CN117156494A - Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication - Google Patents

Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication Download PDF

Info

Publication number
CN117156494A
CN117156494A CN202311422980.XA CN202311422980A CN117156494A CN 117156494 A CN117156494 A CN 117156494A CN 202311422980 A CN202311422980 A CN 202311422980A CN 117156494 A CN117156494 A CN 117156494A
Authority
CN
China
Prior art keywords
ris
user
mec
task
scheduling
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202311422980.XA
Other languages
Chinese (zh)
Other versions
CN117156494B (en
Inventor
许斌
刘丹
程龙刚
柴金铭
庄智超
亓晋
董振江
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Posts and Telecommunications filed Critical Nanjing University of Posts and Telecommunications
Priority to CN202311422980.XA priority Critical patent/CN117156494B/en
Publication of CN117156494A publication Critical patent/CN117156494A/en
Application granted granted Critical
Publication of CN117156494B publication Critical patent/CN117156494B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • H04W28/0925—Management thereof using policies
    • 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
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/06—Testing, supervising or monitoring using simulated traffic
    • 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
    • H04W28/0958—Management thereof based on metrics or performance parameters
    • H04W28/0967—Quality of Service [QoS] parameters

Landscapes

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

Abstract

The invention belongs to the field of mobile edge computing task scheduling and RIS auxiliary wireless communication, and discloses a three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication. The invention can effectively enhance the channel gain, improve the communication quality and solve the problem of server resource waste caused by unreasonable edge task scheduling, thereby reducing the average time delay of users.

Description

Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication
Technical Field
The invention belongs to the field of mobile edge computing task scheduling and RIS (radio resource identifier) auxiliary wireless communication, and particularly relates to a three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication.
Background
With the vigorous development of software and hardware such as artificial intelligence and intelligent terminal equipment, the high integration of the Internet of things and the artificial intelligence promotes the blueprint of the intelligent Internet of things in the 6G era. The concepts of intelligent industrial internet, intelligent traffic, intelligent power grid, intelligent city, etc. are gushing out, and various novel applications are gradually changing the life style of people. The application forms of the future 6G age are more enriched, and the huge quantification of network data can be a huge challenge for the current wireless infrastructure.
The mobile edge computing architecture provides computing and storage services for users by deploying the MEC servers at the network edge, and compared with the cloud computing architecture, the MEC servers are geographically closer to the users, so that delay and energy consumption of data transmission are greatly reduced. The contradiction between the calculation amount and the time delay sensitivity of the novel application is continuously aggravated along with the contradiction between the data amount and the user experience demand. An efficient edge computing task scheduling strategy plays a very important role in a scenario with strict requirements on time delay. In view of the characteristics of the limited performance, the isomerism, the user channel interference, the energy consumption limitation and the like of MEC computing resources, a new challenge is formed for computing migration optimization; in addition, wireless communication between the user and the edge base station is easily hindered by uncertainty of weather, obstacles and other environmental factors, so that the effect of the edge calculation migration technology is affected.
CN2022107236791 discloses a method and system for edge computing task scheduling in industrial internet, which provides a task scheduling policy algorithm (Improved Genetic Algorithm-based Task Scheduling Policy Algorithm, IGA-TSPA) based on improved genetic algorithm, and can effectively select an appropriate task scheduling position according to delay constraint and resource constraint. According to the method, the combined optimization model of task caching, block chain and task scheduling is established, the block chain technology is combined with three-layer architecture of industrial equipment, edge computing and cloud computing, task consumption is reduced, and meanwhile data security in the task scheduling process is improved.
CN202310553022X discloses a method for distributing transmission power of dynamic tasks based on rim-assisted NOMA edge calculation, which comprises the following steps: the RIS-NOMA edge calculation task transmission system is constructed, and the total cost expression of the UE end is established under the constraint of the queue length, but the invention does not consider the partitionability of the user task, takes the transmission power as an optimization target, only considers the quality optimization of wireless communication, adopts a DDQN algorithm, needs a large amount of calculation resource training models, and does not consider the problem of network parameter training caused by the limitation of edge resources and the real-time change of edge scenes.
CN2022102318494 discloses a method for dispatching edge computing tasks in industrial internet, which ignores the quality problem of wireless communication channels, and in a real scene, the wireless channels are extremely susceptible to environmental influence and change dynamically.
Disclosure of Invention
In order to solve the technical problems, the invention provides a three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication, which are used for solving the problem of uncertainty of wireless communication due to environmental factors such as weather, obstacles and the like by constructing a RIS auxiliary edge calculation wireless communication model and solving the problem of task scheduling delay optimization by using a method (AODSO) based on distributed game theory and alternative optimization.
In order to achieve the above purpose, the invention is realized by the following technical scheme:
the invention provides a three-terminal fusion task scheduling model of RIS auxiliary wireless communication, which mainly comprisesBase station->The individual MEC server, MEC-end, is denoted +.>And->Individual client, denoted->Meanwhile, a RIS end is deployed between the user end and the MEC end,
the RIS terminal comprises an RIS auxiliary wireless communication model, the RIS auxiliary wireless communication model is an RIS with N reflecting units, the RIS auxiliary wireless communication model comprises an RIS terminal data transmission module, an RIS parameter optimization module and a control module, an RIS controller is arranged in the control module, the data transmission module of the RIS terminal receives information sent by the MEC terminal and the user terminal, the channel condition of communication between the MEC server providing computing resources by the user terminal and the MEC terminal is obtained, the RIS parameter optimization module adaptively changes parameter information of the RIS according to the channel information, and the control module controls parameters of phase and phase shift of the RIS through the RIS controller;
the user side comprises an MEC task scheduling decision model, the MEC task scheduling decision model comprises a task dividing module, a user side data transmission module and a task scheduling decision module, the task dividing module divides a calculation task into a plurality of subtasks, the user side data transmission module obtains a scheduling decision of a user according to the obtained channel condition, the task scheduling decision module sends the obtained scheduling decision to the user side data transmission module, and the user side data transmission module broadcasts the scheduling decision to an RIS end and an MEC end;
the MEC terminal comprises an MEC terminal data transmission module, a competition decision module and a resource statistics module, and is used for solving the problem that multiple users compete for all server resources together.
The invention further improves that: the RIS-assisted wireless communication model is expressed as:
，
phase shift vectors for N reflection units of the RIS,represented as a scheduling decision of the user,expressed as the number of users,the weight of the user is represented as such,indicating the transmission rate at which the data is to be transmitted,representing beam forming vector, setting the reflection amplitude coefficient of each reflection unit to be the maximum value to maximize the signal reflection power, namely constraintConstraint(s)The power resulting from beamforming, denoted as base station, is less than the maximum transmit power of the base station。
The invention further improves that: at the RIS end, the RIS parameter optimization module adaptively changes the parameter information of the RIS according to the channel information, and specifically comprises the following steps:
step 1.1, after each user broadcasts own scheduling decision, the RIS terminal data transmission module receives scheduling decision information of the user and sends the scheduling decision information to the RIS parameter optimization module;
step 1.2, after the RIS parameter optimization module obtains the scheduling decision information, an AOPO method is used for optimizing RIS parameters;
step 1.3, after the RIS parameter optimization module obtains the optimal parameters, the parameters are sent to the control module;
step 1.4, the control module adjusts the phase and the phase shift of the reflective element of the RIS according to the obtained RIS parameter;
and step 1.5, the data transmission module transmits the current RIS parameters and the beamforming optimization parameters to the MEC terminal.
The invention further improves that: in the user side, the task scheduling method using the MEC task scheduling decision model comprises the following steps:
step 2.1, after a user issues a calculation task, the task division module divides the calculation task into a plurality of dependent subtasks and models the task as oneTo represent the user +.>The tasks of the user can be divided into +.>Subtasks, denoted->，/>Representing the dependency relationship between subtasks;
step 2.2, the user side data transmission module acquires channel information broadcast by an MEC server of an MEC side and use information of MEC server resources;
step 2.3, according to the communication resource and the computing resource information obtained by the data transmission module, the task scheduling decision module obtains a scheduling decision of a user by using a TSO method and sends the scheduling decision to the user side data transmission module;
step 2.4: the task scheduling decision module sends the obtained scheduling decision to the user side data transmission module;
step 2.5: the user data transmission module broadcasts the scheduling decision to the RIS end and the MEC end.
The invention discloses a three-terminal fusion task scheduling method for RIS auxiliary wireless communication, which comprises the following steps:
step 1, a MEC end data transmission module receives a scheduling decision sent by a user end and an RIS endScheduling decisionsThe resource statistics module is sent to the MEC end;
step 2: constructing a multi-user competition game model according to server resources and scheduling decisions:
，
wherein,representing scheduling decisions of users participating in a game, +.>Representing scheduling decisions of other users participating in the game;
step 3, the resource statistics module of the MEC terminal is based on an idle time slot insertion strategy and a scheduling decisionCalculating task completion time of each user +.>；
Step 4, according to the task completion time of each userCalculating the average time delay of the three-terminal fusion task scheduling model；
Step 5, adopting an alternative optimization and distributed game-based method (AODSO) to achieve the balanced states of three independent RIS ends, user ends and MEC ends so as to realize joint optimization of wireless communication quality and system average time delay;
step 6: finally, each user makes a scheduling decision according to the user's ownAnd scheduling the tasks.
The invention further improves that: in step 5, a method (AODSO) based on alternative optimization and distributed gaming specifically comprises the steps of:
step 5.1, after a user side generates a new task, dividing the task into a plurality of subtasks, generating a scheduling decision of the user by a task scheduling decision module by using a TSO method, and transmitting the scheduling decision to a RIS side;
step 5.2, after the RIS terminal receives the scheduling decision sent by the user, calculating by adopting an AOPO method to obtain the optimal RIS parameterAnd->Transmitting the data to the MEC end;
step 5.3, after the MEC receives the parameters from the RIS, according to the scheduling decision, the competition decision module calculates according to the idle time slot insertion strategy to obtain the task completion time of each user and the average time delay of the whole system;
step 5.4, selecting a scheduling decision for the device with the lowest average delay, i.e. the best average delayAs MEC resource competitionA success party;
step 5.5, according to the scheduling decisionRe-acquiring the channel condition and the residual condition of server resources;
step 5.6, the data transmission module of the MEC terminal makes the scheduling decision of the current selectionBroadcasting the current channel condition and the residual condition of the server resource to a user terminal;
and 5.7, repeating the steps 5.1-5.6 until all users do not change any more, and achieving an equilibrium state based on the method of the alternative optimization and the distributed game.
The invention further improves that: in step 5.3, the strategy based on the insertion of the idle time slots specifically comprises the following steps:
step 5.3.1, for each subtask, the MEC server of the computing resource provided by the MEC end queries own task queue;
step 5.3.2 MEC Server calculates the free time between subtasks already in the queued task queue；
Step 5.3.3 if the computation time of the subtaskInserting the subtask into the idle time slot for execution;
step 5.3.4 if the computation time of the subtaskGreater than all free slots, according toCalculating the end time of the subtask, wherein>Representing the earliest start time of the MEC server, < > in->Representing the time required for the computation of the subtask;
and step 5.3.5, repeating the steps 5.3.1-5.3.4 to directly calculate the task execution time of all MEC servers.
The invention further improves that: in the step 5.1, the TSO method specifically includes the following steps:
step 5.1.1, inputting all parameter information including calculation taskInformation, user and edge server information, < >>Representing user +.>Computing power of the device itself, +.>Representing MEC server->Is>Representing a userIs>Calculating the data quantity by subtasks;
step 5.1.2, according to the DAG graph of the user task, arranging the tasks into priority queuesEnsure the dependency relationship between each subtask, let +.>Representing the current subtask->Is (are) successor node set,)>Representing subtasksTo its successor node->Data transmission time,/-, of (2)>And->In the same mode, execute +.>，Representative subtask->The average computation time in MEC mode and local mode, the specific prioritization rules are defined as:
；
step 5.1.3 according toA queue for selecting the sub-task with highest priority and not being scheduled, and obtaining all MEC conditions which can be selected by the task;
step 5.1.4, for each optional MEC, calculating the transmission rates of the user and MEC based on the entered RIS parameters, calculating the userIs>Subtasks transfer to MEC server +.>Time required->The start time of the computing sub-task executable on the MEC is +.>：
；
Step 5.1.5, setRepresenting user +.>Computing power of the device itself, +.>Representing MEC server->Is used in the computing power of the (a) and (b),representing user +.>Is>Subtasks calculate the amount of data, calculate the time needed for the execution of the subtasks +.>：
，
Step 5.1.6, calculating the execution end time of the subtask：
；
Step 5.1.7, repeating steps 5.1.4 through 5.1.6 until all selectable feasible mode calculations are complete, selected such thatMinimum possible mode server +.>Is provided with->Subtask->The scheduling decision of (a) is server->；
Step 5.1.8, repeating the steps 5.1.3 to 5.1.7 until all the subtasks obtain the scheduling decisions, and finally obtaining the scheduling decisions of the whole task。
The invention further improves that: in step 5.2, optimizing the RIS parameters using the AOPO method specifically comprises the steps of:
the method specifically comprises the following steps:
step 5.2.1, scheduling decision according to userObtaining the situation of a user selecting a base station:
，
wherein:representing user +.>Is>Scheduling decisions for subtasks, ">Representing local execution, whereas tasks are performed in MEC server +.>Executing;
step 5.2.2 according toCalculating user +.>Channel gain to MEC terminal +.>Wherein->A phase shift matrix representing the RIS;
step 5.2.3 according to the formulaCalculating signal to noise ratioWherein->Representing a beamforming vector; />Representing background noise;
step 5.2.4, acquiring the transmission rate between the user and the base station；
Step 5.2.5 calculating the phase shift parameters of RIS using Riemann Conjugate Gradient (RCG) algorithmThe formula is:
，
wherein,representing step size->Representing a search direction;
step 5.2.6 phase Shift parameters according to the obtained RISComputing beamforming vector +.>The method specifically comprises the following steps:
，
wherein:indicating the number of users +.>，/>The representation represents user +>Channel gain of>Representation->Identity matrix of>An optimal dual variable representing a transmit power constraint;
step 5.2.7 based on beamforming vectorCalculating the weighted rate sum of the user>；
Step 5.2.8, repeating steps 5.2.5 to 5.2.7 until the WSR reaches a maximum;
step 5.2.9 outputting RIS phase shift parameterAnd a beamforming vector.
The beneficial effects of the invention are as follows:
aiming at the problems that instability and infrastructure of a wireless channel are easily influenced by natural environment and even are partially destroyed, the invention provides a method for introducing RIS auxiliary edge computing network communication, wherein the coverage rate of an edge network and the communication quality of a user are easily influenced by infrastructure faults, bad weather quality and the like.
The invention provides a task splitting method for splitting a user's computing task into a plurality of tasks constrained by a dependency relationship, and constructing a three-terminal (MEC terminal, user terminal, RIS terminal) fused task scheduling model according to the user task and channel conditions, and creatively provides a task scheduling and RIS parameter joint optimization method (AODSO) based on alternative optimization and distributed game by using joint optimization of wireless communication quality and system average time delay, wherein the AODSO method is divided into two sub-modules: the method comprises the steps that TSO and AOPO are adopted by a user side, task scheduling decisions of the user are obtained by the user side through the TSO method and are sent to an MEC side and an RIS side, wireless communication quality is optimized by the RIS side through the AOPO method, and finally joint optimization of the scheduling decisions and the communication quality is achieved by the AODSO method through an alternate optimization strategy. The method disclosed by the invention can effectively solve the problem of server resource waste caused by unreasonable edge task scheduling, thereby reducing the average time delay of users.
The invention is a process of alternate iteration, in one iteration, the user side obtains the dispatching decision locally, after the RIS end and the MEC end receive the dispatching decision, the RIS end and the MEC end obtain the optimal RIS parameter and the beam forming vector of the iteration through the method, and send the optimal RIS parameter and the beam forming vector back to the user side, and repeat the steps until the minimum average time delay of the whole system is reached.
Drawings
FIG. 1 is a schematic diagram of a terminal fusion task scheduling model of the present invention.
FIG. 2 is a schematic diagram of a three-terminal converged task scheduling method for RIS assisted wireless communication of the present invention.
FIG. 3 is a flow chart of a three-terminal converged task scheduling method of RIS assisted wireless communication of the present invention.
FIG. 4 is a graph comparing the convergence of DTSO-free RIS and DTSO-RIS using the AODSO method of the present invention.
Detailed Description
Embodiments of the invention are disclosed in the drawings, and for purposes of explanation, numerous practical details are set forth in the following description. However, it should be understood that these practical details are not to be taken as limiting the invention. That is, in some embodiments of the invention, these practical details are unnecessary.
1-2, considering the RIS auxiliary vehicle-mounted scene, the invention relates to a three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication, which specifically comprises the following steps:
step 1, when a vehicle enters a coverage area of a base station, a user issues a navigation calculation task, and a task division module divides the navigation task into a plurality of dependent subtasks according to a data stream. The specific division:
subtask 1, the user inputs a destination in the navigation application, and the navigation device activates the controller module to obtain the current position of the user.
Subtask 2, get all the alternative paths of the destination.
Subtask 3, obtaining the traffic conditions along the path, such as real-time traffic light conditions, traffic jam conditions and the like.
And 4, interacting with the navigation panel according to the returned result, and displaying the navigation scheme.
The task division module models the subtasks as oneTo represent the user +.>The tasks of the user can be divided into +.>Subtasks, denoted->；/>To represent dependencies between subtasks.
And 2, the user side data transmission module acquires the channel condition broadcasted by the MEC server and the use condition of MEC server resources.
And step 3, according to the conditions of communication resources and computing resources obtained by the user side data transmission module, a task scheduling decision module of the vehicle controller obtains a scheduling decision of a user by using a TSO method and sends the scheduling decision to the user side data transmission module.
The TSO method comprises the following specific steps:
step 3.1: inputting all parameter information, including computing tasksInformation, user and edge server information, < >>Representing user +.>Computing power of the device itself, +.>Representing MEC server->Is>Representing user +.>Is>The subtasks calculate the amount of data.
Step 3.2: ranking tasks into priority queues according to DAG graph of user tasksThereby ensuring the dependency relationship between each subtask, let +.>Representing the current subtask->Is (are) successor node set,)>Representing subtasks->To its successor node->Data transmission time,/-, of (2)>And->In the same mode, execute +.>。Representative subtask->Average computation time in MEC mode and local mode. The specific prioritization rules are defined as:
，
step 3.3: and selecting the sub-tasks with highest priority which are not scheduled according to the priority queue, and acquiring all MEC conditions which can be selected by the tasks.
Step 3.4: for each selectable MEC, calculating transmission rates of users and MECs according to the input RIS parameters, and calculating usersIs>Subtasks transfer to MEC server +.>Time required->Calculating a start time at which a subtask can be executed on the MEC +.>：
，
Step 3.5: is provided withRepresenting user +.>Computing power of the device itself, +.>Representing MEC server->Is used in the computing power of the (a) and (b),representing user +.>Is>Subtasks calculate the amount of data, calculate the time needed for the execution of the subtasks +.>：
，
Step 3.6: calculating execution end time of subtasks：
，
Step 3.7: repeating steps 3.4-3.6 until all possible modes that can be selected are calculated. Selected such thatMinimum possible mode server +.>Is provided with->. Subtask->The scheduling decision of (a) is server->。
Step 3.8: repeating steps 3.3-3.7 until all sub-tasks have obtained their scheduling decisions. Finally, the scheduling decision of the whole task is obtained。
Step 4: and the task scheduling decision module sends the obtained scheduling decision to the data transmission module.
Step 5: the data transmission module broadcasts the scheduling decisions to the RIS end and the MEC end.
The RIS terminal mainly comprises a RIS terminal data transmission module, a RIS parameter optimization module and a control module. The RIS end data transfer module can receive information sent by the MEC end and the user end. Channel conditions of communication between the user and the MEC server are obtained. The parameter adjustment module may adaptively change parameter information of the RIS according to the channel information. The control module is used for controlling parameters such as phase, phase shift and the like of the RIS through the RIS controller. The method comprises the following specific steps:
and 6, after each vehicle user broadcasts the own scheduling decision, the RIS end data transmission module receives the scheduling decision information of the user. And sends the information to the RIS parameter optimization module.
And 7, after the RIS parameter optimization module acquires the decision information, an AOPO method is used for optimizing the RIS parameter, and the AOPO method comprises the following specific steps:
step 7.1 according toObtaining the condition of selecting base station by user, < >>。/>Representing user +.>Is>Scheduling decisions for subtasks, ">Representing local execution, whereas tasks are performed at the MEC serverAnd executing.
Step 7.2 according toCalculating user +.>Channel gain to MEC terminal +.>Wherein->Representing the phase shift matrix of the RIS.
Step 7.3: according to the formulaCalculating signal to noise ratioWherein->Representing the beamforming vector.
Step 7.4: acquiring transmission rate between user and base station。
Step 7.5: calculation of RIS phase shift parameters using Riemann Conjugate Gradient (RCG) algorithm。
Step 7.6: according to the obtainedComputing beamforming vector +.>。
Step 7.7: from beam forming vectorsCalculating the weighted rate sum of the user>。
Step 7.8: repeating the steps 2.5-2.7 until the WSR reaches the maximum value.
Step 7.9: outputting RIS phase shift parametersAnd beamforming vector>。
Step 8: and after the RIS parameter optimization module obtains the optimal parameters, the parameters are sent to the control module.
Step 9: the control module adjusts the phase and phase shift of the reflective element of the RIS according to the obtained RIS parameter value.
Step 10: the RIS end data transmission module transmits the current RIS parameters and the beamforming optimization parameters to the MEC end.
The MEC terminal mainly comprises an MEC terminal data transmission module, a competition decision module and a resource statistics module.
Step 11: the MEC end data transmission module receives the scheduling decision sent by the vehicle user and RIS through a Road Side Unit (RSU)Scheduling decision +.>And sending the result to the competition decision module.
Step 12: the contention decision module is based on an idle slot insertion strategy and a scheduling decisionCalculating task completion time of each user +.>。
Step 13: according to each ofCalculate the average delay of the whole system +.>。
Step 14: scheduling decision for selecting a device with a low average latency, i.e. the best average latencyAnd is according to scheduling decision->The channel condition and the server resource remaining condition are retrieved.
Step 15: the data transmission module makes the scheduling decision of the current selectionThe current channel condition and the residual condition of the server resource are broadcasted to the user terminal.
Step 16: finally, an alternative optimization and distributed game-based method (AODSO) is adopted to achieve the balanced state of three independent RIS ends, MEC ends and user ends, so that the wireless communication quality and the average time delay of the system are jointly optimized. The method is characterized in that the method is an alternate iteration process, in one iteration, a user side obtains a scheduling decision locally, after the RIS side and the MEC side receive the scheduling decision, the RIS side and the MEC side obtain the optimal RIS parameter and the beam forming vector of the iteration through the method, the optimal RIS parameter and the beam forming vector are sent back to the user side, and the steps are repeated until the minimum average time delay of the whole system is reached.
In order to verify the effectiveness of the proposed RIS auxiliary MEC communication model and AODSO method, a comparison experiment is carried out:
1) DTSO-no RIS: a TSO method for distributed game is used to solve the task scheduling optimization model without RIS auxiliary MEC.
2) DTSO-RIS: a distributed game TSO method is used for solving a RIS auxiliary MEC task scheduling optimization model, but does not contain an alternate optimization strategy AOPO of RIS parameters.
3) AODSO: the invention discloses a method based on alternative optimization and distributed game.
As shown in fig. 4, the method of the present invention can stably reach a convergence state. Firstly, the invention verifies the effect of RIS module introduction, and compared with DTSO-without RIS, the RIS auxiliary MEC communication model provided by the invention reduces the average time delay of the used users by 4.8%.
Furthermore, the invention verifies the function of the RIS parameter optimization module, and compared with the DTSO-free RIS and the DTSO-RIS, the method adopts the alternative optimization fusion three-terminal AODSO method to respectively reduce the average time delay of all users by 32ms (18.1%) and 24ms (13.8%).
According to the invention, by constructing the RIS auxiliary edge calculation wireless communication model, the problem that wireless communication is easily subjected to uncertainty of multiple environmental factors such as weather, obstacles and the like is solved, and the task scheduling delay optimization is solved by using a method (AODSO) based on distributed game theory and alternative optimization.
The foregoing description is only illustrative of the invention and is not to be construed as limiting the invention. Various modifications and variations of the present invention will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, or the like, which is within the spirit and principles of the present invention, should be included in the scope of the claims of the present invention.

Claims (9)

1. A three-terminal fusion task scheduling model of RIS auxiliary wireless communication is characterized in that:
the three-terminal fusion task scheduling model mainly comprisesBase station->The individual MEC servers, or MEC ends, are denoted asAnd->Individual client, denoted->Meanwhile, a RIS end is deployed between the user end and the MEC end,
the RIS terminal comprises an RIS auxiliary wireless communication model, the RIS auxiliary wireless communication model is an RIS with N reflecting units, the RIS auxiliary wireless communication model comprises an RIS terminal data transmission module, an RIS parameter optimization module and a control module, an RIS controller is arranged in the control module, the data transmission module of the RIS terminal receives information sent by the MEC terminal and the user terminal, the channel condition of communication between the MEC server providing computing resources by the user terminal and the MEC terminal is obtained, the RIS parameter optimization module adaptively changes parameter information of the RIS according to the channel information, and the control module controls phase and phase shift parameters of the RIS through the RIS controller;
the user side comprises an MEC task scheduling decision model, the MEC task scheduling decision model comprises a task dividing module, a user side data transmission module and a task scheduling decision module, the task dividing module divides a calculation task into a plurality of subtasks, the user side data transmission module obtains a scheduling decision of a user according to the obtained channel condition, the task scheduling decision module sends the obtained scheduling decision to the user side data transmission module, and the user side data transmission module broadcasts the scheduling decision to an RIS end and an MEC end;
the MEC terminal comprises an MEC terminal data transmission module, a competition decision module and a resource statistics module, and is used for solving the problem that multiple users compete for all server resources together.
2. The RIS assisted wireless communication three terminal fusion task scheduling model of claim 1, wherein: the RIS-assisted wireless communication model is expressed as:
，
phase shift vectors for N reflection units of the RIS, and (2)>Scheduling decision expressed as user->Expressed as the number of users,representing the weight of the user->Representing transmission rate +.>Representing beam forming vector, setting the maximum value of the reflection amplitude coefficient of each reflection unit to maximize the signal reflection power, namely the constraint +.>Constraint->The power resulting from the beamforming, denoted base station, is smaller than the maximum transmit power of the base station +.>。
3. The RIS assisted wireless communication three terminal fusion task scheduling model of claim 1, wherein: at the RIS end, the RIS parameter optimization module adaptively changes the parameter information of the RIS according to the channel information, and specifically comprises the following steps:
step 1.1, after each user broadcasts own scheduling decision, the RIS terminal data transmission module receives scheduling decision information of the user and sends the scheduling decision information to the RIS parameter optimization module;
step 1.2, after the RIS parameter optimization module obtains the scheduling decision information, an AOPO method is used for optimizing RIS parameters;
step 1.3, after the RIS parameter optimization module obtains the optimal parameters, the parameters are sent to the control module;
step 1.4, the control module adjusts the phase and the phase shift of the reflective element of the RIS according to the obtained RIS parameter;
and step 1.5, the data transmission module transmits the current RIS parameters and the beamforming optimization parameters to the MEC terminal.
4. The RIS assisted wireless communication three terminal fusion task scheduling model of claim 1, wherein: in the user side, the task scheduling method using the MEC task scheduling decision model comprises the following steps:
step 2.1, after a user issues a calculation task, the task division module divides the calculation task into a plurality of dependent subtasks and models the task as oneTo represent the user +.>The tasks of the user can be divided into +.>Subtasks, denoted->，/>Representing the dependency relationship between subtasks;
step 2.2, the user side data transmission module acquires channel information broadcast by an MEC server of an MEC side and use information of MEC server resources;
step 2.3, according to the communication resource and the computing resource information obtained by the data transmission module, the task scheduling decision module obtains a scheduling decision of a user by using a TSO method and sends the scheduling decision to the user side data transmission module;
step 2.4: the task scheduling decision module sends the obtained scheduling decision to the user side data transmission module;
step 2.5: the user data transmission module broadcasts the scheduling decision to the RIS end and the MEC end.
5. A three-terminal fusion task scheduling method for RIS auxiliary wireless communication is characterized in that: the three-terminal fusion task scheduling method comprises the following steps:
step 1, MEC end data transmission moduleScheduling decision sent by user side and RIS side is receivedScheduling decision +.>The resource statistics module is sent to the MEC end;
step 2: constructing a multi-user competition game model according to server resources and scheduling decisions:
，
wherein,representing scheduling decisions of users participating in a game, +.>Representing scheduling decisions of other users participating in the game;
step 3, the resource statistics module of the MEC terminal is based on an idle time slot insertion strategy and a scheduling decisionCalculating task completion time of each user +.>；
Step 4, according to the task completion time of each userCalculating the average time delay of the three-terminal fusion task scheduling model；
Step 5, adopting an alternative optimization and distributed game-based method (AODSO) to achieve the balanced states of three independent RIS ends, user ends and MEC ends so as to realize joint optimization of wireless communication quality and system average time delay;
step 6: finally, each user makes a scheduling decision according to the user's ownAnd scheduling the tasks.
6. The three-terminal fusion task scheduling method for RIS-assisted wireless communication according to claim 5, wherein: in said step 5, a method (AODSO) based on alternative optimization and distributed gaming comprises in particular the following steps:
step 5.1, after a user side generates a new task, dividing the task into a plurality of subtasks, generating a scheduling decision of the user by a task scheduling decision module by using a TSO method, and transmitting the scheduling decision to a RIS side;
step 5.2, after the RIS terminal receives the scheduling decision sent by the user, calculating by adopting an AOPO method to obtain the optimal RIS parameterAnd->Transmitting the data to the MEC end;
step 5.3, after the MEC receives the parameters from the RIS, according to the scheduling decision, the competition decision module calculates according to the idle time slot insertion strategy to obtain the task completion time of each user and the average time delay of the whole system;
step 5.4, scheduling decision for selecting devices with low average latencyAs a successful party of MEC resource competition;
step 5.5, according to the scheduling decisionRe-acquisition of channel conditionsThe server resource remaining condition;
step 5.6, the data transmission module of the MEC terminal makes the scheduling decision of the current selectionBroadcasting the current channel condition and the residual condition of the server resource to a user terminal;
and 5.7, repeating the steps 5.1-5.6 until all users do not change any more, and achieving an equilibrium state based on the method of the alternative optimization and the distributed game.
7. The three-terminal fusion task scheduling method for RIS-assisted wireless communication according to claim 6, wherein: in the step 5.3, the strategy based on the insertion of the idle time slot specifically includes the following steps:
step 5.3.1, for each subtask, the MEC server of the computing resource provided by the MEC end queries own task queue;
step 5.3.2 MEC Server calculates the free time between subtasks already in the queued task queue；
Step 5.3.3 if the computation time of the subtaskInserting the subtask into the idle time slot for execution;
step 5.3.4 if the computation time of the subtaskGreater than all free slots, then according to +.>Calculating the end time of the subtask, wherein>Representing the earliest start time of the MEC server, < > in->Representing the time required for the computation of the subtask;
and step 5.3.5, repeating the steps 5.3.1-5.3.4 to directly calculate the task execution time of all MEC servers.
8. The three-terminal fusion task scheduling method for RIS-assisted wireless communication according to claim 6, wherein: in the step 5.1, the TSO method specifically includes the following steps:
step 5.1.1, inputting all parameter information including calculation taskInformation, user and edge server information, < >>Representing user +.>Computing power of the device itself, +.>Representing MEC server->Is>Representing user +.>Is the first of (2)Calculating the data quantity by subtasks;
step 5.1.2, arranging the tasks into optimal according to the DAG graph of the user taskFirst order queueEnsure the dependency relationship between each subtask, let +.>Representing the current subtask->Is (are) successor node set,)>Representing subtasks->To its successor node->Data transmission time,/-, of (2)>And->In the same mode, execute +.>，Representative subtask->The average computation time in MEC mode and local mode, the specific prioritization rules are defined as:
；
step 5.1.3 according toA queue for selecting the sub-task with highest priority and not being scheduled, and obtaining all MEC conditions which can be selected by the task;
step 5.1.4, for each optional MEC, calculating the transmission rates of the user and MEC based on the entered RIS parameters, calculating the userIs>Subtasks transfer to MEC server +.>Time required->The start time of the computing sub-task executable on the MEC is +.>：
；
Step 5.1.5, setRepresenting user +.>Computing power of the device itself, +.>Representing MEC server->Is used in the computing power of the (a) and (b),representing user +.>Is>Subtasks calculate the amount of data, calculate the time needed for the execution of the subtasks +.>：
，
Step 5.1.6, calculating the execution end time of the subtask：
；
Step 5.1.7, repeating steps 5.1.4 through 5.1.6 until all selectable feasible mode calculations are complete, selected such thatMinimum possible mode server +.>Is provided with->Subtask->The scheduling decision of (a) is server->；
Step 5.1.8, repeating the steps 5.1.3 to 5.1.7 until all the subtasks obtain the scheduling decisions, and finally obtaining the scheduling decisions of the whole task。
9. The three-terminal fusion task scheduling method for RIS-assisted wireless communication according to claim 6, wherein: in said step 5.2, the optimization of the RIS parameters using the AOPO method comprises in particular the following steps:
step 5.2.1, scheduling decision according to userObtaining the situation of a user selecting a base station:
，
wherein:representing user +.>Is>Scheduling decisions for subtasks, ">Representing local execution, whereas tasks are performed in MEC server +.>Executing;
step 5.2.2 according toCalculating the RIS-assisted wireless communication moduleUnder the->Channel gain to MEC terminal +.>Wherein->A phase shift matrix representing the RIS;
step 5.2.3 according to the formulaCalculating signal to noise ratioWherein->Representing a beamforming vector; />Representing background noise;
step 5.2.4, acquiring the transmission rate between the user and the base station；
Step 5.2.5 calculating the phase shift parameters of RIS using Riemann Conjugate Gradient (RCG) algorithmThe formula is:
，
wherein,representing step size->Representing a search direction;
step 5.2.6 phase Shift parameters according to the obtained RISComputing beamforming vector +.>The method specifically comprises the following steps:
，
wherein:indicating the number of users +.>，/>The representation represents user +>Channel gain of>Representation->Identity matrix of>An optimal dual variable representing a transmit power constraint;
step 5.2.7 based on beamforming vectorCalculating a weighted speed of a userRate sum->；
Step 5.2.8, repeating steps 5.2.5 to 5.2.7 until the WSR reaches a maximum;
step 5.2.9 outputting RIS phase shift parameterAnd beamforming vector>。
CN202311422980.XA 2023-10-31 2023-10-31 Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication Active CN117156494B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311422980.XA CN117156494B (en) 2023-10-31 2023-10-31 Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311422980.XA CN117156494B (en) 2023-10-31 2023-10-31 Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication

Publications (2)

Publication Number Publication Date
CN117156494A true CN117156494A (en) 2023-12-01
CN117156494B CN117156494B (en) 2024-01-19

Family

ID=88901166

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311422980.XA Active CN117156494B (en) 2023-10-31 2023-10-31 Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication

Country Status (1)

Country Link
CN (1) CN117156494B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111182495A (en) * 2019-12-25 2020-05-19 大连理工大学 5G internet of vehicles partial calculation unloading method
CN113746578A (en) * 2021-08-18 2021-12-03 南京邮电大学 Communication system transmission method based on assistance of intelligent reflection surface
CN115567983A (en) * 2022-09-16 2023-01-03 南京邮电大学 Stackelberg game-based D2D calculation unloading method
CN115696451A (en) * 2022-10-10 2023-02-03 广州大学 Optimization method for energy and task scheduling of edge computing system
US20230047993A1 (en) * 2020-07-02 2023-02-16 Zte Corporation Surface element segmentation and node grouping for intelligent reflecting devices
US20230097583A1 (en) * 2021-09-28 2023-03-30 Samsung Electronics Co., Ltd. Apparatus and method for data communication based on intelligent reflecting surface in wireless communication system
CN115988536A (en) * 2023-03-20 2023-04-18 南京邮电大学 Dual-reconfigurable intelligent surface assisted mobile edge computing system optimization method
US20230129288A1 (en) * 2021-10-21 2023-04-27 Samsung Electronics Co., Ltd. Method and device for ris automatic setup
CN116208619A (en) * 2023-03-21 2023-06-02 西安电子科技大学 Intelligent reflection surface-assisted Internet of vehicles safety calculation unloading method, system, equipment and medium
CN116390122A (en) * 2023-03-21 2023-07-04 重庆邮电大学 Multi-RIS auxiliary downlink MIMO-NOMA system optimization method based on sum rate maximization
CN116916386A (en) * 2023-08-02 2023-10-20 南京邮电大学 Large model auxiliary edge task unloading method considering user competition and load

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111182495A (en) * 2019-12-25 2020-05-19 大连理工大学 5G internet of vehicles partial calculation unloading method
US20230047993A1 (en) * 2020-07-02 2023-02-16 Zte Corporation Surface element segmentation and node grouping for intelligent reflecting devices
CN113746578A (en) * 2021-08-18 2021-12-03 南京邮电大学 Communication system transmission method based on assistance of intelligent reflection surface
US20230097583A1 (en) * 2021-09-28 2023-03-30 Samsung Electronics Co., Ltd. Apparatus and method for data communication based on intelligent reflecting surface in wireless communication system
US20230129288A1 (en) * 2021-10-21 2023-04-27 Samsung Electronics Co., Ltd. Method and device for ris automatic setup
CN115567983A (en) * 2022-09-16 2023-01-03 南京邮电大学 Stackelberg game-based D2D calculation unloading method
CN115696451A (en) * 2022-10-10 2023-02-03 广州大学 Optimization method for energy and task scheduling of edge computing system
CN115988536A (en) * 2023-03-20 2023-04-18 南京邮电大学 Dual-reconfigurable intelligent surface assisted mobile edge computing system optimization method
CN116208619A (en) * 2023-03-21 2023-06-02 西安电子科技大学 Intelligent reflection surface-assisted Internet of vehicles safety calculation unloading method, system, equipment and medium
CN116390122A (en) * 2023-03-21 2023-07-04 重庆邮电大学 Multi-RIS auxiliary downlink MIMO-NOMA system optimization method based on sum rate maximization
CN116916386A (en) * 2023-08-02 2023-10-20 南京邮电大学 Large model auxiliary edge task unloading method considering user competition and load

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
BIN XU: "Fine-grained Task Scheduling Based on Priority for Heterogeneous Mobile Edge Computing", 《IEEE》, pages 1 - 6 *
李斌: "智能超表面赋能移动边缘计算部分任务卸载策略", 《电子与信息学报》 *

Also Published As

Publication number Publication date
CN117156494B (en) 2024-01-19

Similar Documents

Publication Publication Date Title
Yang et al. Multi-UAV-enabled load-balance mobile-edge computing for IoT networks
Luo et al. Collaborative data scheduling for vehicular edge computing via deep reinforcement learning
CN113296845B (en) Multi-cell task unloading algorithm based on deep reinforcement learning in edge computing environment
CN114499629B (en) Dynamic allocation method for jumping beam satellite system resources based on deep reinforcement learning
CN110312231A (en) Content caching decision and resource allocation joint optimization method based on mobile edge calculations in a kind of car networking
CN113660681B (en) Multi-agent resource optimization method applied to unmanned aerial vehicle cluster auxiliary transmission
WO2023245835A1 (en) Satellite transmission optimization method based on grid delay prediction
Qi et al. Energy-efficient resource allocation for UAV-assisted vehicular networks with spectrum sharing
CN114900225B (en) Civil aviation Internet service management and access resource allocation method based on low-orbit giant star base
CN109819422B (en) Stackelberg game-based heterogeneous Internet of vehicles multi-mode communication method
CN116489226A (en) Online resource scheduling method for guaranteeing service quality
Gao et al. Multi-armed bandits scheme for tasks offloading in MEC-enabled maritime communication networks
Luo et al. Joint game theory and greedy optimization scheme of computation offloading for UAV-aided network
Wang et al. Multi-access edge computing based vehicular network: Joint task scheduling and resource allocation strategy
Wang et al. Joint optimization of resource allocation and computation offloading based on game coalition in C-V2X
CN114051252A (en) Multi-user intelligent transmitting power control method in wireless access network
CN117156494B (en) Three-terminal fusion task scheduling model and method for RIS auxiliary wireless communication
CN117412391A (en) Enhanced dual-depth Q network-based Internet of vehicles wireless resource allocation method
CN115499875B (en) Satellite internet task unloading method, system and readable storage medium
CN115765826B (en) Unmanned aerial vehicle network topology reconstruction method for on-demand service
CN115514769B (en) Satellite elastic Internet resource scheduling method, system, computer equipment and medium
CN116634450A (en) Dynamic air-ground heterogeneous network user association enhancement method based on reinforcement learning
Song et al. Joint computing, caching and communication resource allocation in the satellite-terrestrial integrated network with UE cooperation
CN114928826A (en) Two-stage optimization method, controller and decision method for software-defined vehicle-mounted task unloading and resource allocation
Zhang et al. Cybertwin-driven multi-intelligent reflecting surfaces aided vehicular edge computing leveraged by deep reinforcement learning

Legal Events

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