CN113286317A - Task scheduling method based on wireless energy supply edge network - Google Patents

Task scheduling method based on wireless energy supply edge network Download PDF

Info

Publication number
CN113286317A
CN113286317A CN202110447548.0A CN202110447548A CN113286317A CN 113286317 A CN113286317 A CN 113286317A CN 202110447548 A CN202110447548 A CN 202110447548A CN 113286317 A CN113286317 A CN 113286317A
Authority
CN
China
Prior art keywords
user terminal
task
user
time
execution
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
CN202110447548.0A
Other languages
Chinese (zh)
Other versions
CN113286317B (en
Inventor
朱琦
朱科宇
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.)
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing University of Posts and Telecommunications
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 Nanjing University of Posts and Telecommunications filed Critical Nanjing University of Posts and Telecommunications
Priority to CN202110447548.0A priority Critical patent/CN113286317B/en
Publication of CN113286317A publication Critical patent/CN113286317A/en
Application granted granted Critical
Publication of CN113286317B publication Critical patent/CN113286317B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/0215—Traffic management, e.g. flow control or congestion control based on user or device properties, e.g. MTC-capable devices
    • H04W28/0221—Traffic management, e.g. flow control or congestion control based on user or device properties, e.g. MTC-capable devices power availability or consumption
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/0231—Traffic management, e.g. flow control or congestion control based on communication conditions
    • H04W28/0236—Traffic management, e.g. flow control or congestion control based on communication conditions radio quality, e.g. interference, losses or delay
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/08—Load balancing or load distribution
    • 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

Abstract

The invention discloses a task scheduling method based on a wireless energy supply edge network, which comprises the following steps: assuming that a total of N user terminals, defining the unloading decision of a user, selecting the task execution sequence of the user terminals unloaded to the edge server, and selecting the local user task execution sequence; each user terminal firstly performs energy acquisition and task processing, and the special energy station supplies energy to all edge user terminals in a wireless energy transmission mode; each user terminal carries out local calculation tasks or unloads the tasks to an edge server for calculation, and an improved Johnson algorithm is adopted to solve the unloading decision and task scheduling of the user terminal; solving the optimal wireless power supply time by adopting a golden section method; and (5) alternately and iteratively using, and converging to obtain an optimal solution. The invention carries out unloading decision and task scheduling on each user task request, greatly reduces the computational complexity and realizes the goal of minimizing the time delay of the user terminal of the whole system.

Description

Task scheduling method based on wireless energy supply edge network
Technical Field
The invention relates to a task scheduling method based on a wireless energy supply edge network, and belongs to the technical field of wireless energy supply communication networks.
Background
In recent years, with the development of technologies such as internet of things, artificial intelligence and virtual reality, high-energy-consumption computing-intensive services are continuously increased, and the conflict between computing-intensive applications and resource-limited mobile computing systems brings unprecedented challenges to the development of future mobile services. To meet this challenge, mobile cloud computing technology MCC is typically employed, offloading computing tasks on mobile terminals to resource-rich remote clouds for completion. However, the conventional MCC method has disadvantages of long delay and low reliability caused by data transmission through a wide area network. In recent years, a mobile edge computing MEC, which can provide cloud computing capability in the vicinity of mobile users, has been proposed as one of the key technologies of 5G. Offloading the user's computing tasks to a nearby MEC server, i.e., mobile edge computing offloading, is considered a promising solution to address the above challenges. Compared with the traditional MCC scheme, the edge computing can realize lower delay and higher reliability and becomes a research hotspot, the technology researches the computation unloading of users and the allocation of computation resources and communication resources, the computation unloading is divided into a partial unloading mode and a complete unloading mode, and the unloading target server also comprises an edge server and a remote cloud server. Meanwhile, due to the development of wireless energy transmission technology (WPT), the wireless energy-supplying MEC network is receiving more and more attention, and the wireless energy-supplying communication network is a special energy supply station deployed near user equipment to perform wireless energy transmission for nearby users, and the key point is the time slot allocation of wireless energy supply, so that edge users can perform tasks by maximally utilizing collected energy.
Wireless power supply technology and computing offloading technology in mobile edge computing are of great importance, and some research has been carried out, but most of the research is that the time for wireless power supply is fixed, and meanwhile, the computing power of an edge server is assumed to be infinite, and the computing delay of an MEC server is ignored. For example, documents [16] l.huang, s.bi and y. -j.a.zhang, "Deep discovery Learning for Online Computing of streaming in Wireless power Mobile-Edge Computing Networks," in IEEE Transactions on Mobile Computing, vol.19, No.11, pp.2581-2593,1nov.2020. have studied the maximum Offloading efficiency based on Wireless energy-providing users in Edge Computing Networks, but neglected the transmission delay of tasks during Offloading of tasks, which is not in practical terms. Documents [21] y.mao, j.zhang, and k.b.letaief, "Joint task of flow scheduling and transmit power allocation for mobile-edge computing systems," in proc.ieee Wireless communication.net.conf. (WCNC),2017, pp.1-6 consider only the task scheduling case when unloaded by the user, and do not pay attention to the local computation case of the user.
Disclosure of Invention
The invention aims to solve the technical problem that in the prior art, the calculation delay or the transmission delay of a task is ignored during the calculation of a mobile edge server under the condition of wireless energy supply, and provides a task scheduling method based on a wireless energy supply edge network.
The invention specifically adopts the following technical scheme to solve the technical problems:
a task scheduling method based on a wireless energy supply edge network comprises the following steps:
step 1, initialization: assuming a total of N user terminals, each having a task to be computed, defining the user's offload decision as
Figure BDA0003037506470000021
Wherein xi1 denotes whether the user terminal i offloads the task to the MEC server for execution; and, define
Figure BDA0003037506470000022
Indicating the order of execution of tasks selected to be offloaded to the user terminal of the edge server, wherein CNcRepresents the Nth unloaded task and defines
Figure BDA0003037506470000023
Indicating the order of execution of user tasks selected locally, where LNlRepresents the Nth locally executed task; the above mentioned requirements
Figure BDA0003037506470000024
And N is Nc + Nl, which respectively represent the number of users offloaded to the MEC server and the number of users executing locally;
step 2, each user terminal firstly collects energy and then performs task processing, in the energy collection stage with wireless energy supply time tau, the special energy station supplies energy to all edge user terminals in a wireless energy transmission mode, and then the ith user terminal collects energy ei；
Step 3, each user terminal utilizes energy obtained by wireless energy supply to perform local calculation tasks or unload the tasks to an edge server for calculation, and under the given wireless energy supply time, an improved Johnson algorithm is adopted to solve unloading decisions and task scheduling of the user terminal;
step 4, solving the unloading decision and task scheduling of the user terminal obtained according to the step 3, and solving the optimal wireless power supply time by adopting a golden section method;
step 5, alternately and iteratively using the step 3 and the step 4 until the unloading decision and the task scheduling are not updated any more, and then converging to obtain an optimal solution to obtain a user task execution sequence selected by the user terminal in the local
Figure BDA0003037506470000025
And selecting the execution order of tasks to be offloaded to the MEC server
Figure BDA0003037506470000026
And an optimal wireless energizing time tau.
Further, as a preferred technical solution of the present invention, the energy e collected by the ith user terminal in step 2iExpressed as: e.g. of the typei＝μPgiτ, where μ ∈ (0, 1) denotes the energy conversion efficiency, P denotes the emission power of the energy-specific station during the energy collection phase, giRepresenting the channel gain between the dedicated energy station and the ith edge user terminal.
Further, as a preferable technique of the present inventionA technical scheme, wherein in the step 3, each user terminal carries out the calculation time delay required by the local calculation task
Figure BDA0003037506470000031
Expressed as:
Figure BDA0003037506470000032
wherein D isiRepresenting the number of CPU cycles required for single task execution; c represents the coefficient of energy consumed per CPU cycle, EiRepresents the energy consumption available for task execution;
computation time delay required for each user terminal to unload tasks to edge servers for computation
Figure BDA0003037506470000033
Expressed as:
Figure BDA0003037506470000034
wherein f isedgeIs the processing power of the MEC server;
further, as a preferred technical solution of the present invention, in step 3, an improved johnson algorithm is adopted to solve the offloading decision and task scheduling of the user terminal, including constructing an optimization objective function of the system model, which is expressed as:
Figure BDA0003037506470000035
Figure BDA0003037506470000036
Figure BDA0003037506470000037
Figure BDA0003037506470000038
C4:0＜τ≤τmax
Figure BDA0003037506470000039
Figure BDA00030375064700000310
C7:Nc+Nl＝N
wherein the content of the first and second substances,
Figure BDA00030375064700000311
and
Figure BDA00030375064700000312
respectively represent locally executing user terminals according to
Figure BDA00030375064700000313
Task completion time of collective execution and offload to MEC server according to
Figure BDA00030375064700000314
Task completion time of collective execution, fiFor the computational power of the user terminal i,
Figure BDA00030375064700000315
indicating the maximum computational power of the user terminal,
Figure BDA00030375064700000316
which represents the transmit power of the user terminal i,
Figure BDA00030375064700000317
representing the maximum transmit power, τ, of the usermaxIndicating the longest time of wireless power.
Further, as a preferred technical solution of the present invention, the step 3 adopts a modified johnson algorithm to solve, and includes the steps of:
the user terminals are divided into two disjoint subsets D and E according to the transmission time of the user terminal tasks and the time of execution at the MEC server, which is expressed as follows:
Figure BDA0003037506470000041
Figure BDA0003037506470000042
wherein the content of the first and second substances,
Figure BDA0003037506470000043
representing the transmission time of the unloading data from the user terminal to the edge server;
Figure BDA0003037506470000044
representing the calculation time delay executed by the calculation task of the user terminal i in the MEC server;
the user terminals in the subset D are arranged in an ascending order according to the transmission time of the tasks, and the user terminals in the subset E are arranged in a descending order according to the execution time of the tasks; adding the first user terminal in the subset D to the set of user terminals
Figure BDA0003037506470000045
In (3), adding the last user terminal in the subset E to the set of user terminals
Figure BDA0003037506470000046
Performing the following steps;
the completion time of the locally executed task in the first i user terminals
Figure BDA0003037506470000047
Expressed as:
Figure BDA0003037506470000048
wherein x isjIndicating whether the user terminal j unloads the task to the MEC server for execution;
Figure BDA0003037506470000049
representing the calculation time delay needed by the local execution selected by the user terminal j;
delaying the completion of task execution of the user terminal unloaded from the first i user terminals
Figure BDA00030375064700000410
Expressed as:
Figure BDA00030375064700000411
Figure BDA00030375064700000412
wherein the content of the first and second substances,
Figure BDA00030375064700000413
representing the transmission time of the unloaded data from the user terminal j to the MEC server;
Figure BDA00030375064700000414
representing the task transmission time of the user terminal unloaded from the first i user terminals;
Figure BDA00030375064700000415
representing the calculation time delay executed by the calculation task of the user terminal j in the MEC server;
Figure BDA00030375064700000416
representing the task transmission time of the user terminal unloaded from the first j user terminals;
Figure BDA00030375064700000417
representing the first j-1 user terminalsThe task transmission time of the unloaded user terminal;
then, the set L is calculated according to the above two formulas*The first user terminal in (1)
Figure BDA00030375064700000418
Completion time of the locally executed task
Figure BDA00030375064700000419
And set C*The first user terminal in (1)
Figure BDA00030375064700000420
Task completion delay for medium-offload user terminals
Figure BDA00030375064700000421
Allocating the user terminals in the subsets D and E according to the time of local and uninstall calculation until the allocation of the user in one of the subsets is completed; adding the remaining unallocated user terminals into the set M, and completely allocating the user terminals in the set M according to the time of local and uninstalled calculation; finally, local user terminal set is obtained
Figure BDA0003037506470000051
Set the user terminal therein to xi0, set of offloaded user terminals
Figure BDA0003037506470000052
Set the user terminal therein to xi＝1。
By adopting the technical scheme, the invention can produce the following technical effects:
(1) the method of the invention aims at the problem of insufficient energy consumption and computational efficiency of the mobile terminal by adopting a wireless energy supply technology and an edge computing unloading technology for a wireless energy supply edge computing network. Under a multi-user binary computation unloading mode, an optimization problem of minimizing the maximum time delay of system execution is established by jointly optimizing wireless energy supply time, unloading decisions of edge users and task scheduling; the invention aims at minimizing the maximum completion time delay of the system user task, carries out unloading decision and task scheduling on each user task request, and can reasonably distribute the unloading decision and task scheduling of edge users on the premise of lower complexity.
(2) The user task scheduling method under the wireless energy supply edge network considers the optimal time of wireless energy supply, and simultaneously adopts a pipeline task scheduling strategy aiming at the unloaded user, so that the transmission delay of the next task is overlapped with the calculation delay of the previous task as far as possible, and the delay of the whole unloading process is minimized. And finally, the time delay of the whole system is minimized through the alternate iterative solution of the wireless energy supply time delay and the task unloading decision. The invention greatly reduces the calculation complexity, improves the user satisfaction and realizes the goal of optimizing the system time delay.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention.
Fig. 2 is a schematic view of a scenario of the present invention applied to a wireless power supply edge network.
FIG. 3 is a diagram illustrating user task offloading according to the present invention.
FIG. 4 is a graph of the convergence of the algorithm of the present invention.
Fig. 5 is a diagram comparing system delays according to the present invention.
Detailed Description
The following describes embodiments of the present invention with reference to the drawings.
Referring to fig. 1, the present invention relates to a task scheduling method based on a wireless energy supply edge network, and a scenario of applying the method to the wireless energy supply edge network is shown in fig. 2, and the method is composed of a dedicated energy station providing energy service, a plurality of user terminals having wireless charging function, and a base station configured with an MEC server. Assuming that all devices are equipped with a single antenna and each edge user terminal is equipped with a limited capacity rechargeable battery, each device has a task that requires calculations. The edge server and the special energy station provide computing service and energy service for the edge user terminal, and the edge user terminal performs local execution or unloads tasks needing computation to the edge server for execution by using the collected energy. Therefore, the task scheduling method based on the wireless energy supply edge network specifically comprises the following steps:
step 1, initialization: assuming a total of N user terminals, each having a task to be computed, defining the user's offload decision as
Figure BDA0003037506470000061
Wherein xi{0,1} represents whether the user terminal offloads the task to the MEC server for execution, and if the user terminal i offloads the task to the MEC server for execution, xi1, otherwise xi＝0；
And, define
Figure BDA0003037506470000062
Indicating the order of execution of tasks selected to be offloaded to the user terminal of the edge server, wherein CNcRepresents the Nth unloaded task and defines
Figure BDA0003037506470000063
Indicating the order of execution of user tasks selected locally, where LNlRepresents the Nth locally executed task; the above mentioned requirements
Figure BDA0003037506470000064
And N is Nc + Nl, which respectively represent the number of users offloaded to the MEC server and the number of users executing locally; wherein the MEC server has a single-core CPU processor and performs processing in the order of first-come-first-executed each time, and further, in order to enable the execution of user tasks to satisfy
Figure BDA0003037506470000065
Assuming that the MEC server has a task cache unit large enough to cache unloaded tasks that are not executing.
And, assuming that each ue needs to collect energy first and then perform task processing, in the energy collection stage with duration τ, the dedicated energy station supplies energy to all edge ues in a wireless energy transmission manner, and in the task offloading stage, the offloaded ues perform uplink task offloading in a time division multiplexing manner by using the collected energy. If the task of the user terminal is executed locally, the time delay is the time of task calculation; if the user terminal selects to unload to the edge server for execution, the time delay comprises task unloading time, task calculating time and task return time.
Step 2, each user terminal firstly collects energy and then performs task processing, in the energy collection stage with wireless energy supply time tau, the special energy station supplies energy to all edge user terminals in a wireless energy transmission mode, and then the ith user terminal collects energy eiThe method comprises the following steps:
offloading the computing tasks to the MEC server for processing may reduce latency and energy consumption compared to local computing, but transferring task data may consume additional latency and energy consumption (i.e., communication latency and energy consumption). As shown in fig. 3, the present invention considers that the wireless energy transmission and the user task transmission occupy the same frequency band, and therefore cannot be performed simultaneously, and the task offloading of the user terminal adopts a time division multiplexing form. According to the Shann-Hartley theorem, the communication model between the ue and the bs can be defined as:
Figure BDA0003037506470000071
where W represents the channel bandwidth, hiRepresenting the channel gain between the user terminal i and the base station,
Figure BDA0003037506470000072
representing the transmission power, σ, of the user terminal i2Representing gaussian white noise.
Within the wireless energy supply time tau, all edges are subjected toThe user carries out wireless energy transmission, so the energy e collected by the ith user terminaliCan be expressed as:
ei＝μPgiτ (2)
wherein, mu epsilon (0, 1) represents energy conversion efficiency, P represents the emission power of the energy special station in the energy collection stage, and giRepresenting the channel gain between the dedicated energy station and the ith edge user terminal.
Step 3, each user terminal performs local computation tasks by using energy obtained by wireless energy supply or offloads the tasks to an edge server for computation, and under the given wireless energy supply time, an improved johnson algorithm is adopted to solve the offloading decision and task scheduling of the user terminal, which is specifically as follows:
step 3.1, constructing a system model
Defining a computing task of a user terminal i as an array
Figure BDA0003037506470000073
Wherein
Figure BDA0003037506470000074
Indicating the size of the task calculation input data, DiIndicating the number of CPU cycles required for the execution of a single task,
Figure BDA0003037506470000075
the size of the response data after the task calculation is shown, and the time delay calculation is discussed in two cases.
(1) When the user terminal i selects local execution, the user terminal selects the calculation delay d required by the local executioni lCan be expressed as:
Figure BDA0003037506470000076
Figure BDA0003037506470000077
substituting equation (3) into equation (4) yields:
Figure BDA0003037506470000078
wherein f isiFor the computing power of the user terminal i, i.e. the number of CPU cycles running per unit time, and the user terminal CPU cycle is satisfied
Figure BDA0003037506470000079
Figure BDA00030375064700000710
Representing the maximum computing power of the user terminal, c representing the coefficient of energy consumed per CPU cycle, EiRepresenting the energy consumption that can be used for task execution.
(2) When the user terminal i selects to unload the task to the MEC for execution, the transmission time of the unloaded data from the user terminal to the MEC server can be obtained according to the communication model
Figure BDA00030375064700000711
Comprises the following steps:
Figure BDA0003037506470000081
user terminal i transmit power
Figure BDA0003037506470000082
Comprises the following steps:
Figure BDA0003037506470000083
the following binding formulae (1), (6) and (7) can be obtained:
Figure BDA0003037506470000084
from the above formula, it can be found that it is difficult to find the expression from the above formula without using the approximate expression
Figure BDA0003037506470000085
The present invention adopts a first-order Taylor expansion to express the above formula, so that
Figure BDA0003037506470000086
Representing the maximum energy that the user terminal i can use to offload a task,
Figure BDA0003037506470000087
representing the maximum transmit power of the user, then:
Figure BDA0003037506470000088
then the task propagation delay for the first order taylor expansion can be expressed as:
Figure BDA0003037506470000089
assume the processing capacity of the MEC server is fedgeCalculating time delay executed by the calculating task of the user terminal i in the MEC server
Figure BDA00030375064700000810
Can be expressed as:
Figure BDA00030375064700000811
step 3.2, constructing an optimized objective function of the system model
The aim of the research of the invention is that a user utilizes the energy obtained by wireless energy supply to process tasks, and the optimal unloading decision and task scheduling are obtained by combining local calculation and MEC server calculation with the aim of minimizing the execution delay of the tasks.
If the task is selected to be executed locally, the locally executed task can be executed simultaneously on different user equipments, and the completion time of the locally executed user task in the first i user terminals can be expressed as:
Figure BDA00030375064700000812
wherein x isjIndicating whether the user terminal j unloads the task to the MEC server for execution;
Figure BDA00030375064700000813
indicating the computational delay required for the user terminal j to choose to execute locally.
If the user chooses to offload a task to an MEC server for execution, then the following two conditions must be met: first, all previous and current task data has been transmitted; second, the processor of the MEC has free time to perform new tasks. Therefore, the present invention can represent the execution delay of the user unloaded from the first i users as:
Figure BDA0003037506470000091
Figure BDA0003037506470000092
wherein the content of the first and second substances,
Figure BDA0003037506470000093
representing the transmission time of the unloaded data from the user terminal j to the MEC server;
Figure BDA0003037506470000094
representing the task transmission time of the user terminal unloaded from the first i user terminals;
Figure BDA0003037506470000095
meter for indicating user terminal jCalculating the calculation time delay of the task executed in the MEC server;
Figure BDA0003037506470000096
representing the task transmission time of the user terminal unloaded from the first j user terminals;
Figure BDA0003037506470000097
indicating the task transmission time of the user terminal offloaded from the first j-1 user terminals.
For the user terminal selected to be unloaded, the execution delay is divided into transmission delay and calculation delay, and because the execution delay of the former task and the transmission delay of the latter task have the overlapping part of the delay, the scheduling problem of the classic flow shop is solved.
Thus, the optimization problem of joint computation offloading, task scheduling and wireless power supply of the present invention can be expressed as:
Figure BDA0003037506470000098
wherein
Figure BDA0003037506470000099
And
Figure BDA00030375064700000910
respectively represent locally executing user terminals according to
Figure BDA00030375064700000911
Task completion time of collective execution and offload to MEC server according to
Figure BDA00030375064700000912
The task completion time of the set execution can be obtained by the calculation of the equations (12) and (14), respectively, and taumaxIndicating the longest time of wireless power. The above expression represents the time delay consumption of the whole system when the user task is completely executed. C1 indicates that the user's task is a two-valued full offload type, and C2 and C3 respectively indicate the maximum of the local deviceComputing power and maximum transmission power, C4 represents power supply time constraint of the wireless power supplier to ensure the reasonability of the wireless power supplier, C5 and C6 represent user terminals which are executed according to the set sequence of the respective pairs and have no repeated execution, and C7 ensures that the tasks of all users can be successfully completed.
Step 3.3, adopting an improved Johnson algorithm to solve the unloading decision and task scheduling of the user terminal, wherein the method comprises the following steps:
for the problem, the problem is decomposed into sub-problems of an upper layer and a lower layer, the user terminal unloading decision and the task scheduling under the given power supply time delay are firstly solved, and then the optimal wireless power supply time delay under the given unloading decision and the task scheduling is solved. Firstly, by utilizing the idea of the Johnson algorithm, the invention provides an improved Johnson algorithm to solve the complete unloading decision and task scheduling of multiple users. The Johnson algorithm is used as a common algorithm for scheduling a flow shop and is very suitable for scheduling user tasks selected for unloading, but locally executed users exist in the method and are difficult to solve by directly adopting the Johnson algorithm, so that the method improves the original Johnson algorithm. The users can be divided into two disjoint subsets D and E according to the transmission time of the user tasks and the time of execution at the MEC, which is expressed as follows:
Figure BDA0003037506470000101
Figure BDA0003037506470000102
the users in subset D execute before subset E, and the core idea of this algorithm is to maximize the CPU utilization of the MEC. According to this improved johnson algorithm, all users can be assigned to local execution and MEC server execution based on the offload latency and execution latency of user tasks. The Johnson algorithm adopted by the invention comprises the following specific steps:
respectively calculating the energy collected by each user terminal according to the initial wireless energy supply time tau, and then obtaining two disjoint subsets D and E according to the formula (16) and the formula (17);
the user terminals in the subset D are arranged in an ascending order according to the transmission time of the tasks, and the user terminals in the subset E are arranged in a descending order according to the execution time of the tasks; adding the first user terminal in the subset D to the set of user terminals C*In (3), the last user terminal in the subset E is added to the set of user terminals L*Performing the following steps;
then, the calculation is performed according to the equations (12) and (14)
Figure BDA0003037506470000103
I.e. the set L is calculated according to the above two formulas*Time of completion of the locally performed task in the first user terminal
Figure BDA0003037506470000104
And set C*Task completion delay of the offloaded user terminal of the first user terminal
Figure BDA0003037506470000105
Wherein
Figure BDA0003037506470000106
A set of representations L*Is determined by the first user terminal in the group,
Figure BDA0003037506470000107
representation set C*A first user terminal of; allocating the user terminals in the subsets D and E according to the time of local and uninstall calculation until the allocation of one set of users is completed; adding the remaining unallocated user terminals into the set M, and completely allocating the user terminals in the set M according to the time of local and uninstalled calculation; finally, local user terminal set is obtained
Figure BDA0003037506470000108
Set the user terminal therein to x i0, set of offloaded user terminals
Figure BDA0003037506470000109
Set the user terminal therein to xi＝1。
And 4, solving the unloading decision and task scheduling of the user terminal obtained according to the step 3, and solving the optimal wireless power supply time by adopting a golden section method.
The user's offload decision and scheduling order can be solved by the above-described modified johnson algorithm. The optimal wireless energy supply time is solved through given unloading decision and task scheduling sequence.
For problem P1, given the offload decision and scheduling order of the user, we can rewrite:
Figure BDA0003037506470000111
the problem P2 is a downward convex function with respect to time τ, and can be solved by using a common one-dimensional search algorithm. Therefore, the invention adopts the golden section method to solve, and the algorithm is commonly used for the unguided one-dimensional search problem and has higher solving efficiency. The method comprises the following specific steps: according to the unloading decision and task scheduling sequence of the user terminal solved by the improved Johnson algorithm, a one-dimensional function about the wireless power supply time tau is obtained, and the size of the wireless power supply time tau is continuously adjusted through the golden section ratio of 0.618 until the time of the whole system is minimum, so that the optimal wireless power supply time tau is obtained.
Step 5, the step 3 and the step 4 are alternately used in an iterative way until the unloading decision and the task scheduling are not updated any more, which indicates that the algorithm has converged to the optimal solution, so that the execution sequence of the user tasks selected by the user terminal of the method to be in the local can be obtained
Figure BDA0003037506470000112
And selecting an order of task execution for offloading to an edge server
Figure BDA0003037506470000113
And an optimal wireless energizing time tau.
The invention discusses the problem of user unloading decision and task scheduling based on a wireless energy supply edge computing network. The wireless energy supply time and user unloading decision are jointly considered, an optimization problem with the aim of minimizing the execution delay of system users is provided, and then a two-layer iterative solution scheme is provided. The upper layer is used for solving the problems of unloading decision and task scheduling of users, an improved Johnson suboptimal algorithm is provided for solving, the lower layer adopts a golden section method to solve the optimal wireless energy supply time, and two layers of alternative iteration solve the optimization problem.
As shown in FIG. 4, the method proposed by the present invention has a fast convergence speed, and tends to be stable around 14 times. The experimental result shows that the algorithm provided by the invention is obviously improved compared with other strategies in the aspect of reducing the system time delay, and the algorithm has higher convergence speed.
As shown in fig. 5, the comparison conditions of the four methods in the aspect of time delay are given, the method of the present invention is obviously superior to other three comparison algorithms, and the demand for resources is higher and higher with the increasing number of user terminals; the method of the invention can simultaneously carry out calculation on multiple tasks in different servers, greatly reduce the calculation waiting time between the tasks and effectively reduce the utility function value of the user terminal under the condition that communication resources and calculation resources are limited. On the premise of wireless power supply, the method can reasonably distribute the unloading decision and task scheduling of the edge user on the premise of lower complexity.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention.

Claims (5)

1. A task scheduling method based on a wireless energy supply edge network is characterized by comprising the following steps:
step 1,Initialization: assuming a total of N user terminals, each having a task to be computed, defining the user's offload decision as
Figure FDA0003037506460000011
Wherein xi1 denotes whether the user terminal i offloads the task to the MEC server for execution; and, define
Figure FDA0003037506460000012
Indicating the order of execution of tasks selected to be offloaded to the user terminal of the edge server, wherein CNcRepresents the Nth unloaded task and defines
Figure FDA0003037506460000013
Indicating the order of execution of user tasks selected locally, where LNlRepresents the Nth locally executed task; wherein
Figure FDA0003037506460000014
And N is Nc + Nl, which respectively represent the number of users offloaded to the MEC server and the number of users executing locally;
step 2, each user terminal firstly carries out energy collection and then carries out task processing, in the energy collection stage of the wireless energy supply time tau, the special energy station supplies energy to all edge user terminals in a wireless energy transmission mode, and then the ith user terminal collects energy ei；
Step 3, each user terminal utilizes energy obtained by wireless energy supply to perform local calculation tasks or unload the tasks to an edge server for calculation, and under the given wireless energy supply time, an improved Johnson algorithm is adopted to solve unloading decisions and task scheduling of the user terminal;
step 4, solving the unloading decision and task scheduling of the user terminal obtained according to the step 3, and solving the optimal wireless power supply time by adopting a golden section method;
step 5, step 3 and stepStep 4, the alternative iteration is used until the unloading decision and the task scheduling are not updated any more, at the moment, the convergence obtains the optimal solution, and the execution sequence of the user tasks selected by the user terminal in the local area is obtained
Figure FDA0003037506460000015
And selecting the execution order of tasks to be offloaded to the MEC server
Figure FDA0003037506460000016
And an optimal wireless energizing time tau.
2. The task scheduling method based on the wireless energy supply edge network as claimed in claim 1, wherein the energy e collected by the ith user terminal in the step 2iExpressed as: e.g. of the typei＝μPgiτ, where μ ∈ (0, 1) denotes the energy conversion efficiency, P denotes the emission power of the energy-specific station during the energy collection phase, giRepresenting the channel gain between the dedicated energy station and the ith edge user terminal.
3. The task scheduling method based on the wireless energy supply edge network as claimed in claim 1, wherein in the step 3, each user terminal performs the calculation delay required for the local calculation task
Figure FDA0003037506460000017
Expressed as:
Figure FDA0003037506460000018
wherein D isiRepresenting the number of CPU cycles required for single task execution; c represents the coefficient of energy consumed per CPU cycle, EiRepresents the energy consumption available for task execution;
computation time delay required for each user terminal to unload tasks to edge servers for computation
Figure FDA0003037506460000021
Expressed as:
Figure FDA0003037506460000022
wherein f isedgeIs the processing power of the MEC server.
4. The method for task scheduling based on wireless energy supply edge network according to claim 1, wherein the step 3 adopts a modified johnson algorithm to solve the offloading decision and task scheduling of the user terminal, and includes constructing an optimization objective function of a system model, which is expressed as:
P1:
Figure FDA0003037506460000023
s.t.C1:
Figure FDA0003037506460000024
C2:
Figure FDA0003037506460000025
C3:
Figure FDA0003037506460000026
C4:0＜τ≤τmax
C5:
Figure FDA0003037506460000027
C6:
Figure FDA0003037506460000028
C7:Nc+Nl＝N
wherein the content of the first and second substances,
Figure FDA0003037506460000029
and
Figure FDA00030375064600000210
respectively represent locally executing user terminals according to
Figure FDA00030375064600000211
Task completion time of collective execution and offload to MEC server according to
Figure FDA00030375064600000212
Task completion time of collective execution, fiFor the computational power of the user terminal i,
Figure FDA00030375064600000213
indicating the maximum computational power of the user terminal,
Figure FDA00030375064600000214
which represents the transmit power of the user terminal i,
Figure FDA00030375064600000215
representing the maximum transmit power, τ, of the usermaxIndicating the longest time of wireless power.
5. The method for task scheduling based on wireless energy supply edge network according to claim 1, wherein said step 3 adopts a modified johnson algorithm to solve, comprising the steps of:
the user terminals are divided into two disjoint subsets D and E according to the transmission time of the user terminal tasks and the time of execution at the MEC server, which is expressed as follows:
Figure FDA00030375064600000216
Figure FDA00030375064600000217
wherein the content of the first and second substances,
Figure FDA00030375064600000218
representing the transmission time of the unloading data from the user terminal to the edge server;
Figure FDA00030375064600000219
representing the calculation time delay executed by the calculation task of the user terminal i in the MEC server;
the user terminals in the subset D are arranged in an ascending order according to the transmission time of the tasks, and the user terminals in the subset E are arranged in a descending order according to the execution time of the tasks; adding the first user terminal in the subset D to the set of user terminals C*In (3), the last user terminal in the subset E is added to the set of user terminals L*Performing the following steps;
the completion time of the locally executed task in the first i user terminals
Figure FDA0003037506460000031
Expressed as:
Figure FDA0003037506460000032
wherein x isjIndicating whether the user terminal j unloads the task to the MEC server for execution;
Figure FDA0003037506460000033
representing the calculation time delay needed by the local execution selected by the user terminal j;
delaying the completion of task execution of the user terminal unloaded from the first i user terminals
Figure FDA0003037506460000034
Expressed as:
Figure FDA0003037506460000035
Figure FDA0003037506460000036
wherein the content of the first and second substances,
Figure FDA0003037506460000037
representing the transmission time of the unloaded data from the user terminal j to the MEC server;
Figure FDA0003037506460000038
representing the task transmission time of the user terminal unloaded from the first i user terminals;
Figure FDA0003037506460000039
representing the calculation time delay executed by the calculation task of the user terminal j in the MEC server;
Figure FDA00030375064600000310
representing the task transmission time of the user terminal unloaded from the first j user terminals;
Figure FDA00030375064600000311
representing the task transmission time of the user terminal unloaded from the first j-1 user terminals;
then, the set L is calculated according to the above two formulas*The first user terminal in (1)
Figure FDA00030375064600000312
Completion time of the locally executed task
Figure FDA00030375064600000313
And set C*The first user terminal in (1)
Figure FDA00030375064600000314
Task completion delay for medium-offload user terminals
Figure FDA00030375064600000315
Allocating the user terminals in the subsets D and E according to the time of local and uninstall calculation until the allocation of one set of users is completed; adding the remaining unallocated user terminals into the set M, and completely allocating the user terminals in the set M according to the time of local and uninstalled calculation; finally, local user terminal set is obtained
Figure FDA00030375064600000316
Set the user terminal therein to xi0, set of offloaded user terminals
Figure FDA00030375064600000317
Set the user terminal therein to xi＝1。
CN202110447548.0A 2021-04-25 2021-04-25 Task scheduling method based on wireless energy supply edge network Active CN113286317B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110447548.0A CN113286317B (en) 2021-04-25 2021-04-25 Task scheduling method based on wireless energy supply edge network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110447548.0A CN113286317B (en) 2021-04-25 2021-04-25 Task scheduling method based on wireless energy supply edge network

Publications (2)

Publication Number Publication Date
CN113286317A true CN113286317A (en) 2021-08-20
CN113286317B CN113286317B (en) 2023-07-28

Family

ID=77277309

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110447548.0A Active CN113286317B (en) 2021-04-25 2021-04-25 Task scheduling method based on wireless energy supply edge network

Country Status (1)

Country Link
CN (1) CN113286317B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114051205A (en) * 2021-11-08 2022-02-15 南京大学 Edge optimization method based on reinforcement learning dynamic multi-user wireless communication scene
CN114513855A (en) * 2022-02-23 2022-05-17 齐鲁工业大学 Edge computing unloading decision and resource allocation method based on wireless energy-carrying communication
CN114915627A (en) * 2022-04-15 2022-08-16 浙江工业大学 FDMA (frequency division multiple Access) -access-based wireless energy supply edge computing network unloading decision method
CN115334669A (en) * 2022-07-29 2022-11-11 西安邮电大学 Non-orthogonal multiple access and energy collection assisted resource allocation method in cacheable network
WO2024065903A1 (en) * 2022-09-29 2024-04-04 福州大学 Joint optimization system and method for computation offloading and resource allocation in multi-constraint-edge environment

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109814951A (en) * 2019-01-22 2019-05-28 南京邮电大学 The combined optimization method of task unloading and resource allocation in mobile edge calculations network
CN111447619A (en) * 2020-03-12 2020-07-24 重庆邮电大学 Joint task unloading and resource allocation method in mobile edge computing network
KR20210026171A (en) * 2019-08-29 2021-03-10 인제대학교 산학협력단 Multi-access edge computing based Heterogeneous Networks System
CN112492626A (en) * 2020-12-07 2021-03-12 南京邮电大学 Method for unloading computing task of mobile user

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109814951A (en) * 2019-01-22 2019-05-28 南京邮电大学 The combined optimization method of task unloading and resource allocation in mobile edge calculations network
KR20210026171A (en) * 2019-08-29 2021-03-10 인제대학교 산학협력단 Multi-access edge computing based Heterogeneous Networks System
CN111447619A (en) * 2020-03-12 2020-07-24 重庆邮电大学 Joint task unloading and resource allocation method in mobile edge computing network
CN112492626A (en) * 2020-12-07 2021-03-12 南京邮电大学 Method for unloading computing task of mobile user

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
张海波;荆昆仑;刘开健;贺晓帆;: "车联网中一种基于软件定义网络与移动边缘计算的卸载策略", 电子与信息学报, no. 03 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114051205A (en) * 2021-11-08 2022-02-15 南京大学 Edge optimization method based on reinforcement learning dynamic multi-user wireless communication scene
CN114051205B (en) * 2021-11-08 2022-09-13 南京大学 Edge optimization method based on reinforcement learning dynamic multi-user wireless communication scene
CN114513855A (en) * 2022-02-23 2022-05-17 齐鲁工业大学 Edge computing unloading decision and resource allocation method based on wireless energy-carrying communication
CN114513855B (en) * 2022-02-23 2024-04-19 齐鲁工业大学 Edge computing unloading decision and resource allocation method based on wireless energy-carrying communication
CN114915627A (en) * 2022-04-15 2022-08-16 浙江工业大学 FDMA (frequency division multiple Access) -access-based wireless energy supply edge computing network unloading decision method
CN114915627B (en) * 2022-04-15 2024-04-02 浙江工业大学 Wireless energy supply edge computing network unloading decision method based on FDMA access
CN115334669A (en) * 2022-07-29 2022-11-11 西安邮电大学 Non-orthogonal multiple access and energy collection assisted resource allocation method in cacheable network
CN115334669B (en) * 2022-07-29 2023-05-16 西安邮电大学 Resource allocation method assisted by non-orthogonal multiple access and energy collection in cacheable network
WO2024065903A1 (en) * 2022-09-29 2024-04-04 福州大学 Joint optimization system and method for computation offloading and resource allocation in multi-constraint-edge environment

Also Published As

Publication number Publication date
CN113286317B (en) 2023-07-28

Similar Documents

Publication Publication Date Title
WO2022121097A1 (en) Method for offloading computing task of mobile user
CN107995660B (en) Joint task scheduling and resource allocation method supporting D2D-edge server unloading
Bi et al. Joint optimization of service caching placement and computation offloading in mobile edge computing systems
CN113286317A (en) Task scheduling method based on wireless energy supply edge network
CN111586720B (en) Task unloading and resource allocation combined optimization method in multi-cell scene
CN111930436B (en) Random task queuing unloading optimization method based on edge calculation
CN110928654B (en) Distributed online task unloading scheduling method in edge computing system
CN108809695B (en) Distributed uplink unloading strategy facing mobile edge calculation
CN111132191B (en) Method for unloading, caching and resource allocation of joint tasks of mobile edge computing server
CN110493360B (en) Mobile edge computing unloading method for reducing system energy consumption under multiple servers
CN111372314A (en) Task unloading method and task unloading device based on mobile edge computing scene
CN109151864B (en) Migration decision and resource optimal allocation method for mobile edge computing ultra-dense network
CN110096362B (en) Multitask unloading method based on edge server cooperation
CN107682443A (en) Joint considers the efficient discharging method of the mobile edge calculations system-computed task of delay and energy expenditure
CN112105062B (en) Mobile edge computing network energy consumption minimization strategy method under time-sensitive condition
CN112000481B (en) Task unloading method for maximizing computing capacity of D2D-MEC system
CN109246761B (en) Unloading method based on alternating direction multiplier method considering delay and energy consumption
CN108964817A (en) A kind of unloading of heterogeneous network combined calculation and resource allocation methods
CN109951873A (en) A kind of Internet of Things mist calculate in information asymmetry do not know lower task unloading mechanism
CN111915142B (en) Unmanned aerial vehicle auxiliary resource allocation method based on deep reinforcement learning
CN114567895A (en) Method for realizing intelligent cooperation strategy of MEC server cluster
CN112860429A (en) Cost-efficiency optimization system and method for task unloading in mobile edge computing system
Tan et al. Joint offloading and resource allocation based on UAV-assisted mobile edge computing
CN113821346B (en) Edge computing unloading and resource management method based on deep reinforcement learning
Liu et al. Distributed location-aware task offloading in multi-uavs enabled edge computing

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant