CN114650568A - Distributed unloading method based on energy collection in mobile Ad Hoc cloud - Google Patents
Distributed unloading method based on energy collection in mobile Ad Hoc cloud Download PDFInfo
- Publication number
- CN114650568A CN114650568A CN202210269015.2A CN202210269015A CN114650568A CN 114650568 A CN114650568 A CN 114650568A CN 202210269015 A CN202210269015 A CN 202210269015A CN 114650568 A CN114650568 A CN 114650568A
- Authority
- CN
- China
- Prior art keywords
- task
- client terminal
- energy
- time slot
- terminal
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000004364 calculation method Methods 0.000 claims abstract description 28
- 230000006870 function Effects 0.000 claims abstract description 20
- 238000003860 storage Methods 0.000 claims abstract description 6
- 239000003795 chemical substances by application Substances 0.000 claims description 71
- 238000005457 optimization Methods 0.000 claims description 30
- 238000005265 energy consumption Methods 0.000 claims description 26
- 238000003306 harvesting Methods 0.000 claims description 21
- 239000000126 substance Substances 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 18
- 230000008901 benefit Effects 0.000 claims description 15
- 230000005540 biological transmission Effects 0.000 claims description 15
- 238000004891 communication Methods 0.000 claims description 15
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 15
- 238000010295 mobile communication Methods 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000004088 simulation Methods 0.000 description 6
- 230000007423 decrease Effects 0.000 description 4
- UGAJKWZVPNVCIO-UHFFFAOYSA-N Terminalin Chemical compound O1C(=O)C(C2=3)=C(C4=C(O)C(O)=C(O)C=C4C(=O)O4)C4=C(O)C=3OC(=O)C3=C2C1=C(O)C(OC1=O)=C3C2=C1C=C(O)C(O)=C2O UGAJKWZVPNVCIO-UHFFFAOYSA-N 0.000 description 3
- QTNGLMWAVBOBLJ-UHFFFAOYSA-N Terminaline Natural products C1CC2C(O)C(O)CCC2(C)C2C1C1CCC(C(C)N(C)C)C1(C)CC2 QTNGLMWAVBOBLJ-UHFFFAOYSA-N 0.000 description 3
- 230000036961 partial effect Effects 0.000 description 3
- 229930183689 terminalin Natural products 0.000 description 3
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 208000001613 Gambling Diseases 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 241000764238 Isis Species 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000001808 coupling effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000005562 fading Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 231100001261 hazardous Toxicity 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- 231100000331 toxic Toxicity 0.000 description 1
- 230000002588 toxic effect Effects 0.000 description 1
Images
Classifications
-
- 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
- H04W28/09—Management thereof
-
- 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
- G06F9/44—Arrangements for executing specific programs
- G06F9/445—Program loading or initiating
- G06F9/44594—Unloading
-
- 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
- H04W28/09—Management thereof
- H04W28/0917—Management thereof based on the energy state of entities
-
- 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
- H04W28/09—Management thereof
- H04W28/0925—Management thereof using policies
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of mobile communication, and particularly relates to a distributed unloading method based on energy collection in a mobile Ad Hoc cloud, which comprises the steps of considering a mobile Ad Hoc cloud network formed by a group of nearby terminal devices with EH functions, and respectively establishing a calculation task model, a task unloading model and an energy collection model; the client terminal is used as a buyer, resources are purchased from the agent terminal according to the self calculation task requirement, and the income maximization problem of the buyer is established; taking the agent terminal as a seller, providing different calculation and storage resources for the client terminal through dynamic resource quotation, and establishing the profit maximization problem of the seller; calculating an optimal task unloading strategy unloaded by the agent terminal selected by the buying direction and an optimal quotation strategy of a seller by utilizing a Lagrange multiplier method and a KKT condition; the invention can effectively improve the system income, stabilize the battery energy and reduce the backlog of the task queue.
Description
Technical Field
The invention belongs to the technical field of mobile communication, and particularly relates to a distributed unloading method based on energy collection in a mobile Ad Hoc cloud.
Background
Under the drive of the rapidly developing internet of things technology, the number of terminal devices and data flow are increased explosively, and computing-intensive and delay-sensitive applications (such as automatic driving, virtual/augmented reality, online games and the like) are continuously started, which puts higher requirements on real-time computing resources of the network. On the other hand, the traditional terminal equipment has limited energy, bandwidth and computing resources, and is difficult to meet the real-time computing requirements and the durable cruising ability of emerging applications characterized by large data and intellectualization. Mobile Cloud Computing (MCC) and Mobile Edge Computing (MEC) alleviate the problem of limited Computing resources of terminal equipment by offloading Computing tasks to a Cloud for processing, but the problems of insufficient Edge Computing power and unbalanced distribution still cannot be effectively solved. In some network scenarios (such as Ad Hoc networks, unmanned aerial vehicle networks, vehicle cloud networks, etc.), there is no available cloud server or local clout, or a computing task cannot be processed in time due to network congestion, which makes it difficult to effectively meet the service experience of a user.
Mobile Ad Hoc cloud computing utilizes free or excess resources of a group of neighboring terminal devices (e.g., smart phones, laptops, AI monitoring devices, vehicle-mounted smart terminals, etc.) to cooperatively process network tasks. Subject to the limitations of volume and hardware cost, the battery capacity of conventional terminal devices is difficult to meet with for long-term endurance requirements, especially when the devices are distributed in remote or toxic and hazardous environments, and power is difficult to supply through rechargeable batteries or conventional power grids. Energy Harvesting (EH) technology support equipment acquires renewable Energy sources (such as solar Energy, wind Energy, mechanical Energy and the like) from the environment to support communication and task processing of the equipment, and the technology becomes an important technology for realizing green mobile communication. Therefore, the integration of the mobile Ad Hoc cloud computing and the EH technology has important significance for improving the network computing performance.
In recent years, green communication combining the MEC and the EH technology is also receiving wide attention, which brings great inspiration to the fusion of the EH technology and the Ad Hoc mobile cloud. Some of the major achievements are: (1) dynamic computational offloading based on incomplete state information in energy harvesting small cell networks: partially Observable random gambling (ref: TANG Q, XIE R, HUANG T, et al. dynamic computing adapting With Imperfect State Information in Energy collecting Small Cell Networks: A partial Observable storage Game [ J ]. IEEE Wireless Communication Letters,2020,9(8):1300-1304.DOI: 10.1109/LWC.2020.2989147.): the algorithm is a dynamic unloading algorithm based on a distributed predictable random game aiming at an MEC heterogeneous environment with EH small base stations and considering average time delay and average energy requirements, and the base stations make optimal unloading decisions by using incomplete state information. (2) Distributed computing offloading in an internet of things fog computing system with energy harvesting functionality: DEC-POMDP method (ref: TANG Q, XIE R, YU F A, et al. decentralized computing of streaming in IoT fog computing system with energy harving: A DEC-POMDP improvement [ J ]. IEEE Internet of today Journal,2020,7(6):4898-4911.DOI: 10.1109/JIOT.2020.2971323.): the algorithm provides a distributed unloading algorithm based on learning aiming at the problem of predictable distributed unloading in an energy collection internet of things fog system, and makes the internet of things equipment make an approximately optimal decision according to a predicted system state under the condition of meeting a time delay constraint. (3) Energy-collection-based task Offloading Energy consumption and latency Tradeoff algorithm in Mobile Edge computation (ref: ZHANG G, ZHANG W, CAO Y, et al. Energy-Delay trade off for Dynamic Offloading in Mobile-Edge Computing System with Energy Harvesting Devices [ J ]. IEEE Transactions on Industrial information, 2018,14(10):4642-4655.DOI: 10.1109/TII.2018.2843365.): the algorithm researches the task unloading problem of the MEC system with the energy collection capability under the constraints of queue backlog and battery power, and minimizes the average weighted sum of the energy consumption and the execution time delay of the mobile equipment through a dynamic unloading algorithm based on Lyapunov.
Different terminal devices with EH functions have different data characteristics and application requirements, which pose a serious challenge to the task offloading efficiency, and most of the existing task offloading schemes only consider that a mobile terminal has the EH function, and do not consider that a base station or an MEC server has the EH function. Therefore, how to develop a task offloading strategy with energy harvesting level in a distributed manner has important research value.
Disclosure of Invention
In view of this, in order to achieve maximization of system revenue and stabilization of queue backlog, the invention provides a distributed unloading method based on energy collection in a mobile Ad Hoc cloud, which specifically includes the following steps:
considering a mobile Ad Hoc cloud network formed by a group of nearby terminal devices with EH functions, respectively establishing a calculation task model, a task unloading model and an energy collection model;
the method comprises the steps that a client terminal is used as a buyer, resources are purchased from an agent terminal according to own computing task requirements, the Lyapunov optimization theory is adopted, and the profit maximization problem of the buyer is established based on a computing task model, a task unloading model and an energy collection model;
taking an agent terminal as a seller, providing different calculation and storage resources for a client terminal through dynamic resource quotation, and establishing the profit maximization problem of the seller based on a calculation task model, a task unloading model and an energy collection model;
according to the task backlog of the client terminal, the battery energy level and the quotation of the proxy terminal, in each time slot, calculating an optimal task unloading strategy unloaded by the proxy terminal selected by the buying direction and an optimal quotation strategy of a seller by using a Lagrange multiplier method and a KKT condition;
and if the optimal task unloading strategy of the buyer and the optimal quotation strategy of the seller meet the SteinKelberg equilibrium solution, the client terminal unloads the tasks to the agent terminal according to the optimal task unloading strategy.
Further, the revenue maximization problem of the buyer is established based on the calculation task model, the task unloading model and the energy collection model and is expressed as follows:
constraint conditions are as follows:
wherein the content of the first and second substances,
representing a profit maximization problem of the buyer at the t-th time slot; viControl parameters indicating the ith client terminal;
representing the total profit of the ith client terminal in the time slot t;
indicating that the unloading profit is related to the task queue;
for the ith client terminal CiThe amount of task arrivals in time slot t,
indicates the ith client terminal CiThe total amount of tasks processed in time slot t;
virtualization of EH device for ith client terminalEnergy queue, denoted as
θiAs a parameter of the disturbance of the EH device,
backlog the energy queue of the EH equipment with the ith client terminal at the beginning of the time slot t; eminRepresents a minimum discharge energy of the battery;
representing the total energy consumption generated by the ith client terminal in the time slot t; emaxRepresents the maximum discharge energy of the battery;
represents the jth agent terminal A in the tth time slotjAssisting the computing energy consumption generated when the ith client terminal computes a task;
representing the energy queue backlog of the EH equipment carried by the ith client terminal at the beginning of the time slot t;
the task amount of an ith client terminal to be unloaded to a jth agent terminal in a tth time slot is represented;
the backlog of the task queue of the ith client terminal at the t time slot is represented;
the average task queue backlog of the ith client terminal in the t time slot is represented; t represents the system running time;
indicating the desire.
Further, the revenue maximization problem of the buyer established based on the calculation task model, the task unloading model and the energy collection model is decomposed into an optimal solution for solving the energy collection and a task unloading optimization problem, wherein the optimal solution for solving the energy collection is expressed as:
when in use
When the temperature of the water is higher than the set temperature,
when in use
When the temperature of the water is higher than the set temperature,
after decoupling the energy harvesting optimization problem, the task offloading optimization problem is represented as:
constraint conditions are as follows:
wherein the content of the first and second substances,
represents an optimal solution for energy harvesting; gamma raymaxRepresenting the maximum amount of energy collected by the EH device at the tth time slot.
Further, when the computing task is processed locally and
and then, solving the optimal local calculation frequency of the t-th time slot by using a Lagrange multiplier method and a KKT condition as follows:
wherein the content of the first and second substances,
κian effective energy cost coefficient for the ith client terminal chip; l is a radical of an alcoholiIndicating the unit processing capacity of the ith client terminal; tau is the unit time slot length; xiiOffloading the benefit weight parameter for the task.
Further, when the computing task is processed locally and
the optimal local computation frequency is:
wherein, the first and the second end of the pipe are connected with each other,
indicating when a computing task is processed locally and
time-optimal local computation frequency;
represents the maximum CPU processing frequency of the ith client;
represents the maximum discharge energy of the battery of the ith client;
the optimal task amount of the ith client terminal unloaded to the jth proxy terminal at the tth time slot is represented;
the transmission power of the ith client terminal for the t time slot for unloading the task to the jth agent terminal;
a task transmission rate for offloading the task to the jth agent terminal for the ith client terminal at the tth time slot; kappaiAn effective energy cost coefficient for the ith client terminal chip; τ is the unit slot length.
Further, the client terminal selects to unload part of tasks to the agent terminal when each time slot starts, the rest of tasks are unloaded locally, and the optimal task amount unloaded from the ith client terminal to the jth agent terminal at the tth time slot is represented as:
wherein ξiOffloading a benefit weight parameter for the task;
for the ith time slot, the ith client terminal CiTo the jth agent terminal AjA unit payment cost for purchasing computing resources;
unit communication cost for offloading a task to a jth agent terminal for the ith client terminal of the t slot;
the transmission power of the ith client terminal for the t time slot to unload the task to the jth agent terminal;
the task transmission rate at which the ith client terminal offloads tasks to the jth proxy terminal for the t slot.
Further, the seller's revenue maximization problem is established based on the calculation task model, the task unloading model and the energy collection model and expressed as:
an objective function:
wherein the content of the first and second substances,
representing a seller's revenue maximization problem; v represents a control parameter;
representing the total income of the jth agent terminal in the tth time slot;
the virtual energy queue of the jth agent terminal for the tth time slot is represented as
θjAs a parameter of the disturbance of the EH device,
the energy queue backlog of the EH equipment carried by the jth agent terminal at the beginning of the time slot t is carried out;
representing the energy actually collected by the EH equipment carried by the jth proxy terminal in the time slot t;
jth proxy terminal A in tth time slotjComputing energy consumption in processing a computing task from an i-th client terminal; eminRepresents a minimum discharge energy of the battery; emaxRepresents the maximum discharge energy of the battery;
for the ith time slot, the ith client terminal CiTo the jth agent terminal AjThe unit of the purchased computing resource pays a cost.
Further, decomposing a profit maximization problem of a seller established based on a calculation task model, a task unloading model and an energy collection model into an optimal solution for solving energy collection and a task unloading optimization problem, wherein the optimal solution for solving energy collection is expressed as:
when in use
When the temperature of the water is higher than the set temperature,
when in use
When the temperature of the water is higher than the set temperature,
after decoupling the energy collection optimization problem, the task offloading optimization problem is represented as;
constraint conditions are as follows:
wherein the content of the first and second substances,
represents an optimal solution for energy harvesting; gamma raymaxThe representation represents the maximum amount of energy collected by the EH device at the tth time slot.
Further, the optimal quotation of the t-th time slot obtained by the Lagrange multiplier method and the KKT condition is as follows:
wherein, the first and the second end of the pipe are connected with each other,
for the ith client terminal CiTo the jth agent terminal AjPurchasing computing resourcesThe optimal unit payment cost;
for the ith client terminal CiThe optimal unloading task amount;
for the ith time slot, the ith client terminal CiTo the jth agent terminal AjA unit payment cost for purchasing computing resources;
a unit communication cost for offloading a task to a jth proxy terminal for a tth time slot ith client terminal; etaijRepresenting the unit energy consumption cost of the agent terminal processing task; xiiOffloading the benefit weight parameter for the task.
The invention provides a distributed dynamic unloading scheme by considering a mobile Ad Hoc cloud network with EH functions of all terminal devices. Meanwhile, rational terminal equipment cannot serve other terminal equipment for a free time, and an incentive mechanism based on dynamic quotation is provided by applying Lyapunov optimization theory and game theory to optimize system income. Compared with the existing scheme, the simulation result shows that the provided scheme can effectively improve the system benefit, stabilize the battery energy and reduce the backlog of the task queue.
Drawings
FIG. 1 is a flow chart of a distributed offloading method based on energy harvesting in a mobile Ad Hoc cloud of the present invention;
FIG. 2 is an Ad Hoc cloud collaboration system model;
FIG. 3 is a price updating diagram during dynamic resource quotation;
fig. 4 is a diagram of CPU frequency update during dynamic resource quotation.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides a distributed unloading method based on energy collection in a mobile Ad Hoc cloud, which comprises the following steps:
considering a mobile Ad Hoc cloud network formed by a group of nearby terminal devices with EH functions, respectively establishing a calculation task model, a task unloading model and an energy collection model;
the method comprises the steps that a client terminal is used as a buyer, resources are purchased from an agent terminal according to own computing task requirements, the Lyapunov optimization theory is adopted, and the profit maximization problem of the buyer is established based on a computing task model, a task unloading model and an energy collection model;
taking the agent terminal as a seller, providing different computing and storage resources for the client terminal through dynamic resource quotation, and establishing the profit maximization problem of the seller based on a computing task model, a task unloading model and an energy collection model;
according to the task backlog of the client terminal, the battery energy level and the quotation of the proxy terminal, in each time slot, calculating an optimal task unloading strategy unloaded by the proxy terminal selected by the buying direction and an optimal quotation strategy of a seller by using a Lagrange multiplier method and a KKT condition;
and if the optimal task unloading strategy of the buyer and the optimal quotation strategy of the seller meet the SteinKelberg equilibrium solution, the client terminal unloads the tasks to the agent terminal according to the optimal task unloading strategy.
The embodiment of the invention respectively explains the scheme of the invention in four aspects of system model, problem description, dynamic distributed unloading scheme based on game theory, simulation result and analysis.
First, system model
Consider a mobile Ad Hoc cloud network consisting of a set of nearby terminal devices with EH functionality, each terminalThe device can be a client terminal or a proxy terminal, and can be used as a client terminal when a computing task is available and can be used as a proxy terminal when the device is in an idle state. Assuming that each Client terminal has a compute intensive task, the system model, as shown in fig. 2, includes M Client terminals with different compute tasks (Client,
and N Agent terminals (agents,
) Spare or excess computing resources may be shared. CiCan be processed locally or offloaded to AjAnd (4) performing collaborative calculation. Where j-0 indicates that the task is processed locally. Definition CiTask offload decision-making
Is represented by CiOffloading tasks to jth agent A at time slot tjIf not, then,
assuming that time can be slotted, the unit slot length is defined as tau, and the slot index is
τdIs represented by CiThe maximum tolerated delay.
1. Computing task model
Hypothesis CiIs lambdaiAnd the computational tasks can be arbitrarily divided, defining the time slot t,
is CiThe amount of task arrivals of (c),
backlog the task queue and satisfy
For maximum task arrival, CiQueue characteristic of
Is represented by, wherein, LiIs represented by CiThe unit processing capacity of (a) is,
is represented by CiThe total amount of tasks processed in the time slot t,
indicating unloading to AjThe amount of the task(s) of (c),
represents the amount of locally processed tasks and satisfies
Is the maximum processable task volume. Wherein the content of the first and second substances,
is AjAnd satisfies the CPU processing frequency of
And
respectively represent maximum and maximumThe frequency of the processing by the small CPU,
is represented by CiThe CPU processing frequency of (1). The backlog of the task queue of the t +1 time slot is
2. Task offloading model
The task unloading process mainly comprises the steps of uploading a task of the client terminal, cooperatively calculating the unloading task by the agent terminal and returning a calculation result. The data volume when returning the result is very small, and the generated time delay and energy consumption can be ignored. For simplicity, only the energy consumption resulting from processing a computing task is considered in the local computation
Wherein the content of the first and second substances,
is the effective energy cost factor associated with the chip architecture.
Suppose that the wireless channel during task uploading is an additive white Gaussian noise channel and the transmission power is
And satisfy
According to the Shannon formula, C in the t time slotiOffloading tasks to AjTask transmission rate of
Wherein the content of the first and second substances,
in order to be the bandwidth of the channel,
in order to obtain the gain of the channel,
is CiAnd AjK is a fading factor, σijIs the average noise power on the channel.
T time slot CiOffloading tasks to AjEnergy consumption of communication is
Wherein the transmission delay is
1 {. is an indication function, and when "is true, 1 {. is equal to 1, and vice versa is 0. A. thejThe self calculation power resource and the battery energy are consumed during the cooperative processing task, and the tth time slot AjThe computing energy consumption when processing computing tasks is
Wherein, κjIs the effective energy cost factor associated with the chip architecture.
3. Energy collection model
Assuming that each mobile terminal device with EH functionality can obtain renewable energy from the environment for battery powering, the EH process is modeled as a continuous energy packet arrival process, g ═ i denotes the client terminal, g ═ j denotes the proxy terminal, and is used as the proxy terminal
Represents the energy collected in the t-th time slot, satisfies
And are independently and equally distributed in different time slots. Due to the randomness and discontinuity in the renewable energy collection, only part of the collected energy is stored in the battery, and therefore, the energy actually collected by the terminal device is
And satisfy
May be used for local and offload computations starting from the next time slot.
By using
And
a set of EH device energy queues representing a client terminal and a proxy terminal, respectively, wherein,
the energy queue backlog at the beginning of the time slot t for the g-th EH device is, in the normal case,
Ciconsidering only energy consumption of local calculation and transmission process
AjConsidering only the calculated energy consumption
To prevent over-discharge of the battery, battery discharge constraints should be satisfied
Wherein E ismaxAnd EminRepresenting the maximum and minimum battery discharge energies, respectively. In addition, in order to guarantee the endurance of the battery of the mobile terminal, the battery power at the beginning of the time slot t must be greater than the energy consumption required by the mobile terminal
Thus, CiThe energy queue backlog at t +1 time slot is
AjEnergy queue at t +1 time slotIs overstocked as
Second, description of the problem
In the t time slot, a communication cost model is defined as
The energy consumption cost model is
Wherein
Local computation does not generate communication energy consumption for unit communication cost, i.e.
ηijRepresenting the unit energy consumption cost of the agent to process the task. Definition CiThe benefit obtained by the offloading task is
Wherein the content of the first and second substances,
ξiand a task unloading benefit weight parameter which is larger than zero. In the mobile Ad Hoc cloud, the unloading decision of the terminal equipment is constrained by the stability of a task queue, the stability of an energy queue and the unloading time of the task, and the invention provides a long-term profit maximization problem in the time-average sense
Wherein the constraint conditions respectively represent CiThe task unloading amount does not exceed the backlog amount of the task queue, and the stability of the queue is restrained.
Dynamic distributed unloading scheme based on game theory
In the mobile Ad Hoc cloud with the energy collection function, the terminal equipment is affected by factors such as the backlog of task queues, computing resources, battery energy and the like, and the collection of state information of all the mobile equipment is difficult. Therefore, the unloading decision and the price strategy of the terminal equipment are modeled through the Lyapunov optimization theory and the game theory, the agent terminal is used as a seller, and different calculation and storage resources are provided for the client terminal through dynamic resource quotation; the client terminal is used as a buyer, and resources are purchased from the agent terminal according to the calculation task requirement of the client terminal, so that the system income is optimized.
1. Deal game model analysis
Since the battery energy is time dependent, the offloading decisions of the terminal devices at different time slots are coupled. Interference deviceThe dynamic weighting method is a method for effectively solving the problems[16]. In order to eliminate the coupling effect, disturbance parameters and a virtual energy queue of the terminal device are first defined.
Definition 1: disturbance parameter θ of EH devicegIs a bounded constant. Namely, it is
Wherein the content of the first and second substances,
representing the maximum energy consumption of the local computation,
which represents the maximum energy consumption for transmission,
representing the maximum energy consumption calculated by the agent. V is a control parameter and satisfies 0 < V < + ∞.
Definition 2: the virtual energy queue is defined as
Representing the battery energy that the mobile terminal device may actually consume. By adjusting the disturbance parameter thetagAnd a control parameter V such that
Stabilized at a disturbance parameter thetagNearby.
1) Buyer gaming model analysis
Defining the tth time slot CiTo AjThe unit payment cost for purchasing computing resources is
CiPaymentThe cost of is
When the task is processed locally,
Cimainly including the cost of paying the agent and the communication cost of offloading the task to the agent, and therefore, the t-th slot CiHas the optimization target of
In order to balance the relationship between client terminal income and queue backlog, a Lyapunov function is introduced to model the queue backlog of each client terminal, wherein the Lyapunov function is C in the current time slotiNon-negation of task queues and energy queuesScalar representation
Wherein, L [ theta ]i(t)]≥0。
Defining the change of the Lyapunov function between two time slots as the Lyapunov drift
Indicating an increase in queue backlog from the tth slot to the t +1 th slot.
By minimizing Δ [ theta ]i(t)]To ensure L [ theta ]i(t)]The stability of (2). Delta [ theta ]i(t)]Is upper bound by
And (4) showing. Wherein the content of the first and second substances,
and phi is more than or equal to 0.
Lyapunov drift plus penalty is expressed as
The goal of optimizing buyer revenue is to minimize the Lyapunov drift plus penalty function, denoted as the P2 problem
2) Seller game model analysis
For the proxy terminal, the benefit obtained by providing the client terminal with computing resources during the time slot t is
Agents maximize their own revenue through optimal pricing, AjConsidering only the cost of energy consumption for executing the calculation task, the tth time slot AjThe maximum profit problem of
Similarly, the Lyapunov drift of the proxy terminal is added with a penalty function of
Wherein the content of the first and second substances,
and gamma is more than or equal to 0.
Representing seller's optimization problem as P3 problem
2. Buy-sell game policy analysis
1) Buyer policy analysis
The optimization problem of the buyer can be divided into EH optimization and task unloading strategy optimization, and the P2 problem can be converted into two sub-problems Q1 and Q2 because the two variables are independent.
a) Solving for optimal values for energy harvesting
When in use
When the utility model is used, the water is discharged,
when in use
When the temperature of the water is higher than the set temperature,
b) task offload optimization strategy
After decoupling the energy harvesting optimization problem, the task offload optimization can be expressed as
In order to maximize the benefit of the client terminal, the queue backlog theta is calculated according to the control parameter Vi(t) and seller's quote, etc. to determine the purchase strategy. When the computing task is processed locally, when
When the temperature of the water is higher than the set temperature,
thus, it is possible to provide
Is about
Because the constraint condition is an affine function, the optimal local calculation frequency of the tth time slot is solved by using a Lagrange multiplier method and a KKT condition
Wherein, the first and the second end of the pipe are connected with each other,
when in use
When the temperature of the water is higher than the set temperature,
is a about
A monotonically increasing function of (1), the optimum local computation frequency being
The battery energy, the resources and the calculation task types of each client terminal are different, and the difference of the calculation resources, the geographical positions and the battery electric quantity among the agent terminals can cause the difference of the agent quotation. When computing tasks are offloaded to agents, battery discharge constraints and vendor quotes can affect the offload decision of the client terminal, so the client terminal will select a reasonable agent and offload the appropriate task at the beginning of each timeslot. The maximum and minimum unloading task amount can be obtained as
And
wherein the content of the first and second substances,
when in use
AjIs satisfied by unit resource quotation
Time, profit of the client terminal
With the amount of offloaded tasks
Is increased. Therefore, when
And is
The client will only offload tasks to proxy execution.
To pair
Calculating a second order partial derivative
The optimal unloading task allocation of the t time slot can be obtained by applying the Lagrange multiplier method and the KKT condition as
Wherein the content of the first and second substances,
2) seller policy analysis
Unit resource quotation of agent terminal
The higher the agent profit, the higher the cost that the client terminal needs to pay, however, the purchase will be reduced, which will result in a reduction in the profit of the agent terminal. Thus, where the broker has an optimal bid to balance the buyer's and seller's revenue, the P3 problem may be divided into EH optimization and resource bid optimization problems Q3 and Q4.
a) Solving for optimal values for energy harvesting
When the temperature is higher than the set temperature
When the temperature of the water is higher than the set temperature,
when the temperature is higher than the set temperature
When the utility model is used, the water is discharged,
b) resource quotation optimization strategy
After decoupling the energy harvesting optimization problem, the resource quote optimization can be expressed as
AjShould be greater than zero, so that
Price quoted
Is the minimum value of the quote, i.e.
When the temperature is higher than the set temperature
In time, there are:
therefore, the number of the first and second electrodes is increased,
is about
The constraint condition is affine function, and the optimal quotation of the t time slot is obtained by Lagrange multiplier method and KKT condition
Stackelberg equilibrium analysis
According to the analysis, in order to maximize the benefit of the client terminal, the client terminal can make an unloading decision according to the attribute of the agent and the channel resource, and select a proper agent to cooperatively calculate the unloading task; and the agent can decide the optimal resource pricing according to the relation between the task amount of the client terminal and the income of the agent. The optimal solution is then demonstrated
Is the SE equalization solution.
Definition 3: when acting on the quote of the terminal
At a certain time, the temperature of the liquid crystal display panel is controlled,
when the client terminal unloads the task
At a certain time, the temperature of the liquid crystal display panel is controlled,
if it is
Is SE equilibrium solution, then
This is demonstrated by the following three lemmas.
Introduction 1: when acting on the quote of the terminal
At a given time, the benefit of the client terminal
In that
The maximum value is obtained.
And (3) proving that: due to the fact that
Therefore, it is not only easy to use
Is about
And thus, the profit of the client terminal
In that
The maximum value is obtained. According to the definition 3, the first and second,
is SE equilibrium solution
After the certificate is confirmed
2, leading: when the client terminal unloads the task
At a given time, the benefit of the terminal is proxied
In that
The maximum value is obtained.
And (3) proving that: when proxy quote
When the temperature of the water is higher than the set temperature,
therefore, it is not only easy to use
Is about
And thus, the benefit of the proxy terminal
In that
The maximum value is obtained. According to the definition 3, the first and second,
is SE equilibrium solution
Introduction and management3: optimal offloading tasks for client terminals
With the quotation of the agent terminal
Is increased and decreased.
And (3) proving that: optimum value
To pair
To obtain a first order partial derivative
After the certificate is confirmed
Therefore, the temperature of the molten metal is controlled,
for SE equalisation, i.e.
Fourth, simulation result and analysis
The invention carries out simulation analysis on dynamic quotation of resources, battery energy, queue backlog, system income and the like. The method provided by the invention verifies the effectiveness of the method provided by the invention through simulation comparison with a local execution scheme, a centralized scheme and a greedy algorithm.
1. Simulated scene setting
The main simulation parameters are set as follows, the utility gain factor xi i2, the minimum CPU cycle frequency f of the client terminali min100MHz, maximum CPU cycle frequency fi max2000 MHz; minimum CPU cycle frequency of proxy terminal
Maximum CPU cycle frequency fi max6000 MHz. Control parameter V e (0,600)]Unit communication cost of
Cost per unit energy consumption etaij∈[1×10-7,8×10-7]The task processing density L is equal to [800,1500 ]]cycle/Mbit, energy harvesting Power Pi h∈[5,30]mw，
The unit time slot length tau is 1s, and the price updating step length is 5 x 10-7。
Figures 3 and 4 describe the gaming process where the mobile terminal CPU calculates the frequency and dynamic price. Suppose that the system has 1 client terminal and 4 agent terminals, and the unit communication cost and the energy consumption cost are respectively rho1＝0.06，η1＝1×10-7，Vj1＝100，ρ2＝0.12，η2＝2×10-7，Vj2＝150，ρ3＝0.18，η3＝4×10-7，Vj3＝200，ρ4＝0.24，η4＝8×10-7，Vj4＝250。
As can be seen from fig. 3, the proxy bid increases as the number of iterations increases until it converges to the optimal bid. As can be seen from fig. 4, the cost price of the initial agent is low, and the frequency of the CPU purchased by the client terminal increases first as the number of iterations increases. As the proxy offer increases, the customer's willingness to purchase the CPU frequency decreases, the purchased CPU frequency decreases as the number of iterations increases, and finally, as the proxy offer becomes stable, the purchased CPU frequency also becomes stable. Further, the higher the unit communication cost and the unit calculation cost of the agent terminal, the higher the dynamic resource offer, and the less the CPU frequency purchased by the client terminal.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (9)
1. A distributed unloading method based on energy collection in a mobile Ad Hoc cloud is characterized by comprising the following steps:
considering a mobile Ad Hoc cloud network formed by a group of nearby terminal devices with EH functions, respectively establishing a calculation task model, a task unloading model and an energy collection model;
the method comprises the steps that a client terminal is used as a buyer, resources are purchased from an agent terminal according to the computing task demand of the client terminal, the Lyapunov optimization theory is adopted, and the profit maximization problem of the buyer is established on the basis of a computing task model, a task unloading model and an energy collection model;
taking the agent terminal as a seller, providing different computing and storage resources for the client terminal through dynamic resource quotation, and establishing the profit maximization problem of the seller based on a computing task model, a task unloading model and an energy collection model;
according to the task backlog of the client terminal, the battery energy level and the quotation of the proxy terminal, in each time slot, calculating an optimal task unloading strategy unloaded by the proxy terminal selected by the buying direction and an optimal quotation strategy of a seller by using a Lagrange multiplier method and a KKT condition;
and if the optimal task unloading strategy of the buyer and the optimal quotation strategy of the seller meet the SteinKelberg equilibrium solution, the client terminal unloads the tasks to the agent terminal according to the optimal task unloading strategy.
2. The distributed unloading method based on energy collection in the mobile Ad Hoc cloud of claim 1, wherein the establishment of the buyer's profit maximization problem based on the calculation task model, the task unloading model and the energy collection model is represented as:
constraint conditions are as follows:
wherein the content of the first and second substances,
representing a profit maximization problem of the buyer at the t-th time slot; viTo representControl parameters of the ith client terminal;
representing the total profit of the ith client terminal in the time slot t;
indicating that the unloading profit is related to the task queue;
for the ith client terminal CiThe amount of task arrivals in time slot t,
indicates the ith client terminal CiThe total amount of tasks processed in time slot t;
the virtual energy queue of the EH device for the ith client terminal is represented as
θiAs a parameter of the disturbance of the EH device,
backlog the energy queue of the EH equipment with the ith client terminal at the beginning of the time slot t; eminRepresents a minimum discharge energy of the battery;
representing the total energy consumption generated by the ith client terminal in the time slot t; emaxRepresents the maximum discharge energy of the battery;
represents the jth agent terminal A in the tth time slotjAssisting the ith client terminal in calculating the calculation energy consumption generated during the task;
representing the energy queue backlog of the EH equipment carried by the ith client terminal at the beginning of the time slot t;
the task amount of an ith client terminal to be unloaded to a jth agent terminal in a tth time slot is represented; qi tThe backlog of the task queue of the ith client terminal at the t time slot is represented;
the average task queue backlog of the ith client terminal in the t time slot is represented; t represents the system running time;
indicating the desire.
3. The distributed unloading method based on energy collection in the mobile Ad Hoc cloud according to claim 2, wherein the buyer's profit maximization problem established based on the computation task model, the task unloading model and the energy collection model is decomposed into an optimal solution for solving energy collection and a task unloading optimization problem, wherein the optimal solution for solving energy collection is expressed as:
min:
when the temperature is higher than the set temperature
When the temperature of the water is higher than the set temperature,
when the temperature is higher than the set temperature
When the utility model is used, the water is discharged,
after decoupling the energy harvesting optimization problem, the task offloading optimization problem is represented as:
constraint conditions are as follows:
wherein the content of the first and second substances,
represents an optimal solution for energy harvesting; gamma raymaxRepresenting the maximum amount of energy collected by the EH device at the tth time slot.
4. The distributed offloading method for energy harvesting in a mobile Ad Hoc cloud of claim 3, wherein when the computing task is processed locally and the energy harvesting is performed locally
And then, solving the optimal local calculation frequency of the t-th time slot by using a Lagrange multiplier method and a KKT condition as follows:
wherein the content of the first and second substances,
κian effective energy cost coefficient for the ith client terminal chip; l is a radical of an alcoholiA unit processing capacity of an ith client terminal; tau is the unit time slot length; xiiOffloading the benefit weight parameter for the task.
5. The distributed offloading method for energy harvesting in a mobile Ad Hoc cloud of claim 3, wherein when the computing task is processed locally and the energy harvesting is performed locally
The optimal local computation frequency is:
wherein the content of the first and second substances,
indicating when a computing task is processed locally and
time-optimal local computation frequency;
represents the maximum CPU processing frequency of the ith client;
represents the maximum discharge energy of the battery of the ith client;
the optimal task amount of the ith client terminal to be unloaded to the jth agent terminal in the tth time slot is represented;
the transmission power of the ith client terminal for the t time slot for unloading the task to the jth agent terminal;
a task transmission rate for offloading the task to the jth agent terminal for the ith client terminal at the tth time slot; kappaiAn effective energy cost coefficient for the ith client terminal chip; τ is the unit slot length.
6. The distributed unloading method based on energy collection in the mobile Ad Hoc cloud of claim 3, wherein the client terminal selects to unload part of tasks to the agent terminal at the beginning of each time slot, the rest of tasks are unloaded locally, and the optimal task amount unloaded by the ith client terminal to the jth agent terminal at the tth time slot is expressed as:
wherein ξiOffloading a benefit weight parameter for the task;
for the ith time slot, the ith client terminal CiTo the jth agent terminal AjA unit payment cost for purchasing computing resources;
unit communication cost for offloading a task to a jth agent terminal for the ith client terminal of the t slot;
the transmission power of the ith client terminal for the t time slot to unload the task to the jth agent terminal;
the task transmission rate at which the ith client terminal offloads tasks to the jth proxy terminal for the t slot.
7. The distributed unloading method based on energy collection in the mobile Ad Hoc cloud according to claim 1, wherein the seller's profit maximization problem is established based on the calculation task model, the task unloading model and the energy collection model and is expressed as:
an objective function:
wherein the content of the first and second substances,
representing a seller's revenue maximization problem; v represents a control parameter;
representing the total income of the jth agent terminal in the tth time slot;
the virtual energy queue of the jth agent terminal for the tth time slot is represented as
θjAs a parameter of the disturbance of the EH device,
the energy queue backlog of the EH equipment carried by the jth agent terminal at the beginning of the time slot t is carried out;
representing the energy actually collected by the EH equipment carried by the jth proxy terminal in the time slot t;
jth proxy terminal A in tth time slotjComputing energy consumption in processing a computing task from an i-th client terminal; eminRepresents a minimum discharge energy of the battery; emaxRepresents the maximum discharge energy of the battery;
for the ith time slot, the ith client terminal CiTo the jth agent terminal AjThe unit of the purchased computing resource pays a cost.
8. The distributed unloading method based on energy collection in the mobile Ad Hoc cloud according to claim 7, wherein the seller-based yield maximization problem established based on the computation task model, the task unloading model and the energy collection model is decomposed into an optimal solution for solving energy collection and a task unloading optimization problem, wherein the optimal solution for solving energy collection is expressed as:
min:
when in use
When the temperature of the water is higher than the set temperature,
when in use
When the temperature of the water is higher than the set temperature,
after decoupling the energy collection optimization problem, the task offloading optimization problem is represented as;
constraint conditions are as follows:
wherein the content of the first and second substances,
represents an optimal solution for energy harvesting; gamma raymaxThe representation represents the maximum amount of energy collected by the EH device at the tth time slot.
9. The distributed unloading method based on energy collection in the mobile Ad Hoc cloud according to claim 8, wherein the optimal quotation of the t-th time slot obtained by the Lagrangian multiplier method and the KKT condition is as follows:
wherein, the first and the second end of the pipe are connected with each other,
for the ith client terminal CiTo the jth agent terminal AjPurchasing an optimal unit payment cost for the computing resource;
for the ith client terminal CiThe optimal unloading task amount;
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210269015.2A CN114650568A (en) | 2022-03-18 | 2022-03-18 | Distributed unloading method based on energy collection in mobile Ad Hoc cloud |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210269015.2A CN114650568A (en) | 2022-03-18 | 2022-03-18 | Distributed unloading method based on energy collection in mobile Ad Hoc cloud |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114650568A true CN114650568A (en) | 2022-06-21 |
Family
ID=81994880
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210269015.2A Pending CN114650568A (en) | 2022-03-18 | 2022-03-18 | Distributed unloading method based on energy collection in mobile Ad Hoc cloud |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114650568A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115843070A (en) * | 2023-02-23 | 2023-03-24 | 山东省计算中心（国家超级计算济南中心） | Ocean sensor network computing unloading method and system based on energy collection technology |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170013495A1 (en) * | 2015-07-10 | 2017-01-12 | Lg Electronics Inc. | Method and apparatus for an input data processing via a local computing or offloading based on power harvesting in a wireless communication system |
| CN107295109A (en) * | 2017-08-16 | 2017-10-24 | 重庆邮电大学 | Task unloading and power distribution joint decision method in self-organizing network cloud computing |
| CN110113195A (en) * | 2019-04-26 | 2019-08-09 | 山西大学 | A kind of method of joint unloading judgement and resource allocation in mobile edge calculations system |
| CN113064665A (en) * | 2021-03-18 | 2021-07-02 | 四川大学 | Multi-server computing unloading method based on Lyapunov optimization |
| CN113114733A (en) * | 2021-03-24 | 2021-07-13 | 重庆邮电大学 | Distributed task unloading and computing resource management method based on energy collection |
-
2022
- 2022-03-18 CN CN202210269015.2A patent/CN114650568A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170013495A1 (en) * | 2015-07-10 | 2017-01-12 | Lg Electronics Inc. | Method and apparatus for an input data processing via a local computing or offloading based on power harvesting in a wireless communication system |
| CN107295109A (en) * | 2017-08-16 | 2017-10-24 | 重庆邮电大学 | Task unloading and power distribution joint decision method in self-organizing network cloud computing |
| CN110113195A (en) * | 2019-04-26 | 2019-08-09 | 山西大学 | A kind of method of joint unloading judgement and resource allocation in mobile edge calculations system |
| CN113064665A (en) * | 2021-03-18 | 2021-07-02 | 四川大学 | Multi-server computing unloading method based on Lyapunov optimization |
| CN113114733A (en) * | 2021-03-24 | 2021-07-13 | 重庆邮电大学 | Distributed task unloading and computing resource management method based on energy collection |
Non-Patent Citations (2)
| Title |
|---|
| SHICHAO XIA: "Online Distributed Offloading and Computing Resource Management With Energy Harvesting for Heterogeneous MEC-Enabled IoT", IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL.20.NO.10，OCTOBER 2021, 5 March 2021 (2021-03-05) * |
| 李长云； 黎建波； 徐曦； 李亭立: "具有能量收集设备的移动边缘计算资源分配研究", 系统仿真学报第34卷第11期, 14 September 2021 (2021-09-14) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115843070A (en) * | 2023-02-23 | 2023-03-24 | 山东省计算中心（国家超级计算济南中心） | Ocean sensor network computing unloading method and system based on energy collection technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xia et al. | Online distributed offloading and computing resource management with energy harvesting for heterogeneous MEC-enabled IoT | |
| CN113114733B (en) | Distributed task unloading and computing resource management method based on energy collection | |
| CN112512056B (en) | Multi-objective optimization calculation unloading method in mobile edge calculation network | |
| CN107295109A (en) | Task unloading and power distribution joint decision method in self-organizing network cloud computing | |
| CN113810233B (en) | Distributed computation unloading method based on computation network cooperation in random network | |
| CN111262944B (en) | Method and system for hierarchical task offloading in heterogeneous mobile edge computing network | |
| CN111401744B (en) | Dynamic task unloading method in uncertainty environment in mobile edge calculation | |
| Siew et al. | Dynamic pricing for resource-quota sharing in multi-access edge computing | |
| CN111949409A (en) | Method and system for unloading calculation tasks in electric wireless heterogeneous network | |
| Huang et al. | Market-based dynamic resource allocation in Mobile Edge Computing systems with multi-server and multi-user | |
| Liu et al. | Multi-item auction based mechanism for mobile data offloading: A robust optimization approach | |
| CN114650568A (en) | Distributed unloading method based on energy collection in mobile Ad Hoc cloud | |
| CN113747450B (en) | Service deployment method and device in mobile network and electronic equipment | |
| Singh et al. | Profit optimization for mobile edge computing using genetic algorithm | |
| CN114449490A (en) | Multi-task joint computing unloading and resource allocation method based on D2D communication | |
| Lyu et al. | Multi-leader multi-follower Stackelberg game based resource allocation in multi-access edge computing | |
| Chang et al. | Flat-rate pricing for green edge computing with latency guarantee: A Stackelberg game approach | |
| Binh et al. | A reinforcement learning algorithm for resource provisioning in mobile edge computing network | |
| Tütüncüoğlu et al. | Optimal pricing for service caching and task offloading in edge computing | |
| Wang et al. | PSOGT: PSO and game theoretic based task allocation in mobile edge computing | |
| Zu et al. | An auction-based mechanism for task offloading in fog networks | |
| Wang et al. | Auction pricing-based task offloading strategy for cooperative edge computing | |
| Xu et al. | BASIC: Distributed task assignment with auction incentive in UAV-enabled crowdsensing system | |
| Su et al. | Game-based pricing and task offloading in mobile edge computing enabled edge-cloud systems | |
| Zhang et al. | System revenue maximization for offloading decisions in mobile 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 |