CN112423320A - Multi-user computing unloading method based on QoS and user behavior prediction - Google Patents

Multi-user computing unloading method based on QoS and user behavior prediction Download PDF

Info

Publication number
CN112423320A
CN112423320A CN202011253694.1A CN202011253694A CN112423320A CN 112423320 A CN112423320 A CN 112423320A CN 202011253694 A CN202011253694 A CN 202011253694A CN 112423320 A CN112423320 A CN 112423320A
Authority
CN
China
Prior art keywords
task
user
calculation
terminal equipment
qos
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
Application number
CN202011253694.1A
Other languages
Chinese (zh)
Inventor
李志刚
张睿
杨凯
陈光晓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Northwestern Polytechnical University
Original Assignee
Northwestern Polytechnical University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Northwestern Polytechnical University filed Critical Northwestern Polytechnical University
Priority to CN202011253694.1A priority Critical patent/CN112423320A/en
Publication of CN112423320A publication Critical patent/CN112423320A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/16—Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/16—Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]
    • H04W28/24—Negotiating SLA [Service Level Agreement]; Negotiating QoS [Quality of Service]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Abstract

The invention provides a multi-user computing unloading method based on QoS and user behavior prediction, which comprises the following steps: the method comprises the following steps: transmitting the calculation tasks to be processed to the terminal equipment; step two: the terminal equipment completes calculation unloading decision, then executes calculation task, and finally returns processing result. And the unloading decision of the calculation task is carried out by comprehensively considering the current and future behaviors of the user within a period of time, so that the execution time of the calculation task is optimized, the energy consumption of a terminal is optimized, and the privacy leakage risk brought by long-distance transmission is reduced. The overhead of the calculation task executed in the edge server and the local is comprehensively considered, the calculation resource is reasonably distributed, and the QoS (quality of service) provided for the user is improved.

Description

Multi-user computing unloading method based on QoS and user behavior prediction
Technical Field
The invention belongs to the field of computational data processing, and particularly relates to a multi-user computation unloading method based on QoS and user behavior prediction.
Background
With the rapid development of communication technology, smart mobile devices have become an indispensable important part in social life, and have spawned a large number of high-quality services and applications, such as automatic driving, augmented reality, and the like. Data required by various applications are generated by data acquisition nodes distributed on terminal equipment such as a smart phone and the like and nearby, the data are collected and then gathered to the terminal equipment, and computing tasks applied by the terminal equipment and the terminal equipment need to be processed. And in the data processing stage, the computing task can be sent to the cloud computing center for processing besides local execution.
The cloud computing center has strong computing power and large-scale computing equipment, and provides abundant network and computing services for the terminal equipment. However, the cloud computing center is usually deployed at a place far away from the user, and the long transmission delay is needed when the computing task is unloaded to the cloud computing center, which is difficult to be applied to the computing task and application with strong real-time performance. Computing tasks from the outside of user terminal equipment and the terminal are transmitted to a cloud computing center through a network for computing and storing, so that the problems of large delay, high bandwidth occupation, privacy disclosure caused by long-distance transmission and the like are caused.
Compared with the computing unloading of a cloud computing center, the unloading technology in the mobile edge computing unloads the computing task to the edge network, so that the defects of the terminal equipment in the aspects of resource storage, computing performance and the like can be overcome, and the processing of the edge server and the terminal equipment which are close to a user can meet the time delay requirement of the computing task with high real-time performance.
However, since the energy consumption of the terminal, the completion time of the computing task, the load of the terminal itself, and other factors all affect the user experience, considering only the computing offloading method aiming at reducing the energy consumption may cause the completion time of the computing task not to meet the delay requirement thereof and directly affect the user experience and quality of service (QoS).
Disclosure of Invention
Aiming at the defects, the invention provides a multi-user computing unloading method based on user behavior modeling and prediction, which can well meet the requirements of computing tasks on QoS (quality of service) indexes such as response delay, terminal equipment energy consumption and the like, can optimize the response delay of the computing tasks, and can prolong the service time of terminal equipment such as smart phones and the like.
The technical scheme of the invention is as follows: a multi-user computational offload method based on QoS and user behavior prediction, the method comprising the steps of: the method comprises the following steps: transmitting the calculation tasks to be processed to the terminal equipment; step two: the terminal equipment completes calculation unloading decision, then executes calculation task, and finally returns processing result.
Further, a multi-user computing offloading method based on QoS and user behavior prediction, wherein the computing task to be processed in the first step comprises: computing tasks from outside the terminal equipment and computing tasks applied by the terminal equipment; and the computing tasks except the terminal equipment are transmitted to the terminal equipment through the network.
Further, in the second step, only one computation task is considered at a time when the computation offload decision is made.
Further, the computation unloading decision result of each computation task is executed locally or completely unloaded to an edge server for execution.
Further, the unloading to the edge server comprises three stages, namely uploading, processing and descending, specifically, a task is executed under an unloading decision, the terminal equipment firstly uploads data to a wireless access point and then uploads the data to the edge server for processing, and the last edge server returns a result to the terminal equipment.
Further, the computation decision is made according to computation resources required by the computation task, computation cost of the task executed locally and computation cost of the task executed at an edge server.
Further, the multi-user calculation unloading method based on QoS and user behavior prediction optimizes calculation unloading through constraint conditions during calculation unloading decision, wherein the constraint conditions comprise edge server residual equipment resources and current equipment residual equipment resources.
The invention has the beneficial effects that: according to the invention, the unloading decision of the calculation task is carried out by comprehensively considering the current and future behaviors of the user, so that the execution time of the calculation task is optimized, the energy consumption of the terminal is optimized, and the privacy disclosure risk caused by long-distance transmission is reduced. The overhead of the calculation task executed in the edge server and the local is comprehensively considered, the calculation resource is reasonably distributed, and the QoS (quality of service) provided for the user is improved.
Drawings
FIG. 1 is a scene architecture diagram of a multi-user computation offload method based on QoS and user behavior prediction according to the present invention.
Detailed Description
The technical solution of the present invention is further described below with reference to the accompanying drawings: as shown in fig. 1, a method for multi-user computational offloading based on QoS and user behavior prediction, the method comprising the steps of:
the method comprises the following steps: transmitting the calculation tasks to be processed to the terminal equipment;
the computing tasks needing to be processed comprise: computing tasks from outside the terminal equipment and computing tasks applied by the terminal equipment; and the computing tasks except the terminal equipment are transmitted to the terminal equipment through the network.
Step two: the terminal equipment completes calculation unloading decision, then executes calculation task, and finally returns processing result. And the calculation unloading decision result of each calculation task is executed locally or completely unloaded to the edge server for execution.
The unloading to the edge server comprises three stages, namely uploading, processing and descending, specifically, a task is executed under an unloading decision, the terminal equipment firstly uploads data to the wireless access point, then uploads the data to the edge server for processing, and finally the edge server returns a result to the terminal equipment.
The computation decision makes a computation offload decision by computing resources required for the task, computation cost of the task being executed locally, computation cost of the task being executed at the edge server.
The computation offload decision considers only one computation task at a time.
The computational decision phase is described in detail below, and the present invention considers the case of multi-user (i.e., multi-device), single AP, single edge server, with N devices { d }1，d2，d3，…，dNAnd the edge server is connected with the AP. Only one computing task is considered at a time in the computing offload phase, and each computing task can only be executed locally or offloaded to the edge server entirely. Edge servers are resource-limited and therefore do not allow all devices to offload all their computing tasks simultaneously to an edge server.
Calculating a task model: for each task
Figure BDA0002772434800000041
di can only be executed locally or offloaded to the edge server for execution.
Figure BDA0002772434800000051
By(s)i,ci,mi) To indicate.
And (3) communication model: the signal to interference and noise ratio of the signal received by the AP may be expressed as
Figure BDA0002772434800000052
Calculating a model: for a computing task
Figure BDA0002772434800000053
Its processing policy is either local execution or off-load to the edge server execution.
(1) Local execution
Latency of local execution: let Tl iRepresenting delay of local execution, fl iRepresenting the cycle frequency of the CPU (i.e., CPU cycles per second of execution), then the local execution latency can be expressed as
Figure BDA0002772434800000054
Processing tasks
Figure BDA0002772434800000055
With an energy consumption of
Figure BDA0002772434800000056
Thus, the task is executed locally
Figure BDA0002772434800000057
Has a total cost of
Figure BDA0002772434800000058
Where α, (1- α) is the weight of execution delay and energy consumption. Alpha is more than or equal to 0 and less than or equal to 1.
(2) Execute after uninstall
Performing tasks under offload decisions
Figure BDA0002772434800000059
diThe data is first uploaded to the AP and then to the edge server for processing. The edge server then allocates resources and then performs the computational task
Figure BDA00027724348000000510
Finally, the AP returns the result to di. The execution after unloading is divided into three stages, namely uploading, processing and descending.
Uploading:
order to
Figure BDA00027724348000000511
Denotes diThe data rate of the uplink radio link to the AP, w represents the bandwidth,
Figure BDA00027724348000000512
(according to the shannon formula). Where Si is the average power of the transmitted signal within the channel and Ni represents the gaussian noise power within the channel. Specifically, there are
Figure BDA00027724348000000513
From diThe time delay for uploading to the AP is expressed as
Figure BDA0002772434800000061
Energy consumption
Figure BDA0002772434800000062
And (3) treatment:
edge server processing computing tasks
Figure BDA0002772434800000063
At a time of
Figure BDA0002772434800000064
The energy consumption for normal operation of the mobile phone is
Figure BDA0002772434800000065
Descending:
the result goes down to diAt a time of
Figure BDA0002772434800000066
Wherein ω isiIs the size of the data of the result,
Figure BDA0002772434800000067
representing from AP to diThe downlink data rate. Let diThe power required for downloading the execution result is
Figure BDA0002772434800000068
The energy consumption of the reception processing result is expressed as
Figure BDA0002772434800000069
The total time delay and the energy consumption are respectively
Figure BDA00027724348000000610
Figure BDA00027724348000000611
The computational cost of the offload is
Figure BDA00027724348000000612
The total cost of all equipment is expressed as
Figure BDA00027724348000000613
In addition, in order to optimize the calculation decision, the invention firstly describes the problem, and then sets the constraint conditions:
Minimize Ot
s.t.
Figure BDA00027724348000000614
(requirement for time delay)
0≤fi≤R，
Figure BDA00027724348000000615
(requirement for resources)
Figure BDA00027724348000000616
(requirement for resources)
The symbols in the above description have the following meanings:
AP-Wireless Access Point;
n is the number of terminal devices;
di-the ith terminal device;
Figure BDA00027724348000000712
-the ith calculation task;
Si-the size of the i-th computation task input data;
ci-the number of CPU cycles required to process the ith computation task;
mi-maximum tolerated delay for the ith computation task;
SINRi-the AP receives the signal to interference plus noise ratio of the signal transmitted by the ith device;
Figure BDA0002772434800000071
-latency of local execution of the ith computation task;
Figure BDA0002772434800000072
-latency of local execution of the ith computation task;
Figure BDA0002772434800000073
-the cycle frequency of the CPU;
Figure BDA0002772434800000074
-locally executing the task epsiloniThe total cost of;
Figure BDA0002772434800000075
-processing task epsiloniEnergy consumption of (2);
k-energy efficiency coefficient (constant associated with the chip itself);
α — weight of execution latency;
Figure BDA0002772434800000076
-data rate of the uplink radio link to the AP;
w-bandwidth of channel
Si-the average power of the transmission signal in the channel;
Ni-Gaussian noise power within the channel;
Figure BDA0002772434800000077
time delay for uploading of ith computation task to AP；
Figure BDA0002772434800000078
-time delay for the ith computation task to upload to the AP;
Figure BDA0002772434800000079
-edge server processing computing tasks
Figure BDA00027724348000000710
The time of (d);
Figure BDA00027724348000000711
energy consumption of normal operation of the mobile phone;
Figure BDA0002772434800000081
energy consumption of normal operation of the mobile phone;
Figure BDA0002772434800000082
-the result goes down to diThe time of (d);
ωi-data size of the result;
Figure BDA0002772434800000083
-data rate from AP to down to ith terminal device;
di-the power required for downloading the execution results is
Figure BDA0002772434800000084
Figure BDA0002772434800000085
-receiving the energy consumption of the processing result;
Figure BDA0002772434800000086
-i < th > computation task total offload delay
Figure BDA0002772434800000087
-the ith calculation task total unloading energy consumption
Figure BDA0002772434800000088
Computation cost of the i-th computation task offload
R-computing resources provided by edge servers
OtTotal cost of all equipment
And optimizing the calculation unloading through constraint conditions during the calculation unloading decision, wherein the constraint conditions comprise the residual equipment resources of the edge server and the residual equipment resources of the current equipment.
Constraint conditions are as follows: re, Rd
Figure BDA0002772434800000089
Figure BDA0002772434800000091
The symbols in the above description have the following meanings:
re-edge Server residual device resources
Rd-device remaining device resources
S-offload decision
cur _ Ed-current remaining capacity of equipment
Alpha-weight of remaining charge
Ti _ r-computing resources required by task i
Theta-cost of task execution locally
O-the computational cost of the task being performed at the edge server.

Claims (7)

1. A multi-user computing unloading method based on QoS and user behavior prediction is characterized in that: the method comprises the following steps:
the method comprises the following steps: transmitting the calculation tasks to be processed to the terminal equipment;
step two: the terminal equipment completes calculation unloading decision, then executes calculation task, and finally returns processing result.
2. The method of claim 1, wherein the QoS and user behavior prediction based multi-user computing offload is as follows: the computing task needing to be processed in the first step comprises the following steps: computing tasks from outside the terminal equipment and computing tasks applied by the terminal equipment; and the computing tasks except the terminal equipment are transmitted to the terminal equipment through the network.
3. The method of claim 1, wherein the QoS and user behavior prediction based multi-user computing offload is as follows: in the second step, only one calculation task is considered each time when the unloading decision is calculated.
4. The method of claim 1, wherein the QoS and user behavior prediction based multi-user computing offload is as follows: and the calculation unloading decision result of each calculation task is executed locally or completely unloaded to the edge server for execution.
5. The method of claim 5, wherein the QoS and user behavior prediction based multi-user computing offload is as follows: the unloading to the edge server comprises three stages, namely uploading, processing and descending, specifically, a task is executed under an unloading decision, the terminal equipment firstly uploads data to the wireless access point, then uploads the data to the edge server for processing, and finally the edge server returns a result to the terminal equipment.
6. The method of claim 1, wherein the QoS and user behavior prediction based multi-user computing offload is as follows: the computation decision makes a computation offload decision by computing resources required for the task, computation cost of the task being executed locally, computation cost of the task being executed at the edge server.
7. The method of claim 6, wherein the QoS and user behavior prediction based multi-user computational offloading is performed by: and optimizing the calculation unloading through constraint conditions during the calculation unloading decision, wherein the constraint conditions comprise the residual equipment resources of the edge server and the residual equipment resources of the current equipment.
CN202011253694.1A 2020-11-11 2020-11-11 Multi-user computing unloading method based on QoS and user behavior prediction Pending CN112423320A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011253694.1A CN112423320A (en) 2020-11-11 2020-11-11 Multi-user computing unloading method based on QoS and user behavior prediction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011253694.1A CN112423320A (en) 2020-11-11 2020-11-11 Multi-user computing unloading method based on QoS and user behavior prediction

Publications (1)

Publication Number Publication Date
CN112423320A true CN112423320A (en) 2021-02-26

Family

ID=74781423

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011253694.1A Pending CN112423320A (en) 2020-11-11 2020-11-11 Multi-user computing unloading method based on QoS and user behavior prediction

Country Status (1)

Country Link
CN (1) CN112423320A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113518122A (en) * 2021-06-17 2021-10-19 中南林业科技大学 Task unloading method, device, equipment and medium for ensuring low-delay transmission by edge intelligent network
CN113778550A (en) * 2021-08-17 2021-12-10 南京邮电大学 Task unloading system and method based on mobile edge calculation

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109240818A (en) * 2018-09-04 2019-01-18 中南大学 Task discharging method based on user experience in a kind of edge calculations network

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109240818A (en) * 2018-09-04 2019-01-18 中南大学 Task discharging method based on user experience in a kind of edge calculations network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
杨天，田霖，等: "移动边缘计算中基于用户体验的计算卸载方案", 《CNK-计算机工程》 *
陈龙险: "移动边缘计算中高能效任务卸载决策", 《CNKI-信息技术》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113518122A (en) * 2021-06-17 2021-10-19 中南林业科技大学 Task unloading method, device, equipment and medium for ensuring low-delay transmission by edge intelligent network
CN113778550A (en) * 2021-08-17 2021-12-10 南京邮电大学 Task unloading system and method based on mobile edge calculation
CN113778550B (en) * 2021-08-17 2022-05-03 南京邮电大学 Task unloading system and method based on mobile edge calculation

Similar Documents

Publication Publication Date Title
CN109814951B (en) Joint optimization method for task unloading and resource allocation in mobile edge computing network
CN113612843B (en) MEC task unloading and resource allocation method based on deep reinforcement learning
CN111240701B (en) Task unloading optimization method for end-side-cloud collaborative computing
CN112492626B (en) Method for unloading computing task of mobile user
CN108920279B (en) Mobile edge computing task unloading method under multi-user scene
CN108934002B (en) Task unloading method based on D2D communication cooperation
CN109951869B (en) Internet of vehicles resource allocation method based on cloud and mist mixed calculation
CN111372314A (en) Task unloading method and task unloading device based on mobile edge computing scene
CN110096362B (en) Multitask unloading method based on edge server cooperation
Ti et al. Computation offloading leveraging computing resources from edge cloud and mobile peers
Cao et al. An energy-optimal offloading algorithm of mobile computing based on HetNets
CN109756912B (en) Multi-user multi-base station joint task unloading and resource allocation method
CN110545584A (en) Communication processing method of full-duplex mobile edge computing communication system
CN112416603B (en) Combined optimization system and method based on fog calculation
CN113727362B (en) Unloading strategy method of wireless power supply system based on deep reinforcement learning
CN112423320A (en) Multi-user computing unloading method based on QoS and user behavior prediction
CN112738822A (en) NOMA-based security offload and resource allocation method in mobile edge computing environment
CN115473896A (en) Electric power internet of things unloading strategy and resource configuration optimization method based on DQN algorithm
CN111796880A (en) Unloading scheduling method for edge cloud computing task
CN113365290B (en) Game theory calculation unloading method based on greedy strategy in world fusion network
CN115811793B (en) Intelligent reflector-assisted energy acquisition mobile edge calculation method
Ha et al. Computation capacity constrained joint transmission design for c-rans
CN114615705B (en) Single-user resource allocation strategy method based on 5G network
CN114025359B (en) Resource allocation and calculation unloading method, system, equipment and medium based on deep reinforcement learning
CN111372268B (en) Method for optimizing joint wireless resource management and task unloading in ultra-dense network

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20210226