cellular data mobile device, a network mobile device and a 5G edge server; the 5G edge server is in communication connection with the network mobile equipment through the base station; and
the cellular-free data mobile device requesting performance of a large-scale complexity task;
the network-free mobile equipment selects an idle network-equipped mobile equipment as an assistant (helper) equipment, and judges whether the task is cached in a 5G edge server according to a given initial caching strategy;
if the task is not cached in the 5G edge server, optimizing and partially unloading the task from the cellular data-free mobile equipment to the corresponding helper equipment, wherein the execution of the rest tasks of the cellular data-free mobile equipment and the unloading of the partial task to the helper equipment are simultaneously executed; after the unloading is finished, the helper device performs secondary unloading of the partial tasks, continues to partially unload the task data to the 5G edge server, and at the moment, the tasks are simultaneously executed by the non-network mobile device, the helper device and the 5G edge server, and gradually returns results;
if the task is cached in the 5G edge server, task unloading is not needed, and the 5G edge server directly performs calculation of the task and returns the result to the helper device and then to the mobile device without cellular data.
CN202011212941.3A 2020-11-03 2020-11-03 Energy-saving method and system for unloading large-scale tasks to 5G edge server Active CN113114714B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011212941.3A CN113114714B (en) 2020-11-03 2020-11-03 Energy-saving method and system for unloading large-scale tasks to 5G edge server

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011212941.3A CN113114714B (en) 2020-11-03 2020-11-03 Energy-saving method and system for unloading large-scale tasks to 5G edge server

Publications (2)

Publication Number Publication Date
CN113114714A true CN113114714A (en) 2021-07-13
CN113114714B CN113114714B (en) 2022-03-01

Family

ID=76709004

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011212941.3A Active CN113114714B (en) 2020-11-03 2020-11-03 Energy-saving method and system for unloading large-scale tasks to 5G edge server

Country Status (1)

Country Link
CN (1) CN113114714B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113553160A (en) * 2021-08-03 2021-10-26 上海紫邦电气技术有限公司 Task scheduling method and system for edge computing node of artificial intelligence Internet of things
CN113938957A (en) * 2021-12-06 2022-01-14 太平洋电信股份有限公司 Calculation distribution method and system of network edge equipment
CN114866557A (en) * 2022-05-06 2022-08-05 东北林业大学 Computing resource sharing platform and method under edge computing and V2V converged network

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108934002A (en) * 2018-07-18 2018-12-04 广东工业大学 A kind of task unloading algorithm based on D2D communication cooperation
CN109819046A (en) * 2019-02-26 2019-05-28 重庆邮电大学 A kind of Internet of Things virtual computing resource dispatching method based on edge cooperation
CN110062026A (en) * 2019-03-15 2019-07-26 重庆邮电大学 Mobile edge calculations resources in network distribution and calculating unloading combined optimization scheme
CN110941667A (en) * 2019-11-07 2020-03-31 北京科技大学 Method and system for calculating and unloading in mobile edge calculation network
EP3648436A1 (en) * 2018-10-29 2020-05-06 Commissariat à l'énergie atomique et aux énergies alternatives Method for clustering cache servers within a mobile edge computing network
CN111148134A (en) * 2019-12-19 2020-05-12 南京大学 Multi-user multi-task unloading method based on mobile edge calculation
CN111182582A (en) * 2019-12-30 2020-05-19 东南大学 Multitask distributed unloading method facing mobile edge calculation
CN111240701A (en) * 2019-12-31 2020-06-05 重庆大学 Task unloading optimization method for end-edge-cloud collaborative computing
CN111556516A (en) * 2020-05-25 2020-08-18 南京邮电大学 Distributed wireless network task cooperative distribution method facing delay and energy efficiency sensitive service
CN111585916A (en) * 2019-12-26 2020-08-25 国网辽宁省电力有限公司电力科学研究院 LTE electric power wireless private network task unloading and resource allocation method based on cloud edge cooperation

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108934002A (en) * 2018-07-18 2018-12-04 广东工业大学 A kind of task unloading algorithm based on D2D communication cooperation
EP3648436A1 (en) * 2018-10-29 2020-05-06 Commissariat à l'énergie atomique et aux énergies alternatives Method for clustering cache servers within a mobile edge computing network
CN109819046A (en) * 2019-02-26 2019-05-28 重庆邮电大学 A kind of Internet of Things virtual computing resource dispatching method based on edge cooperation
CN110062026A (en) * 2019-03-15 2019-07-26 重庆邮电大学 Mobile edge calculations resources in network distribution and calculating unloading combined optimization scheme
CN110941667A (en) * 2019-11-07 2020-03-31 北京科技大学 Method and system for calculating and unloading in mobile edge calculation network
CN111148134A (en) * 2019-12-19 2020-05-12 南京大学 Multi-user multi-task unloading method based on mobile edge calculation
CN111585916A (en) * 2019-12-26 2020-08-25 国网辽宁省电力有限公司电力科学研究院 LTE electric power wireless private network task unloading and resource allocation method based on cloud edge cooperation
CN111182582A (en) * 2019-12-30 2020-05-19 东南大学 Multitask distributed unloading method facing mobile edge calculation
CN111240701A (en) * 2019-12-31 2020-06-05 重庆大学 Task unloading optimization method for end-edge-cloud collaborative computing
CN111556516A (en) * 2020-05-25 2020-08-18 南京邮电大学 Distributed wireless network task cooperative distribution method facing delay and energy efficiency sensitive service

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
PENG LIU，GAOCHAO XU等: "Jointly Optimized Energy-Minimal Resource Allocation in Cache-Enhanced Mobile Edge Computing Systems", 《IEEE ACCESS》 *
YANG LI, GAOCHAO XU等: "Communication and computation cooperation in wireless network for mobile edge computing", 《IEEE ACCESS》 *
YANG LI, GAOCHAO XU等: "Jointly Optimizing Helpers Selection and Resource Allocation in D2D Mobile Edge Computing", 《 2020 IEEE WIRELESS COMMUNICATIONS AND NETWORKING CONFERENCE (WCNC)》 *
吴柳青等: "基于边-端协同的任务卸载资源分配联合优化算法", 《电信科学》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113553160A (en) * 2021-08-03 2021-10-26 上海紫邦电气技术有限公司 Task scheduling method and system for edge computing node of artificial intelligence Internet of things
CN113938957A (en) * 2021-12-06 2022-01-14 太平洋电信股份有限公司 Calculation distribution method and system of network edge equipment
CN114866557A (en) * 2022-05-06 2022-08-05 东北林业大学 Computing resource sharing platform and method under edge computing and V2V converged network
CN114866557B (en) * 2022-05-06 2024-01-23 东北林业大学 Computing resource sharing platform and method under edge computing and V2V fusion network

Also Published As

Publication number Publication date
CN113114714B (en) 2022-03-01

Similar Documents

Publication Publication Date Title
CN113114714B (en) Energy-saving method and system for unloading large-scale tasks to 5G edge server
CN111586696B (en) Resource allocation and unloading decision method based on multi-agent architecture reinforcement learning
CN113950066B (en) Single server part calculation unloading method, system and equipment under mobile edge environment
CN108809695B (en) Distributed uplink unloading strategy facing mobile edge calculation
CN111010684B (en) Internet of vehicles resource allocation method based on MEC cache service
CN111132191B (en) Method for unloading, caching and resource allocation of joint tasks of mobile edge computing server
CN110941667A (en) Method and system for calculating and unloading in mobile edge calculation network
CN110798849A (en) Computing resource allocation and task unloading method for ultra-dense network edge computing
Li et al. An energy-aware task offloading mechanism in multiuser mobile-edge cloud computing
CN111163143B (en) Low-delay task unloading method for mobile edge calculation
CN110719641B (en) User unloading and resource allocation joint optimization method in edge computing
CN110928691A (en) Traffic data-oriented edge collaborative computing unloading method
CN113286317B (en) Task scheduling method based on wireless energy supply edge network
CN112188551B (en) Computation migration method, computation terminal equipment and edge server equipment
CN111200831B (en) Cellular network computing unloading method fusing mobile edge computing
CN113835878A (en) Resource allocation method and device, computer equipment and storage medium
CN112969163A (en) Cellular network computing resource allocation method based on self-adaptive task unloading
Wu et al. A mobile edge computing-based applications execution framework for Internet of Vehicles
Gu et al. Energy-efficient computation offloading and transmit power allocation scheme for mobile edge computing
CN114064294B (en) Dynamic resource allocation method and system in mobile edge computing environment
CN111158893B (en) Task unloading method, system, equipment and medium applied to fog computing network
CN113364860B (en) Method and system for joint calculation resource allocation and unloading decision in MEC
CN109818788B (en) Secondary-mode optimization-based calculation resource allocation method in edge cache C-RAN
CN112486685A (en) Computing task allocation method and device of power Internet of things and computer equipment
CN116347522A (en) Task unloading method and device based on approximate computation multiplexing under cloud edge cooperation

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