CN117076121A - Intelligent task allocation method for wireless energy supply assisted mobile edge calculation - Google Patents
Intelligent task allocation method for wireless energy supply assisted mobile edge calculation Download PDFInfo
- Publication number
- CN117076121A CN117076121A CN202311064179.2A CN202311064179A CN117076121A CN 117076121 A CN117076121 A CN 117076121A CN 202311064179 A CN202311064179 A CN 202311064179A CN 117076121 A CN117076121 A CN 117076121A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- task
- unloading
- mobile edge
- unloading path
- 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.)
- Pending
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 35
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000005457 optimization Methods 0.000 claims abstract description 48
- 238000005265 energy consumption Methods 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 6
- 238000010606 normalization Methods 0.000 claims abstract description 4
- 230000009471 action Effects 0.000 claims description 14
- 230000002787 reinforcement Effects 0.000 claims description 12
- 230000008878 coupling Effects 0.000 claims description 6
- 238000010168 coupling process Methods 0.000 claims description 6
- 238000005859 coupling reaction Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 238000013468 resource allocation Methods 0.000 abstract description 4
- 238000011156 evaluation Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000254173 Coleoptera Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 210000003608 fece Anatomy 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/092—Reinforcement learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/502—Proximity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/509—Offload
Abstract
The invention provides an intelligent task allocation method for wireless energy supply assisted mobile edge computing, which relates to the technical field of mobile edge computing resource allocation and comprises the following steps: s1, establishing a basic framework of a wireless energy supply auxiliary mobile edge calculation model in a dynamic task scene; s2, generating a scheduling scheme according to the basic framework, performing task allocation and resource scheduling on equipment calculation tasks in a scene by utilizing an enhanced immune differential optimization algorithm, and performing normalization time delay and energy consumption minimum weighted sum calculation on the premise of ensuring that the energy can finish the group of tasks. The invention optimizes the charging time of the wireless energy station and the unloading path selection of the terminal equipment to assist the calculation of the mobile edge, and improves the processing efficiency of the mobile edge calculation server on the task through the system.
Description
Technical Field
The invention relates to the technical field of mobile edge computing resource allocation, in particular to an intelligent task allocation method for wireless energy supply assisted mobile edge computing.
Background
With the rapid development of the internet of things and the popularization of the fifth generation mobile communication technology in recent years, network equipment generates massive data to be processed at any moment. Mobile cloud computing technology has clearly failed to process these data efficiently, and therefore the academia proposed and advocated research and application of mobile edge computing (Mobile Edge Computing, MEC). The server for edge computing is very close to the equipment terminal in geographic position, so that the transmission quality problem and the safety problem existing in the mobile cloud computing technology can be well solved. The mobile edge computing technology not only can reduce the use of data transmission bandwidth, but also can better protect data related to personal privacy, thereby reducing the probability of sensitive data leakage. On the other hand, the edge computing supports mutual collaboration among the adjacent device terminals, so that idle computing resources and the like can be shared among the terminal devices, and meanwhile, the interactivity in a mobile scene is improved.
In the existing mobile edge computing technology, the WPT-MEC system can obtain an efficient resource allocation method by balancing computing resources, energy supply and communication resources, so that the problem of double limitation of computing capacity and energy of network edge equipment is solved. In multi-user wireless powered mobile edge computing systems, it is a worth studying question how to trade off communication resources, computing resources and energy supply and how to overcome the "dual near far effect" to minimize system latency or maximize device energy efficiency.
The total task processing time delay or the equipment energy efficiency of the WPT-MEC system is one of the key performance indexes. In recent years, although research on a WPT-MEC system is increased year by year, most algorithms used by students in models are traditional algorithms, and although the traditional optimization technology pursues theoretical accuracy, perfection and high convergence speed, the traditional optimization technology is based on calculus, so that the difficulty of the upper hand is high, the upper hand needs a strong mathematical work, the application scale of the traditional optimization algorithm is small, and the solving result is strongly dependent on an initial value. In addition, the problems in the WPT-MEC model basically belong to non-convex integer programming problems, so that the traditional optimization algorithm needs to be preprocessed before being applied, and the researched problems are converted into convex optimization problems. Compared with the traditional mathematical optimization method, the heuristic algorithm or the intelligent evolution algorithm has no any requirement on the mathematical property of the solved target problem, is a robust algorithm for the target problem, and has the characteristic of wide application. However, heuristic algorithms have better properties in small-scale data solutions, but the effectiveness of the algorithm decreases as the data size increases. The intelligent evolution algorithm is easy to sink into local optimum, so that the strategy effect is often output generally.
Disclosure of Invention
Therefore, the present invention aims to provide an intelligent task allocation method for wireless energy-supply assisted mobile edge calculation, so as to solve the problem of resource scheduling of wireless energy-supply assisted mobile edge calculation.
The invention adopts the following technical means:
an intelligent task allocation method for wireless energy supply assisted mobile edge calculation comprises the following steps:
s1, establishing a basic framework of a wireless energy supply auxiliary mobile edge calculation model in a dynamic task scene;
s2, generating a scheduling scheme according to the basic framework, performing task allocation and resource scheduling on equipment calculation tasks in a scene by utilizing an enhanced immune differential optimization algorithm, and performing normalization time delay and energy consumption minimum weighted sum calculation on the premise of ensuring that the energy can finish the group of tasks.
Further, S1 specifically includes the following steps:
s11, acquiring channel state information of terminal equipment through a wireless energy station, and carrying out wireless charging on the terminal equipment to obtain collected electric quantity information;
s12, acquiring position information and channel state information of the terminal equipment through a server, and generating an optimal unloading path by adopting the position information, the channel state information and the collected electric quantity information acquired through the server;
and S13, generating a proper task data unloading proportion of the terminal equipment by adopting the collected electric quantity information and the optimal unloading path.
Further, S11 specifically includes the following steps:
s111, acquiring position information and channel state information of the terminal equipment, and generating downlink channel gain h of each terminal equipment according to the position information and the channel state information of the terminal equipment down ；
S112, wireless charging is carried out on the terminal equipment by the wireless energy station at intervals, and calculation of charging time is accessed into an immune differential optimization algorithm and is based on the downlink channel gain h down And optimizing the delay by utilizing an optimization algorithm to obtain the optimized charging time, and entering S12 if the charging time reaches the standard.
Further, S12 specifically includes the following steps:
s121, acquiring position information and channel state information through a server, and calculating uplink channel gain h of the device according to the position information and the channel state information acquired by the server up ；
S122, task information is collected, and the task information is based on the uplink channel gain h up And possible unloading energy consumption and time delay of the task information acquisition equipment;
s123, determining whether the offloaded task is relayed or not by using the position information, the channel state information and the task information, and generating an offloading path decision.
Further, S13 specifically includes the following steps:
s131, generating a computing resource scheduling scheme by using an immune difference optimization algorithm according to task data amount collected by task information and unloading path decision;
s132, the terminal equipment performs unloading, calculation and data feedback of calculation tasks according to the resource scheduling scheme.
S133, the terminal equipment judges the energy consumption and the time delay after executing the current calculation task, if the energy consumption and the time delay meet the requirements, the task is continuously executed, and if not, the task output result is ended.
Further, the enhancing immune differential optimization algorithm of S2 includes:
the Q learning algorithm is used as an auxiliary algorithm of the immune difference optimization algorithm, the resource scheduling problem is preprocessed, and an unloading path under the current situation is obtained;
the antibody is used as a core in an immune differential optimization algorithm;
the intelligent combination algorithm starts with a Q learning algorithm, performs population preprocessing, rapidly makes an unloading path decision and adds the unloading path decision into an enhanced immune difference optimization algorithm, the immune difference optimization algorithm performs global convergence enhancement on the basis of an immune algorithm, the immune difference optimization algorithm uses the unloading path decision as an unloading path scheme, an antibody is used as an iterative operator, and interacts with the environment, and the two algorithms alternately cooperate to find a resource scheduling strategy.
Further, the Q learning algorithm includes:
the unloading path decision is used as an agent for learning in the environment, the Q value is updated through the selection of actions, the purpose of learning is achieved, each terminal device is provided with N+M-1 actions, N-1 is the number of terminal devices, M is the number of edge servers, the judgment of rewards is the change of the adaptability of the states of the front action and the back action, and the Q table in Q learning is updated;
selecting the action with the maximum Q value under each state according to a Q table generated by interaction of the intelligent agent and the environment, namely, unloading path decision of each device;
in the decision of offloading the path per device, the offloading priority to the edge server is highest, i.e. the same Q value is offloaded to the edge server.
Further, the offload path decision comprises:
the resource scheduling of the terminal equipment relates to a task unloading path scheme of the terminal equipment, and the coding problem is solved by using real number coding;
the offloading path decision is represented by a set of integers, each integer representing an offloading path decision code of a terminal device, i.e. offloading to the terminal device for relay or offloading to an edge server;
after the intelligent agent finishes the unloading path decision, processing the current unloading path decision at an algorithm interface to enable the current unloading path decision to get rid of the influence of the coupling degree of the current unloading path decision, the charging time and the unloading proportion, and taking the current unloading path decision as a precondition for generating an enhanced immune differential optimization algorithm population;
the knowledge learned by the intelligent agent without coupling degree in the process of interacting with the environment is reflected on the selection of the unloading path by the terminal equipment, and finally the learned strategy is output.
Further, processing the current offload path decision at the algorithm interface includes the steps of:
the interface is used for communicating the reinforcement learning algorithm with the reinforcement immune difference optimization algorithm, and generating a precondition required by the antibody, namely unloading path decision, through the coding interface;
and according to the unloading path decision generated by the reinforcement learning algorithm, as preprocessing for resource scheduling, generating charging time and unloading proportion by the reinforcement immune difference optimization algorithm, and distributing time resources according to the unloading path decision.
Compared with the prior art, the invention has the following advantages:
1. the enhanced immunity differential optimization algorithm, also called as Q-IADE algorithm, is formed by adding an iterative operator of a differential evolution algorithm into an immune algorithm, and the algorithm realizes the improvement of the local searching capability of a feasible solution on the premise of keeping good population individual diversity owned by the immune algorithm;
2. and secondly, combining the Q learning algorithm with the improved enhanced immune differential optimization algorithm, so that the dynamic planning problem is better solved by using the algorithm, and the unloading path decision is more accurately generated.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to the drawings without inventive effort to a person skilled in the art.
Fig. 1 is a basic frame diagram of the present invention.
Fig. 2 is a flow chart of the algorithm of the present invention.
FIG. 3 is a graph of the improved results of the algorithm of the present invention.
Fig. 4 is a graph comparing algorithm results when the number of edge servers is 1 and no relay condition exists.
Fig. 5 is a graph comparing algorithm results when the number of edge servers is 4 and no relay condition exists.
Fig. 6 is a graph comparing algorithm results when the number of edge servers is 1 and a relay condition exists.
Fig. 7 is a graph comparing algorithm results when the number of edge servers is 4 and a relay condition exists.
FIG. 8 is a graph showing the comparison of the results of four algorithms when the number of servers is increased.
Fig. 9 is a comparison chart of four algorithm results when the number of terminal devices is increased.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
As shown in fig. 1 and 2, the present invention provides an intelligent task allocation method for wireless energy supply assisted mobile edge calculation, which includes:
s1, establishing a basic framework of a wireless energy supply auxiliary mobile edge calculation model in a dynamic task scene.
S11, the wireless energy station acquires the channel state information of the terminal equipment and carries out wireless charging on the channel state information.
S111, acquiring position information and channel state information of the terminal equipment, thereby generating downlink channel gain h of each terminal equipment down 。
S112, wireless charging is carried out on the terminal equipment by the wireless energy station at intervals, calculation of charging time is accessed into an enhanced immune differential optimization algorithm (Q-IADE algorithm), delay is optimized by the optimization algorithm, and the optimized charging time reaches the standard and then enters the next step.
S12, acquiring the position information of the terminal equipment and the channel state information, and generating an optimal unloading path.
S121, acquiring position information and channel state information for calculating uplink channel gain h of the device up 。
S122, task information collection is used for acquiring possible unloading energy consumption and possible time delay of the equipment.
S123, determining whether to relay by utilizing the information collected by the two modules, and generating an unloading path decision.
S13, generating a proper task data unloading proportion of the terminal equipment.
S131, analyzing and calculating a resource scheduling scheme by using an intelligent task allocation method according to the task data quantity and the unloading path decision.
S132, the terminal equipment performs unloading, calculation and data feedback of calculation tasks according to the resource scheduling scheme.
The computing task offload acts as a server resource schedule, manifesting as a relay device or edge server that the terminal device selects to offload.
The computing resource allocation is embodied in the terminal device selecting the offload proportion of the task data.
S133, the terminal equipment judges the energy consumption and the time delay after executing the current calculation task, if the energy consumption and the time delay meet the requirements, the task is continuously executed, and if not, the task output result is ended.
S2, performing task allocation and resource scheduling on equipment calculation tasks in a scene according to an intelligent task allocation method by using a scheduling scheme for wireless energy supply assisted mobile edge calculation, and finally performing normalization time delay and energy consumption minimum weighted sum calculation on the premise of ensuring that energy can complete the group of tasks.
And the Q learning algorithm is used as an auxiliary algorithm of the enhanced immune differential optimization algorithm, the resource scheduling problem is preprocessed, and an unloading path under the current situation is found.
The unloading path decision is used as an agent to learn in the environment, the Q value is updated through the selection of actions, so that the purpose of learning is achieved, and each terminal device (state) has (N+M-1) actions which can be selected (N-1 terminal devices and M edge servers which can be selected). The evaluation of the reward is a change in fitness of the states of the front and rear actions (embodied as current and historical states but excluding fitness changes of future states) and updates the Q table in Q learning.
The terminal equipment resource scheduling relates to a terminal equipment task unloading path scheme, and the coding problem is solved by using real number coding.
The "offload path decision" is represented by a set of integers, each representing an offload path decision code for a terminal device, i.e., offload to a terminal device for relay or offload to an edge server.
After the intelligent agent finishes the unloading path decision, the current unloading path decision is required to be processed at an algorithm interface (connection of two algorithms), so that the influence of the coupling degree between the current unloading path decision and the charging time and the unloading proportion is eliminated, and the current unloading path decision is used as a precondition for population generation of the enhanced immune differential optimization algorithm.
The interface communicates the reinforcement learning algorithm with the reinforcement immune differential optimization algorithm, and generates preconditions required by the antibody, namely unloading path decisions, through the coding interface.
And according to the unloading path decision generated by the reinforcement learning algorithm, as preprocessing for resource scheduling, the charge time and the unloading proportion are generated by the reinforcement immune difference optimization algorithm, so that the complexity of coding is reduced. The specific scheme is as follows: and distributing the time resources according to the unloading path decision.
The knowledge learned by the intelligent agent without coupling degree in the process of interacting with the environment is reflected on the selection of the unloading path by the terminal equipment, and finally the learned strategy is output.
And at the moment, selecting the action with the maximum Q value under each state according to a Q table generated by interaction of the intelligent agent and the environment, namely, unloading path decision of each device.
The priority of offloading to the edge server is highest in the decision of offloading the path per device, i.e. offloading to the edge server is prioritized with the Q value.
The antibody is used as a core in the evolutionary algorithm, is an important tool for iterative optimization of the evolutionary algorithm, and is still used as an iterative core of the algorithm in the combination algorithm.
The intelligent combination algorithm starts with a Q learning algorithm, performs population preprocessing, rapidly makes an unloading path decision and adds the unloading path decision into an enhanced immune difference optimization algorithm, wherein the enhanced immune difference optimization algorithm is used for enhancing global convergence on the basis of an immune algorithm, then the enhanced immune difference optimization algorithm uses the decision as an unloading path scheme, uses an antibody as an iterative operator, interacts with the environment, and alternately cooperates with the two algorithms to find a resource scheduling strategy.
The auxiliary effect of the enhanced immune differential optimization algorithm is as follows: (1) While maintaining the diversity of the solution, the local searching capability and global convergence of the solution are enhanced (2) to find the charging time and task offloading ratio.
According to the wireless energy supply (WPT) auxiliary mobile edge calculation intelligent task allocation method based on the intelligent combination algorithm, an improved immune algorithm, namely an enhanced immune differential optimization algorithm, is used, and the intelligent combination algorithm combined with a Q learning algorithm takes the minimum weighted sum of normalized time delay and energy consumption when a task is completed as an optimization target.
In the embodiment, experiments are performed in an actual task scene, and tests are performed under the number of edge servers and the number of terminal devices in different scales respectively. The comparison algorithm herein adopts an adaptive weighted particle swarm Algorithm (AWPSO), a dung beetle algorithm (DBO) and a fire hawk algorithm (FHO).
As shown in fig. 3, the number of edge servers is 1, the number of terminal devices is 10, and the algorithm improvement results are compared with the same evaluation times.
As shown in fig. 4, the number of edge servers is 1, the number of terminal devices is 10, and when no relay condition exists (the difference between the heterogeneous devices or the task amount is large), the four algorithms are compared with each other under the same evaluation times.
As shown in fig. 5, the number of edge servers is 4, the number of terminal devices is 60, and when no relay condition exists (the difference between the heterogeneous devices or the task amount is large), the four algorithms are compared with each other under the same evaluation times.
As shown in fig. 6, the number of edge servers is 1, the number of terminal devices is 10, and when relay conditions (heterogeneous devices or large difference in task amount) exist, four algorithm comparison diagrams are obtained under the same evaluation times.
As shown in fig. 7, the number of edge servers is 4, the number of terminal devices is 60, and when relay conditions (heterogeneous devices or large difference in task amount) exist, the four algorithms are compared with each other under the same evaluation times.
As can be seen from fig. 6 and 7, the Q-IADE algorithm can greatly reduce the overhead because it can better find the offload path or the relay offload path.
When the number of terminal devices is 10 and the number of edge servers is gradually increased, two algorithm pairs with the same evaluation times are shown in fig. 8.
The number of edge servers is 4, and when the number of terminal devices is gradually increased, two algorithm pairs with the same evaluation times are shown in fig. 9.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention.
Claims (9)
1. An intelligent task allocation method for wireless energy supply assisted mobile edge calculation is characterized by comprising the following steps:
s1, establishing a basic framework of a wireless energy supply auxiliary mobile edge calculation model in a dynamic task scene;
s2, generating a scheduling scheme according to the basic framework, performing task allocation and resource scheduling on equipment calculation tasks in a scene by utilizing an enhanced immune differential optimization algorithm, and performing normalization time delay and energy consumption minimum weighted sum calculation on the premise of ensuring that the energy can finish the group of tasks.
2. The intelligent task allocation method for wireless energy supply assisted mobile edge computing according to claim 1, wherein S1 specifically comprises the following steps:
s11, acquiring channel state information of terminal equipment through a wireless energy station, and carrying out wireless charging on the terminal equipment to obtain collected electric quantity information;
s12, acquiring position information and channel state information of the terminal equipment through a server, and generating an optimal unloading path by adopting the position information, the channel state information and the collected electric quantity information acquired through the server;
and S13, generating a proper task data unloading proportion of the terminal equipment by adopting the collected electric quantity information and the optimal unloading path.
3. The intelligent task allocation method for wireless energy supply assisted mobile edge computing according to claim 2, wherein S11 specifically comprises the steps of:
s111, acquiring position information and channel state information of the terminal equipment, and generating downlink channel gain h of each terminal equipment according to the position information and the channel state information of the terminal equipment down ；
S112, wireless charging is carried out on the terminal equipment by the wireless energy station at intervals, and calculation of charging time is accessed into an immune differential optimization algorithm and is based on the downlink channel gain h down And optimizing the delay by utilizing an optimization algorithm to obtain the optimized charging time, and entering S12 if the charging time reaches the standard.
4. The intelligent task allocation method for wireless energy supply assisted mobile edge computing according to claim 3, wherein S12 specifically comprises the steps of:
s121, acquiring position information and channel state information through a server, and calculating uplink channel gain h of the device according to the position information and the channel state information acquired by the server up ；
S122, task information is collected, and the task information is based on the uplink channel gain h up And possible unloading energy consumption and time delay of the task information acquisition equipment;
s123, determining whether the offloaded task is relayed or not by using the position information, the channel state information and the task information, and generating an offloading path decision.
5. The intelligent task allocation method for wireless energy supply assisted mobile edge computing according to claim 4, wherein S13 specifically comprises the steps of:
s131, generating a computing resource scheduling scheme by using an immune difference optimization algorithm according to task data amount collected by task information and unloading path decision;
s132, the terminal equipment performs unloading, calculation and data feedback of calculation tasks according to the resource scheduling scheme.
S133, the terminal equipment judges the energy consumption and the time delay after executing the current calculation task, if the energy consumption and the time delay meet the requirements, the task is continuously executed, and if not, the task output result is ended.
6. The intelligent task allocation method for wireless energy-assisted mobile edge computing according to claim 1, wherein S2 the enhanced immune differential optimization algorithm comprises:
the Q learning algorithm is used as an auxiliary algorithm of the immune difference optimization algorithm, the resource scheduling problem is preprocessed, and an unloading path under the current situation is obtained;
the antibody is used as a core in an immune differential optimization algorithm;
the intelligent combination algorithm starts with a Q learning algorithm, performs population preprocessing, rapidly makes an unloading path decision and adds the unloading path decision into an enhanced immune difference optimization algorithm, the immune difference optimization algorithm performs global convergence enhancement on the basis of an immune algorithm, the immune difference optimization algorithm uses the unloading path decision as an unloading path scheme, an antibody is used as an iterative operator to interact with the environment, and the immune algorithm and the differential evolution algorithm are alternately matched to search a resource scheduling strategy.
7. The wireless energy assisted mobile edge computing oriented intelligent task allocation method of claim 6 wherein said Q learning algorithm comprises:
the unloading path decision is used as an agent for learning in the environment, the Q value is updated through the selection of actions, the purpose of learning is achieved, each terminal device is provided with N+M-1 actions, N-1 is the number of terminal devices, M is the number of edge servers, the judgment of rewards is the change of the adaptability of the states of the front action and the back action, and the Q table in Q learning is updated;
selecting the action with the maximum Q value under each state according to a Q table generated by interaction of the intelligent agent and the environment, namely, unloading path decision of each device;
in the decision of offloading the path per device, the offloading priority to the edge server is highest, i.e. the same Q value is offloaded to the edge server.
8. The wireless energy assisted mobile edge computing oriented intelligent task allocation method of claim 7 wherein said offload path decision comprises:
the resource scheduling of the terminal equipment relates to a task unloading path scheme of the terminal equipment, and the coding problem is solved by using real number coding;
the offloading path decision is represented by a set of integers, each integer representing an offloading path decision code of a terminal device, i.e. offloading to the terminal device for relay or offloading to an edge server;
after the intelligent agent finishes the unloading path decision, processing the current unloading path decision at an algorithm interface to enable the current unloading path decision to get rid of the influence of the coupling degree of the current unloading path decision, the charging time and the unloading proportion, and taking the current unloading path decision as a precondition for generating an enhanced immune differential optimization algorithm population;
the knowledge learned by the intelligent agent without coupling degree in the process of interacting with the environment is reflected on the selection of the unloading path by the terminal equipment, and finally the learned strategy is output.
9. The wireless energy assisted mobile edge computing oriented intelligent task allocation method of claim 8 wherein processing the current offload path decision at the algorithm interface comprises the steps of:
the interface is used for communicating the reinforcement learning algorithm with the reinforcement immune difference optimization algorithm, and generating a precondition required by the antibody, namely unloading path decision, through the coding interface;
and according to the unloading path decision generated by the reinforcement learning algorithm, as preprocessing for resource scheduling, generating charging time and unloading proportion by the reinforcement immune difference optimization algorithm, and distributing time resources according to the unloading path decision.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311064179.2A CN117076121A (en) | 2023-08-21 | 2023-08-21 | Intelligent task allocation method for wireless energy supply assisted mobile edge calculation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311064179.2A CN117076121A (en) | 2023-08-21 | 2023-08-21 | Intelligent task allocation method for wireless energy supply assisted mobile edge calculation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117076121A true CN117076121A (en) | 2023-11-17 |
Family
ID=88705679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311064179.2A Pending CN117076121A (en) | 2023-08-21 | 2023-08-21 | Intelligent task allocation method for wireless energy supply assisted mobile edge calculation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117076121A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117545017A (en) * | 2024-01-09 | 2024-02-09 | 大连海事大学 | Online computing and unloading method for wireless energy supply mobile edge network |
-
2023
- 2023-08-21 CN CN202311064179.2A patent/CN117076121A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117545017A (en) * | 2024-01-09 | 2024-02-09 | 大连海事大学 | Online computing and unloading method for wireless energy supply mobile edge network |
| CN117545017B (en) * | 2024-01-09 | 2024-03-19 | 大连海事大学 | Online computing and unloading method for wireless energy supply mobile edge network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xu et al. | Intelligent offloading for collaborative smart city services in edge computing | |
| Xu et al. | A computation offloading method over big data for IoT-enabled cloud-edge computing | |
| Huang et al. | A lightweight collaborative deep neural network for the mobile web in edge cloud | |
| WO2019184836A1 (en) | Data analysis device, and multi-model co-decision system and method | |
| CN117076121A (en) | Intelligent task allocation method for wireless energy supply assisted mobile edge calculation | |
| CN114051254B (en) | Green cloud edge collaborative computing unloading method based on star-ground fusion network | |
| Ren et al. | Collaborative edge computing and caching with deep reinforcement learning decision agents | |
| CN111836284B (en) | Energy consumption optimization calculation and unloading method and system based on mobile edge calculation | |
| CN112637806A (en) | Transformer substation monitoring system based on deep reinforcement learning and resource scheduling method thereof | |
| CN109375999A (en) | A kind of MEC Random Task moving method based on Bayesian network | |
| CN110851277A (en) | Task scheduling strategy based on edge cloud cooperation in augmented reality scene | |
| Cheng et al. | Distributed green offloading and power optimization in virtualized small cell networks with mobile edge computing | |
| Liao et al. | Joint multi-user DNN partitioning and task offloading in mobile edge computing | |
| Shi et al. | Multiuser co-inference with batch processing capable edge server | |
| CN113139639B (en) | MOMBI-oriented smart city application multi-target computing migration method and device | |
| Sadatdiynov et al. | An intelligent hybrid method: Multi-objective optimization for MEC-enabled devices of IoE | |
| Gao et al. | Fast adaptive task offloading and resource allocation via multiagent reinforcement learning in heterogeneous vehicular fog computing | |
| CN114169506A (en) | Deep learning edge computing system framework based on industrial Internet of things platform | |
| Cui et al. | Multi-Agent Reinforcement Learning Based Cooperative Multitype Task Offloading Strategy for Internet of Vehicles in B5G/6G Network | |
| CN115361453B (en) | Load fair unloading and migration method for edge service network | |
| Sun et al. | Semantic-driven computation offloading and resource allocation for UAV-assisted monitoring system in vehicular networks | |
| CN115622603A (en) | Age minimization optimization method for auxiliary transmission information | |
| CN115914230A (en) | Adaptive mobile edge computing unloading and resource allocation method | |
| CN112492651B (en) | Resource scheduling scheme optimization method and device | |
| CN114697978A (en) | Intelligent reflector assisted uplink moving edge calculation method and system |
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 |