CN112995913B - Unmanned aerial vehicle track, user association and resource allocation joint optimization method - Google Patents

Unmanned aerial vehicle track, user association and resource allocation joint optimization method Download PDF

Info

Publication number
CN112995913B
CN112995913B CN202110252974.9A CN202110252974A CN112995913B CN 112995913 B CN112995913 B CN 112995913B CN 202110252974 A CN202110252974 A CN 202110252974A CN 112995913 B CN112995913 B CN 112995913B
Authority
CN
China
Prior art keywords
unmanned aerial
aerial vehicle
mec
drone
user
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.)
Active
Application number
CN202110252974.9A
Other languages
Chinese (zh)
Other versions
CN112995913A (en
Inventor
董超
游文静
经宇骞
刘青昕
屈毓锛
陶婷
吴启晖
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 Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
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 Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN202110252974.9A priority Critical patent/CN112995913B/en
Publication of CN112995913A publication Critical patent/CN112995913A/en
Application granted granted Critical
Publication of CN112995913B publication Critical patent/CN112995913B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/029—Location-based management or tracking services
    • 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
    • H04W28/0215—Traffic management, e.g. flow control or congestion control based on user or device properties, e.g. MTC-capable devices
    • H04W28/0221—Traffic management, e.g. flow control or congestion control based on user or device properties, e.g. MTC-capable devices power availability or consumption
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W84/00—Network topologies
    • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Abstract

The invention discloses an unmanned aerial vehicle track, user association and resource allocation joint optimization method, which considers an air-air communication network, and provides a mobile edge calculation supporting patrol system consisting of a plurality of user unmanned aerial vehicles and a plurality of MEC unmanned aerial vehicles, wherein each user unmanned aerial vehicle is provided with a sensor for data acquisition, the MEC unmanned aerial vehicle is provided with an edge calculation server for assisting the user unmanned aerial vehicle to perform calculation tasks, the user unmanned aerial vehicle moves in a specified area to cover task points, the calculation tasks can be unloaded to the edge calculation unmanned aerial vehicle for calculation and can also be unloaded to a cluster head unmanned aerial vehicle for calculation, and in order to reduce the energy consumption of the user unmanned aerial vehicle, the user association, the calculation resource allocation and the track of the MEC unmanned aerial vehicle between the cluster head unmanned aerial vehicle and the MEC unmanned aerial vehicle are jointly optimized. The invention can minimize the total energy consumption of the unmanned aerial vehicle of the user under the constraint of ensuring the computing resources and the task completion time.

Description

Unmanned aerial vehicle track, user association and resource allocation joint optimization method
Technical Field
The invention relates to the technical field of unmanned aerial vehicle ad hoc network mobile edge computing, in particular to a mobile edge computing unmanned aerial vehicle track, user association and computing resource allocation joint optimization method under a clustering architecture.
Background
In applications of the area coverage problem, performing computationally intensive tasks is one of the challenges facing drones. In recent years, with the development of communication and artificial intelligence, computing-intensive applications are leading to a revolution in mobile applications, which can significantly improve the quality of experience of mobile users. However, this presents a significant challenge for mobile devices that are computationally weak and have limited battery capacity. Moving edge computing is considered one of the promising techniques to address this challenge, and is receiving increasing research attention from both the industry and academia. Since the mobile edge computing server is typically deployed in a location close to the end user, the computing performance of the mobile user is significantly improved, and it is cost effective and can save energy consumption. Unmanned aerial vehicles possess a number of advantages, and through integrating unmanned aerial vehicles into a mobile edge computing network, unmanned aerial vehicle-assisted mobile edge computing architecture is proposed. In these architectures, a drone may be considered a user that has a computing task to perform, or a relay that assists the user in offloading the computing task, or a mobile edge computing server that performs the computing task. Compared with the traditional ground mobile edge computing network, the unmanned aerial vehicle-assisted mobile edge computing network has several outstanding advantages, can be flexibly deployed under most conditions, and even can be used in places where the ground edge mobile computing network cannot be conveniently and reliably established in wilderness, deserts and complex terrains. Furthermore, since there is a high probability that a short-range line-of-sight link exists for offloading the computing task and downloading the computing results, computing performance can be improved. Meanwhile, the track of the unmanned aerial vehicle can be optimized, and the user computing performance is further improved. When the cost of drones drops low enough, drone-assisted mobile edge computing networks will be widely deployed. The existing research mostly considers the performance factors such as system energy consumption, resource allocation, calculation rate and time delay.
Disclosure of Invention
The invention provides a joint optimization method for unmanned aerial vehicle track, user association and resource allocation, which aims at the defects in the prior art, and takes an air-air communication network into consideration, wherein the inspection system is composed of a plurality of user unmanned aerial vehicles and a plurality of MEC unmanned aerial vehicles and supports mobile edge calculation, each user unmanned aerial vehicle is provided with a sensor for data acquisition, the MEC unmanned aerial vehicles are provided with an edge calculation server for assisting the user unmanned aerial vehicles to carry out calculation tasks, the user unmanned aerial vehicles move in a specified area to cover task points, the calculation tasks can be unloaded to the edge calculation unmanned aerial vehicles for calculation and can also be unloaded to the cluster head unmanned aerial vehicles for calculation, and in order to reduce the energy consumption of the user unmanned aerial vehicles, the joint optimization is carried out on the user association of the cluster head unmanned aerial vehicles and the MEC unmanned aerial vehicles, the calculation resource allocation and the track of the MEC unmanned aerial vehicles. The invention can minimize the total energy consumption of the unmanned aerial vehicle of the user under the constraint of ensuring the computing resources and the task completion time.
In order to achieve the purpose, the invention adopts the following technical scheme:
a joint optimization method for unmanned aerial vehicle track, user association and resource allocation is provided, wherein a network comprises a user unmanned aerial vehicle set
Figure GDA0003506035700000014
And MEC unmanned plane set
Figure GDA0003506035700000013
A task set is represented by S ═ {1, 2.., S }; the user unmanned aerial vehicle cluster is previously divided into clusters and cluster head clusters
Figure GDA0003506035700000011
Set of cluster members
Figure GDA0003506035700000012
The joint optimization method comprises the following steps:
s1, modeling the task points in the cluster head unmanned aerial vehicle collaborative access area as a multi-traveler problem, and planning the cluster head track by adopting a genetic algorithm;
s2, constructing a network model, a calculation task scheduling model and an energy consumption model based on the planning result of the step S1;
s3, according to the constructed network model, the calculation task scheduling model and the energy consumption model, the problem of minimizing the energy consumption of the user unmanned aerial vehicle under the constraint of calculation resources and task completion time is solved;
s4, decomposing the problem of minimizing the energy consumption of the user unmanned aerial vehicle into two sub-problems, wherein the first sub-problem is a user association and resource allocation optimization problem, and the second sub-problem is an MEC unmanned aerial vehicle track optimization problem;
s5, converting the second sub-problem from the non-convex problem into a convex optimization problem by introducing a continuous relaxation variable and adopting a continuous convex approximation method;
and S6, solving the first subproblem by adopting a Mosek optimization solver, solving the second subproblem by adopting a CVX optimization solver, and iteratively solving the two subproblems based on a block coordinate descent method.
In order to optimize the technical scheme, the specific measures adopted further comprise:
further, in step S2, the process of constructing the network model, the computation task scheduling model, and the energy consumption model includes the following steps:
s21, in the network model, considering an unmanned aerial vehicle auxiliary moving edge calculation patrol system consisting of N rotor unmanned aerial vehicles and M fixed wing unmanned aerial vehicles, wherein S ground task points are arranged in the area;
s22, during a designated task, dividing a task cycle into T slots, wherein the slot length is equally divided into τ, and the slot set is denoted as T ═ 1, 2.., T }; the position of cluster drone j is represented as
Figure GDA0003506035700000021
Where j ∈ H, the position of cluster member drone k is represented as
Figure GDA0003506035700000022
Where k ∈ CM. The flight direction of the MEC unmanned plane in each time slot is
Figure GDA0003506035700000023
Distance of flight
Figure GDA0003506035700000024
Wherein the content of the first and second substances,
Figure GDA0003506035700000025
for the maximum flight distance, v, of the ith MEC unmanned aerial vehicle in each time slotiThe flight speed of the ith MEC unmanned aerial vehicle is taken as the flight speed of the ith MEC unmanned aerial vehicle; in the calculation task scheduling model, each timeThe unmanned aerial vehicle l of the frame user generates a task at the t time slot
Figure GDA0003506035700000026
Figure GDA0003506035700000027
Wherein l ∈ (H ≦ CM),
Figure GDA0003506035700000028
indicating completion
Figure GDA0003506035700000029
The total number of CPU cycles required for the CPU,
Figure GDA00035060357000000210
representing the data amount required to be calculated by the user unmanned plane l;
Figure GDA00035060357000000211
a value of 1 indicates that in the t-th time slot, the task on the cluster head drone j is unloaded to be executed on the MEC drone i,
Figure GDA00035060357000000212
a value of 0 indicates that a task is executed on the cluster head unmanned aerial vehicle j in the t-th time slot;
s23, in the energy consumption model, the total task energy consumption in the t-th time slot is represented as:
Figure GDA00035060357000000213
wherein the content of the first and second substances,
Figure GDA0003506035700000031
for the transmission of cluster heads to the MEC server,
Figure GDA0003506035700000032
the energy consumption for the transmission of cluster members to the cluster head,
Figure GDA0003506035700000033
calculating consumed energy consumption, k, locally for cluster headsjMore than or equal to 0 is effective switch capacitance, vjIs a constant; euclidean distance from cluster member to cluster head
Figure GDA0003506035700000034
The Euclidean distance from a cluster member k to a cluster head unmanned aerial vehicle j;
Figure GDA0003506035700000035
representing the Euclidean distance from the cluster head unmanned aerial vehicle j to the MEC unmanned aerial vehicle i; b is a channel bandwidth and is a channel bandwidth,
Figure GDA0003506035700000036
the node transmit power for cluster head drone j,
Figure GDA0003506035700000037
for node transmit power of drone k, α ═ g0G0/σ2，G0≈2.2846，g0Channel gain, σ, per unit distance2Is the noise power;
Figure GDA0003506035700000038
is the local computing resource provided by the cluster head unmanned aerial vehicle j at the moment t, vjIs a local calculation of the energy consumption parameter, vjIs a local calculation of the energy consumption parameter, HiIs the flying height of MEC unmanned plane i, hjIs the flight height of the cluster head drone j.
Further, in step S3, the process of proposing the problem of minimizing energy consumption of the user drone under the constraints of computing resources and task completion time includes the following steps:
s31, defining the MEC unmanned aerial vehicle track variable as
Figure GDA0003506035700000039
The correlation variable of the MEC unmanned aerial vehicle and the cluster head unmanned aerial vehicle is defined as
Figure GDA00035060357000000310
A computing resource allocation variable is defined as
Figure GDA00035060357000000311
S32, modeling the problem of minimizing the energy consumption of the user unmanned aerial vehicle cluster as:
Figure GDA00035060357000000312
where U, A, F is an optimization variable, problem (1-2) C1 indicates that cluster-head drones have and only offload tasks to a certain edge computing drone or perform computing tasks on themselves, and problem (1-2) C2 indicates that each MEC drone i can only serve at most simultaneously
Figure GDA00035060357000000313
A cluster head drone is erected, a problem (1-2) C3 shows that the sum of the task unloading capacity of a cluster head drone j cannot exceed the amount of computing resources that an MEC drone i can provide, a problem (1-2) C4 shows that a cluster head drone to be unloaded is to be within the communication coverage of an MEC drone, and a problem (1-2) C5 shows that each task must be at TmaxIs finished within time;
Figure GDA00035060357000000314
is the computing resource provided by the ith MEC unmanned plane for the jth cluster head unmanned plane in the tth time slot, fi maxIs the largest computational resource that the i-th MEC drone can provide in each time slot,
Figure GDA0003506035700000041
for the communication coverage of MEC drone i,
Figure GDA0003506035700000042
and the total task completion time of the jth cluster head unmanned aerial vehicle is obtained.
Further, in step S4, the process of decomposing the problem of minimizing the energy consumption of the user drone into two sub-problems includes the following two steps:
s41, given MEC unmanned aerial vehicle trajectory U, the first subproblem of solving user association variable a and resource allocation variable F is represented as:
Figure GDA0003506035700000043
s42, given the user association variable a and the resource allocation variable F, the second sub-problem of solving the MEC unmanned aerial vehicle trajectory U is represented as:
Figure GDA0003506035700000044
s.t.
Figure GDA0003506035700000045
Figure GDA0003506035700000046
further, in step S5, the process of converting the second sub-problem from the non-convex problem to the convex optimization problem by introducing a continuous relaxation variable and using the continuous convex approximation method includes the following steps:
s51, introducing a continuous relaxation variable
Figure GDA0003506035700000051
Wherein the content of the first and second substances,
Figure GDA0003506035700000052
after SCA treatment, the problem (1-4) is equivalently converted into:
Figure GDA0003506035700000053
problem (1-6) is a convex optimization problem, with constraints on
Figure GDA0003506035700000054
The convex function of (a), wherein,
Figure GDA0003506035700000055
is a cluster of trajectories of the cluster head node,
Figure GDA0003506035700000056
is the data transmission rate of member k to cluster head drone j. Further, problem (1-3) C5 is morphed according to task offload objects:
if the task is offloaded to the MEC drone for computation, question (1-3) C5 is written as:
Figure GDA0003506035700000057
if the task is computed locally at the cluster head drone, the question (1-3) C5 may be written as:
Figure GDA0003506035700000058
problems (1-3)) are re-expressed as:
Figure GDA0003506035700000061
s.t.
Figure GDA0003506035700000062
Figure GDA0003506035700000063
Figure GDA0003506035700000064
Figure GDA0003506035700000065
Figure GDA0003506035700000066
wherein the content of the first and second substances,
Figure GDA0003506035700000067
is a cluster of trajectories of the cluster head node,
Figure GDA0003506035700000068
is the data transmission rate of member k to cluster head drone j.
Further, in step S6, the process of iteratively solving the two sub-problems based on the block coordinate descent method includes the following steps:
s61, initializing the user association variable A and the calculation resource allocation variable F, MEC unmanned aerial vehicle track variable U to obtain A0、F0、U0The initial value r of the iteration times is 1;
s62, calculating the objective function value V of the problem P10＝Obj(A0,F0,U0)；
S63, fixing Ur-1To obtain Ar,Fr；
S64, fixing Ar,FrTo obtain Ur；
S65, calculating Vr＝Obj(Ar,Fr,Ur)；
S66, changing the iteration number r to r + 1;
s67, repeating S63 to S66 until | V is satisfiedr–Vr-1|/Vr-1<＝ε。
The invention has the beneficial effects that:
the method aims at minimizing the total energy consumption of the unmanned aerial vehicle of the user under the constraint of computing resources and task completion time, and converts a non-convex problem into a convex optimization problem by introducing a continuous relaxation variable and adopting a continuous convex approximation method. In order to minimize the total energy consumption of the user unmanned aerial vehicle, an iterative optimization algorithm based on a block coordinate descent method is developed to simultaneously optimize the trajectory of the MEC unmanned aerial vehicle, the association between the MEC unmanned aerial vehicle and the cluster head unmanned aerial vehicle and the allocation of computing resources.
Drawings
Fig. 1 is a schematic diagram of a system supporting mobile edge computing application under an unmanned aerial vehicle ad hoc network clustering architecture according to the present invention.
Fig. 2 is a diagram of a simulation result of a trajectory planning experiment of the cluster head unmanned aerial vehicle and the MEC unmanned aerial vehicle.
Fig. 3 is a diagram showing the experimental simulation result of the relationship between the total energy consumption of the user unmanned aerial vehicle and the number of cycles of the CPU required by the unmanned aerial vehicles of different users.
Fig. 4 is a diagram of an experimental simulation result of a relationship between total energy consumption of the user unmanned aerial vehicle and the maximum number of unmanned aerial vehicles in the cluster head allowed to access by different MEC unmanned aerial vehicles.
Fig. 5 is a diagram of an experimental simulation result of the relationship between the total energy consumption of the user unmanned aerial vehicle and the number of different MEC unmanned aerial vehicles.
Fig. 6 is a flowchart of a joint optimization method for trajectory, user association, and resource allocation of an unmanned aerial vehicle according to the present invention.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings.
It should be noted that the terms "upper", "lower", "left", "right", "front", "back", etc. used in the present invention are for clarity of description only, and are not intended to limit the scope of the present invention, and the relative relationship between the terms and the terms is not limited by the technical contents of the essential changes.
With reference to fig. 6, the present invention provides a joint optimization method for trajectory, user association, and resource allocation of an unmanned aerial vehicle, where a network includes a set of user unmanned aerial vehicles N ═ {1, 2.·, N } and a set of MEC unmanned aerial vehicles M ═ 1, 2.·, M }, and a task set is represented by S ═ {1, 2.·, S }; the user unmanned aerial vehicle cluster is divided in advanceCluster or cluster head set
Figure GDA0003506035700000071
Set of cluster members
Figure GDA0003506035700000072
The joint optimization method comprises the following steps:
s1, modeling the task points in the cluster head unmanned aerial vehicle collaborative access area as a multi-traveler problem, and planning the cluster head track by adopting a genetic algorithm.
And S2, constructing a network model, a calculation task scheduling model and an energy consumption model based on the planning result of the step S1.
And S3, according to the constructed network model, the calculation task scheduling model and the energy consumption model, solving the problem of minimizing the energy consumption of the user unmanned aerial vehicle under the constraint of calculation resources and task completion time.
And S4, decomposing the problem of minimizing the energy consumption of the user unmanned aerial vehicle into two sub-problems, wherein the first sub-problem is a user association and resource allocation optimization problem, and the second sub-problem is an MEC unmanned aerial vehicle track optimization problem.
S5, converting the second sub-problem from the non-convex problem to a convex optimization problem by introducing a continuous relaxation variable and adopting a continuous convex approximation method.
And S6, solving the first subproblem by adopting a Mosek optimization solver, solving the second subproblem by adopting a CVX optimization solver, and iteratively solving the two subproblems based on a block coordinate descent method.
In the joint optimization method for track, user association and computing resource allocation of the mobile edge computing unmanned aerial vehicle under the clustering architecture, provided by the invention, N user unmanned aerial vehicles detect a certain area, M mobile edge computing unmanned aerial vehicles (MEC unmanned aerial vehicles) provide computing service for the N user unmanned aerial vehicles, S ground tasks are provided, user unmanned aerial vehicle clusters are clustered in advance, and a cluster head cluster is formed
Figure GDA0003506035700000073
Set of cluster members
Figure GDA0003506035700000074
In the track optimization of the mobile edge computing unmanned aerial vehicle under the clustering architecture, the problems of communication performance, time delay, user resource allocation and the like exist, and the following challenges are specifically provided: 1) how to minimize the long term energy consumption of all cluster head drones by selecting the appropriate user association, i.e. whether or not a cluster head drone should offload a task, and if so, which cluster head drones to offload to the MEC in case of multiple cluster head drones); 2) by considering the limited number of airborne resources, the MEC should allocate how much computing resources to each cluster head drone to offload the mission; 3) how to control the flight trajectory, i.e. flight direction and distance, of each drone in real time.
Specifically, the method for joint optimization of the trajectory of the mobile edge computing unmanned aerial vehicle, the user association and the computing resource allocation under the clustering architecture comprises the following steps:
in step 2, a network model, a calculation task scheduling model and an energy consumption model are constructed. In the network model, consider a patrol system of unmanned aerial vehicle-assisted mobile edge calculation, which is composed of N rotor unmanned aerial vehicles and M fixed-wing unmanned aerial vehicles, and there are S ground task points in the area, where the set of user unmanned aerial vehicles N ═ {1, 2.·, N }, the set of MEC unmanned aerial vehicles M ═ {1, 2.·, M }, and the set of tasks S ═ {1, 2.·, S }. The user unmanned aerial vehicle cluster is previously divided into clusters and cluster head clusters
Figure GDA0003506035700000081
Set of cluster members
Figure GDA0003506035700000082
During a given task, the task cycle is divided into T slots, the slot length is equally divided into τ, and the set of slots is denoted as T ═ 1, 2. The position of cluster drone j is represented as
Figure GDA0003506035700000083
Where j ∈ H, the position of cluster member drone k is represented as
Figure GDA0003506035700000084
Where k ∈ CM. The flight direction of the MEC unmanned plane in each time slot is
Figure GDA0003506035700000085
Distance of flight
Figure GDA0003506035700000086
Wherein the content of the first and second substances,
Figure GDA0003506035700000087
for the maximum flight distance, v, of the ith MEC unmanned aerial vehicle in each time slotiThe flight speed of the ith MEC unmanned aerial vehicle is shown. Assume the initial position of the MEC drone is
Figure GDA0003506035700000088
The position at the time t is then indicated as
Figure GDA0003506035700000089
Wherein T ∈ T:
Figure GDA00035060357000000810
Figure GDA00035060357000000811
association variable of cluster head unmanned aerial vehicle and MEC unmanned aerial vehicle
Figure GDA00035060357000000812
Wherein i belongs to M, j belongs to H, and T belongs to T.
Figure GDA00035060357000000813
A value of 1 indicates that in the t-th time slot, the task on the cluster head drone j is unloaded to be executed on the MEC drone i,
Figure GDA00035060357000000814
a value of 0 indicates that the task is executed on cluster head drone j during the t-th time slot. On-line meterIn the task calculation scheduling model, each user unmanned aerial vehicle generates tasks at the t-th time slot
Figure GDA00035060357000000815
Where T ∈ T. In particular, the present invention relates to a method for producing,
Figure GDA00035060357000000816
can be described as
Figure GDA00035060357000000817
Wherein l is ∈ (H ≦ CM), Fl tIndicating completion
Figure GDA00035060357000000818
The total number of CPU cycles required for the CPU,
Figure GDA00035060357000000819
representing the amount of data that the user drone needs to calculate. The calculation task can be unloaded to the cluster head unmanned aerial vehicle for execution, and can also be unloaded to the MEC unmanned aerial vehicle for execution. Assuming that each MEC drone is equipped with a directional antenna with a fixed beam width beam theta, the maximum amount of computational resources that can be provided by each MEC drone is fi max. If the jth cluster head drone decides to offload the task to the ith MEC drone at the tth time slot, then the cluster head drone should be within the communication range of MEC drone i, then there are:
Figure GDA00035060357000000820
wherein, the communication coverage range of the MEC unmanned aerial vehicle i is
Figure GDA0003506035700000091
Euclidean distance from cluster head j to MEC unmanned aerial vehicle i
Figure GDA0003506035700000092
Expressed as:
Figure GDA0003506035700000093
the uplink data transmission rate from the cluster head drone j to the MEC drone i is represented as:
Figure GDA0003506035700000094
euclidean distance from cluster member to cluster head
Figure GDA0003506035700000095
Expressed as:
Figure GDA0003506035700000096
the data transmission rate from cluster member k to cluster head drone j is represented as:
Figure GDA0003506035700000097
where B is the channel bandwidth, PTrFor node transmit power, α ═ g0G0/σ2，G0≈2.2846，g0Channel gain, σ, per unit distance2Is the noise power. Consider that each drone employs Orthogonal Frequency Division Multiplexing (OFDM) channels with no interference between them. Within each time period, it is assumed that each MEC drone i can only be served at most simultaneously
Figure GDA0003506035700000098
The shelf user unmanned aerial vehicle then has:
Figure GDA0003506035700000099
in each time period, since the computational resources that each MEC drone i can provide are limited, there are:
Figure GDA00035060357000000910
wherein the content of the first and second substances,
Figure GDA00035060357000000911
is the computing resource provided by the ith MEC unmanned plane for the jth cluster head unmanned plane in the tth time slot, fi maxIs the largest computing resource that the i-th MEC drone can provide in each time slot.
The task completion time is divided into transmission time and calculation time, assuming that the calculation result return time is negligible. The transmission time is divided into two stages, firstly, the cluster member k transmits the cluster head unmanned aerial vehicle j, and then the cluster head unmanned aerial vehicle j transmits the task of the whole cluster to the MEC unmanned aerial vehicle i. The calculation time is divided into local cluster head calculation time and MEC unmanned plane calculation time. Therefore, the total task completion time of the jth cluster head drone may be expressed as:
Figure GDA00035060357000000912
wherein the content of the first and second substances,
Figure GDA00035060357000000913
the transmission time for the cluster member k to offload its computation task to the cluster head drone j in the time slot t is represented as:
Figure GDA0003506035700000101
Figure GDA0003506035700000102
indicating the transmission time for the cluster head drone j to offload aggregated tasks of the whole cluster to the MEC drone i in the t time slot, which is expressed as:
Figure GDA0003506035700000103
Figure GDA0003506035700000104
representing the computation time required for the task to be offloaded onto the MEC drone for execution at the t slot, is represented as:
Figure GDA0003506035700000105
Figure GDA0003506035700000106
the calculation time required for the task to execute on the cluster head unmanned aerial vehicle in the t time slot is represented as:
Figure GDA0003506035700000107
suppose each task needs to be at TmaxCompleted within time, namely:
Figure GDA0003506035700000108
in the energy consumption model, the total energy consumption of tasks in the t-th time slot is represented as:
Figure GDA0003506035700000109
wherein the content of the first and second substances,
Figure GDA00035060357000001010
for the transmission of cluster heads to the MEC server,
Figure GDA00035060357000001011
the energy consumption for the transmission of cluster members to the cluster head,
Figure GDA00035060357000001012
calculating consumed energy consumption, k, locally for cluster headsjMore than or equal to 0 is effective switch capacitance, vjIs a constant.
And 3, according to the constructed model, solving the problem of minimizing the energy consumption of the unmanned aerial vehicle of the user under the constraint of computing resources and task completion time. MEC unmanned aerial vehicle trajectory variable is defined as
Figure GDA00035060357000001013
The correlation variable of the MEC unmanned aerial vehicle and the cluster head unmanned aerial vehicle is defined as
Figure GDA00035060357000001014
A computing resource allocation variable is defined as
Figure GDA00035060357000001015
The problem of minimizing the energy consumption of a cluster of user drones can be modeled as:
Figure GDA0003506035700000111
where U, A, F is an optimization variable, problem (1-2) C1 indicates that cluster-head drones have and only offload tasks to a certain edge computing drone or perform computing tasks on themselves, and problem (1-2) C2 indicates that each MEC drone i can only serve at most simultaneously
Figure GDA0003506035700000112
A cluster head drone is erected, problem (1-2) C3 indicates that the sum of the task offload amounts of cluster head drone j cannot exceed the amount of computing resources that MEC drone i can provide, problem (4-17) C4 indicates that the cluster head drone to be offloaded is to be within the communication coverage of MEC drone, and problem (1-2) C5 indicates that each task must be at TmaxAnd is finished within time.
In step 4, the problem is decomposed into two sub-problems, the first sub-problem is a user association and resource allocation optimization problem, and the second sub-problem is an MEC unmanned aerial vehicle trajectory optimization problem. Sub-problem one, given MEC drone trajectory U, the sub-problem (P1) to solve user association variable a and resource allocation variable F may be represented as:
Figure GDA0003506035700000113
wherein, problem (1-3) C5 may be morphed according to task off-load objects, and if tasks are off-loaded to MEC drones for computation, then problem (1-3) C5 may be written as:
Figure GDA0003506035700000114
if the task is computed locally at the cluster head drone, then the question (4-18) C5 may be written as:
Figure GDA0003506035700000121
then the problem (4-18) can be re-expressed as:
Figure GDA0003506035700000122
wherein the content of the first and second substances,
Figure GDA0003506035700000123
is the trajectory of the cluster head node.
Given a user association variable a and a resource allocation variable F, the problem (1-4) of solving the MEC unmanned aerial vehicle trajectory U can be expressed as:
Figure GDA0003506035700000124
in step 5, the second sub-problem is transformed from the non-convex problem to the convex optimization problem by introducing a continuous relaxation variable and using a continuous convex approximation method. Problems (1-4) are a continuous function of MEC drone trajectory variable U, but the objective function and constraints of this sub-problem are both non-convex. To solve the problem, a continuous relaxation is introducedVariables of
Figure GDA0003506035700000131
Wherein the content of the first and second substances,
Figure GDA0003506035700000132
then the question (1-4) can be converted into a question (1-28),
Figure GDA0003506035700000133
it can be seen that the objective function of the problem (1-28) is convex, with constraints on
Figure GDA0003506035700000134
The constraint is therefore a non-convex constraint with respect to the variable U, and a continuous convex approximation is used to solve the problem. Given the local value of the last iteration
Figure GDA0003506035700000135
The strict lower bounds for C1 and C2 in problems (1-28) are:
Figure GDA0003506035700000136
wherein the content of the first and second substances,
Figure GDA0003506035700000137
Figure GDA0003506035700000138
after SCA treatment, the problems (1-28) are equivalently converted into:
Figure GDA0003506035700000139
the problem (1-32) is a convex optimization problem.
In the step 6, a Mosek optimization solver is adopted to solve the first subproblem, a CVX optimization solver is adopted to solve the second subproblem, and the two subproblems are subjected to iterative solution based on a block coordinate descent method. It can be seen from the problems (1-3) that the objective functions and constraints are linear on both A and F, and the problems (1-3) are a classical Multi-Dimensional Multi-Choice Knapsack Problem (Multiple-Choice Multi-Dimensional 0-1 Knapack Problem, MMKP), which is usually a NP-hard Problem but can be best solved using the branch-defining method using the Standard optimizer MOSEK [89 ]. Problem (1-4) is a convex optimization problem, which is solved efficiently using CVX. The iterative solution of the two subproblems based on the block coordinate descent method comprises the following steps:
s61, initializing the user association variable A and the calculation resource allocation variable F, MEC unmanned aerial vehicle track variable U to obtain A0、F0、U0The initial value r of the number of iterations is 1.
S62, calculating the objective function value V of the problem P10＝Obj(A0,F0,U0)。
S63, fixing Ur-1To obtain Ar,Fr。
S64, fixing Ar,FrTo obtain Ur。
S65, calculating Vr＝Obj(Ar,Fr,Ur)。
S66, changing the iteration number r to r + 1.
S67, repeating S63 to S66 until | V is satisfiedr–Vr-1|/Vr-1<＝ε。
The invention discloses a moving edge computing unmanned aerial vehicle track, user association and computing resource allocation joint optimization method under a clustering framework, and FIG. 1 is a scene of an unmanned aerial vehicle ad hoc network supporting moving edge computing application under the clustering framework provided by the invention.
Fig. 2 is a plot of cluster head drones and MEC drones trajectories. Each cluster head unmanned aerial vehicle starts from different initial positions and finally returns to the takeoff position, and each closed line represents the flight track of one cluster head unmanned aerial vehicle. And 4, completing the access of the task points in the coverage area by the cooperation of the unmanned planes with the cluster heads in the shortest distance. The initial position coordinates of the two MEC unmanned aerial vehicles are respectively (200,0), (200 ), the end point coordinates are respectively (0,200), (0,0), and the dotted line is the optimized track of the two MEC unmanned aerial vehicles.
Fig. 3 shows the total energy consumption of the user drone versus the number of CPU cycles required for the user drone workload. As the number of CPUs required for user drones increases, more user drones may have to perform tasks at the local cluster head due to the relatively limited computing power of each MEC drone, resulting in an increase in the overall energy consumption of all user drones. The performance of the AFU algorithm is better than other two references, and the FixU is Local. This is because the calculation tasks of the Local scheme are all executed locally at the cluster head, resulting in very high total energy consumption of the user unmanned aerial vehicle, while in the fix u scheme, the MEC unmanned aerial vehicle flies in a fixed diagonal, which can provide calculation service for the user unmanned aerial vehicle, and because the MEC unmanned aerial vehicle trajectory is not optimized, it is not possible to provide calculation service for more user unmanned aerial vehicles for a long time, so the energy consumption performance is worse than AFU. More specifically, AFU reduces overall energy consumption of the user drones, on average, 294.63% and 43.77% respectively, compared to Local, fix u, when the workload of each user drone varies from 4 × 108 CPU cycles to 109 CPU cycles.
Fig. 4 shows the total energy consumption of user drones versus the maximum number of cluster head drones that the MEC drone is allowed to access. As can be seen from the figure, with the change of the maximum number of cluster-head drones allowed to be accessed by the MEC drones, the total energy consumption of the user drones in other algorithms is reduced except for the algorithm for executing the calculation tasks at all local cluster heads, because the MEC drones have stronger communication access capability, the cluster-head drones can unload more tasks to the MEC drones for execution. It can be seen that when the maximum number of cluster head unmanned aerial vehicles allowed to access is greater than 1, the total energy consumption of the user unmanned aerial vehicles in the AFU and fix u algorithms is greatly reduced. Compared with Local and FixU, the AFU reduces 409.21% and 128.14% of the total energy consumption of the unmanned aerial vehicle of the user on average.
Fig. 5 shows the total energy consumption of the user drones versus the number of MEC drones. It can be seen from the figures, except for all
Besides the algorithm that tasks are executed on the local cluster head, the total energy consumption of the user unmanned aerial vehicle in the AFU algorithm and the FixU algorithm follows MEC
The increase of the number of the unmanned aerial vehicles is reduced, the energy consumption of the AFU is lower than that of the FixU, and when the number of the MEC unmanned aerial vehicles is increased to 4, the total energy consumption of the user unmanned aerial vehicles of the AFU and the FixU is close. This is because as the number of MEC drons increases, cluster head drones may have more opportunities to be within the communication coverage of MEC drons, that is, cluster head drones can offload more computing services to MEC drons, thereby reducing the total energy consumption of user drons. Meanwhile, the number of the MEC unmanned aerial vehicles is increased to a certain value, even if the path of the MEC unmanned aerial vehicles is not optimized, a plurality of MEC unmanned aerial vehicles can cover most of areas, and calculation service is provided for more cluster head unmanned aerial vehicles, so that after the number of the MEC unmanned aerial vehicles is increased to a certain value, the AFU and the FixU approach to the energy consumption performance. When the MEC drone count increases from 1 to 4, AFU reduces 496.96% and 80.66% of the overall energy consumption of the user drone, on average, compared to Local, fix u.
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.

Claims (7)

1. A joint optimization method for unmanned aerial vehicle track, user association and resource allocation is characterized in that a network is set to comprise a user unmanned aerial vehicle set
Figure FDA00035060356900000115
And MEC unmanned plane set
Figure FDA00035060356900000116
A task set is represented by S ═ {1, 2.., S }; the user unmanned aerial vehicle cluster is previously divided into clusters and cluster head clusters
Figure FDA0003506035690000011
Set of cluster members
Figure FDA0003506035690000012
The joint optimization method comprises the following steps:
s1, modeling the task points in the cluster head unmanned aerial vehicle collaborative access area as a multi-traveler problem, and planning the cluster head track by adopting a genetic algorithm;
s2, constructing a network model, a calculation task scheduling model and an energy consumption model based on the planning result of the step S1;
s3, according to the constructed network model, the calculation task scheduling model and the energy consumption model, the problem of minimizing the energy consumption of the user unmanned aerial vehicle under the constraint of calculation resources and task completion time is solved;
s4, decomposing the problem of minimizing the energy consumption of the user unmanned aerial vehicle into two sub-problems, wherein the first sub-problem is a user association and resource allocation optimization problem, and the second sub-problem is an MEC unmanned aerial vehicle track optimization problem;
s5, converting the second sub-problem from the non-convex problem into a convex optimization problem by introducing a continuous relaxation variable and adopting a continuous convex approximation method;
and S6, solving the first subproblem by adopting a Mosek optimization solver, solving the second subproblem by adopting a CVX optimization solver, and iteratively solving the two subproblems based on a block coordinate descent method.
2. The joint optimization method for unmanned aerial vehicle trajectory, user association and resource allocation as claimed in claim 1, wherein in step S2, the process of constructing the network model, the computation task scheduling model and the energy consumption model comprises the following steps:
s21, in the network model, considering an unmanned aerial vehicle auxiliary moving edge calculation patrol system consisting of N rotor unmanned aerial vehicles and M fixed wing unmanned aerial vehicles, wherein S ground task points are arranged in the area;
s22, during a designated task, dividing a task cycle into T slots, wherein the slot length is equally divided into τ, and the slot set is denoted as T ═ 1, 2.., T }; the position of cluster drone j is represented as
Figure FDA0003506035690000013
Where j ∈ H, the position of cluster member drone k is represented as
Figure FDA0003506035690000014
Wherein k belongs to CM; the flight direction of the MEC unmanned plane in each time slot is
Figure FDA0003506035690000015
Distance of flight
Figure FDA0003506035690000016
Wherein the content of the first and second substances,
Figure FDA0003506035690000017
for the maximum flight distance, v, of the ith MEC unmanned aerial vehicle in each time slotiThe flight speed of the ith MEC unmanned aerial vehicle is taken as the flight speed of the ith MEC unmanned aerial vehicle; in the calculation task scheduling model, each user unmanned aerial vehicle l generates a task at the t-th time slot
Figure FDA0003506035690000018
Figure FDA0003506035690000019
Wherein l ∈ (H ≦ CM),
Figure FDA00035060356900000110
indicating completion
Figure FDA00035060356900000111
Total number of CPU cycles required，
Figure FDA00035060356900000112
Representing the data amount required to be calculated by the user unmanned plane l;
Figure FDA00035060356900000113
a value of 1 indicates that in the t-th time slot, the task on the cluster head drone j is unloaded to be executed on the MEC drone i,
Figure FDA00035060356900000114
a value of 0 indicates that a task is executed on the cluster head unmanned aerial vehicle j in the t-th time slot;
s23, in the energy consumption model, the total task energy consumption in the t-th time slot is represented as:
Figure FDA0003506035690000021
wherein the content of the first and second substances,
Figure FDA0003506035690000022
for the transmission of cluster heads to the MEC server,
Figure FDA0003506035690000023
the energy consumption for the transmission of cluster members to the cluster head,
Figure FDA0003506035690000024
calculating consumed energy consumption, k, locally for cluster headsjMore than or equal to 0 is an effective switch capacitor;
Figure FDA0003506035690000025
is the Euclidean distance from the cluster member k to the cluster head unmanned plane j;
Figure FDA0003506035690000026
representing the Euclidean distance from the cluster head unmanned aerial vehicle j to the MEC unmanned aerial vehicle i; b is a channel bandwidth and is a channel bandwidth,
Figure FDA0003506035690000027
the node transmit power for cluster head drone j,
Figure FDA0003506035690000028
for node transmit power of drone k, α ═ g0G0/σ2，G0≈2.2846，g0Channel gain, σ, per unit distance2Is the noise power;
Figure FDA0003506035690000029
is the local computing resource provided by the cluster head unmanned aerial vehicle j at the moment t, vjIs a local calculation of the energy consumption parameter, HiIs the flying height of MEC unmanned plane i, hjIs the flying height of the cluster head drone j,
Figure FDA00035060356900000210
is the local calculated time of cluster head drone j.
3. The joint optimization method for unmanned aerial vehicle trajectory, user association and resource allocation according to claim 2, wherein in step S3, the process of proposing the problem of minimizing energy consumption of the user unmanned aerial vehicle under the constraint of computation resources and task completion time includes the following steps:
s31, defining the MEC unmanned aerial vehicle track variable as
Figure FDA00035060356900000211
The correlation variable of the MEC unmanned aerial vehicle and the cluster head unmanned aerial vehicle is defined as
Figure FDA00035060356900000212
A computing resource allocation variable is defined as
Figure FDA00035060356900000213
S32, modeling the problem of minimizing the energy consumption of the user unmanned aerial vehicle cluster as:
Figure FDA00035060356900000214
where U, A, F is an optimization variable, problem (1-2) C1 indicates that cluster-head drones have and only offload tasks to a certain edge computing drone or perform computing tasks on themselves, and problem (1-2) C2 indicates that each MEC drone i can only serve at most simultaneously
Figure FDA0003506035690000031
A cluster head drone is erected, a problem (1-2) C3 shows that the sum of the task unloading capacity of a cluster head drone j cannot exceed the amount of computing resources that an MEC drone i can provide, a problem (1-2) C4 shows that a cluster head drone to be unloaded is to be within the communication coverage of an MEC drone, and a problem (1-2) C5 shows that each task must be at TmaxIs finished within time;
Figure FDA0003506035690000032
is the computing resource provided by the ith MEC unmanned plane for the jth cluster head unmanned plane in the tth time slot, fi maxIs the largest computational resource that the i-th MEC drone can provide in each time slot,
Figure FDA0003506035690000033
for the communication coverage of MEC drone i,
Figure FDA0003506035690000034
and the total task completion time of the jth cluster head unmanned aerial vehicle is obtained.
4. The joint optimization method for unmanned aerial vehicle trajectory, user association and resource allocation as claimed in claim 3, wherein in step S4, the process of decomposing the problem of minimizing energy consumption of user unmanned aerial vehicle into two sub-problems comprises the following two steps:
s41, given MEC unmanned aerial vehicle trajectory U, the first subproblem of solving user association variable a and resource allocation variable F is represented as:
Figure FDA0003506035690000035
s42, given the user association variable a and the resource allocation variable F, the second sub-problem of solving the MEC unmanned aerial vehicle trajectory U is represented as:
Figure FDA0003506035690000041
s.t.
Figure FDA0003506035690000042
Figure FDA0003506035690000043
Figure FDA0003506035690000044
representing the horizontal coordinate of cluster head drone j at time t.
5. The joint optimization method for unmanned aerial vehicle trajectory, user association and resource allocation as claimed in claim 4, wherein in step S5, the process of transforming the second sub-problem from the non-convex problem to the convex optimization problem by introducing a continuous relaxation variable and using a continuous convex approximation method comprises the following steps:
s51, introducing a continuous relaxation variable
Figure FDA0003506035690000045
Wherein the content of the first and second substances,
Figure FDA0003506035690000046
after SCA treatment, the problem (1-4) is equivalently converted into:
Figure FDA0003506035690000047
problem (1-6) is a convex optimization problem, with constraints on
Figure FDA0003506035690000048
The convex function of (a), wherein,
Figure FDA0003506035690000051
is a cluster of trajectories of the cluster head node,
Figure FDA0003506035690000052
is the data transmission rate of member k to cluster head drone j,
Figure FDA0003506035690000053
and
Figure FDA0003506035690000054
there is no physical meaning for the intermediate variables.
6. The joint optimization method for unmanned aerial vehicle trajectory, user association and resource allocation according to claim 4, wherein the problem (1-3) C5 is transformed according to task offload objects:
if the task is offloaded to the MEC drone for computation, question (1-3) C5 is written as:
Figure FDA0003506035690000055
Figure FDA0003506035690000056
representing the uplink data transmission rate from the cluster head unmanned aerial vehicle j to the MEC unmanned aerial vehicle i;
if the task is computed locally at the cluster head drone, the question (1-3) C5 may be written as:
Figure FDA0003506035690000057
problems (1-3)) are re-expressed as:
Figure FDA0003506035690000058
s.t.
Figure FDA0003506035690000059
Figure FDA00035060356900000510
Figure FDA00035060356900000511
Figure FDA00035060356900000512
Figure FDA00035060356900000513
wherein the content of the first and second substances,
Figure FDA00035060356900000514
is a cluster of trajectories of the cluster head node,
Figure FDA00035060356900000515
is the data transmission rate of member k to cluster head drone j.
7. The joint optimization method for unmanned aerial vehicle trajectory, user association and resource allocation as claimed in claim 1, wherein in step S6, the process of iteratively solving two sub-problems based on the block coordinate descent method includes the following steps:
s61, initializing the user association variable A and the calculation resource allocation variable F, MEC unmanned aerial vehicle track variable U to obtain A0、F0、U0The initial value r of the iteration times is 1;
s62, calculating an objective function value V for minimizing the energy consumption problem of the user unmanned aerial vehicle cluster0＝Obj(A0,F0,U0)；
S63, fixing Ur-1To obtain Ar,Fr；
S64, fixing Ar,FrTo obtain Ur；
S65, calculating Vr＝Obj(Ar,Fr,Ur)；
S66, changing the iteration number r to r + 1;
s67, repeating S63 to S66 until | V is satisfiedr–Vr-1|/Vr-1<ε is a precision threshold and is a constant.
CN202110252974.9A 2021-03-08 2021-03-08 Unmanned aerial vehicle track, user association and resource allocation joint optimization method Active CN112995913B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110252974.9A CN112995913B (en) 2021-03-08 2021-03-08 Unmanned aerial vehicle track, user association and resource allocation joint optimization method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110252974.9A CN112995913B (en) 2021-03-08 2021-03-08 Unmanned aerial vehicle track, user association and resource allocation joint optimization method

Publications (2)

Publication Number Publication Date
CN112995913A CN112995913A (en) 2021-06-18
CN112995913B true CN112995913B (en) 2022-04-08

Family

ID=76335957

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110252974.9A Active CN112995913B (en) 2021-03-08 2021-03-08 Unmanned aerial vehicle track, user association and resource allocation joint optimization method

Country Status (1)

Country Link
CN (1) CN112995913B (en)

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113472426B (en) * 2021-07-01 2022-06-28 云南大学 Fair perception task scheduling and resource allocation method
CN113556764B (en) * 2021-07-30 2022-05-31 云南大学 Method and system for determining calculation rate based on mobile edge calculation network
CN113825145B (en) * 2021-09-15 2022-08-12 云南大学 Unmanned aerial vehicle system service method and system for user experience
CN113840329B (en) * 2021-09-15 2023-05-30 云南大学 Collaborative computing and cache scheduling policy method and system in unmanned aerial vehicle network
CN114035610B (en) * 2021-11-15 2024-03-26 武汉大学 Unmanned intelligent cluster joint track design method
CN114143893A (en) * 2021-12-08 2022-03-04 中国石油大学(华东) Unmanned aerial vehicle resource allocation and track optimization method based on mobile edge calculation and microwave energy transmission
CN114245436B (en) * 2021-12-30 2024-04-02 杭州电子科技大学 Unmanned aerial vehicle auxiliary mobile edge network clustering method and device
CN114531193B (en) * 2022-01-04 2023-11-10 无锡市市政设施养护管理有限公司 Bridge state monitoring method based on unmanned aerial vehicle cellular topology networking and mobile edge calculation
CN114500533B (en) * 2022-01-18 2023-06-23 南京邮电大学 Unmanned aerial vehicle auxiliary mobile edge computing system optimization method based on user cooperation
CN114125708B (en) * 2022-01-20 2022-04-15 南京信息工程大学 Unmanned aerial vehicle cluster trajectory optimization and task unloading method based on digital twinning
CN114520991B (en) * 2022-01-27 2023-07-28 重庆邮电大学 Unmanned aerial vehicle cluster-based edge network self-adaptive deployment method
CN114422060B (en) * 2022-03-29 2022-06-17 军事科学院系统工程研究院网络信息研究所 Method and system for constructing unmanned aerial vehicle communication channel model
CN115243285A (en) * 2022-06-14 2022-10-25 北京理工大学长三角研究院(嘉兴) Safety calculation unloading method based on unmanned aerial vehicle network
CN115412966A (en) * 2022-07-28 2022-11-29 国网内蒙古东部电力有限公司信息通信分公司 Green energy-saving unloading method based on multi-edge node cooperation under power Internet of things
CN116541153B (en) * 2023-07-06 2023-10-03 南昌工程学院 Task scheduling method and system for edge calculation, readable storage medium and computer

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3746999A1 (en) * 2018-01-29 2020-12-09 Interdigital Patent Holdings, Inc. Methods of a mobile edge computing (mec) deployment for unmanned aerial system traffic management (utm) system applications
CN110730031B (en) * 2019-10-22 2022-03-11 大连海事大学 Unmanned aerial vehicle track and resource allocation joint optimization method for multi-carrier communication
CN112399375B (en) * 2020-06-19 2023-01-31 南京邮电大学 Unmanned aerial vehicle auxiliary edge computing unloading method based on terminal energy efficiency optimization

Also Published As

Publication number Publication date
CN112995913A (en) 2021-06-18

Similar Documents

Publication Publication Date Title
CN112995913B (en) Unmanned aerial vehicle track, user association and resource allocation joint optimization method
CN110730031B (en) Unmanned aerial vehicle track and resource allocation joint optimization method for multi-carrier communication
CN111552313B (en) Multi-unmanned aerial vehicle path planning method based on edge calculation dynamic task arrival
CN113543176A (en) Unloading decision method of mobile edge computing system based on assistance of intelligent reflecting surface
CN112399375B (en) Unmanned aerial vehicle auxiliary edge computing unloading method based on terminal energy efficiency optimization
CN113282352B (en) Energy-saving unloading method based on multi-unmanned aerial vehicle cooperative auxiliary edge calculation
CN115827108B (en) Unmanned aerial vehicle edge calculation unloading method based on multi-target deep reinforcement learning
Liao et al. Energy minimization for UAV swarm-enabled wireless inland ship MEC network with time windows
Sha et al. DRL-based task offloading and resource allocation in multi-UAV-MEC network with SDN
Wei et al. Joint UAV trajectory planning, DAG task scheduling, and service function deployment based on DRL in UAV-empowered edge computing
Ren et al. Computation offloading game in multiple unmanned aerial vehicle‐enabled mobile edge computing networks
CN112579290B (en) Computing task migration method of ground terminal equipment based on unmanned aerial vehicle
Huda et al. Deep reinforcement learning-based computation offloading in uav swarm-enabled edge computing for surveillance applications
Shi et al. Energy-efficient UAV-enabled computation offloading for industrial internet of things: a deep reinforcement learning approach
CN116887355A (en) Multi-unmanned aerial vehicle fair collaboration and task unloading optimization method and system
Khan et al. Dqn-based Proactive Trajectory Planning of UAVs in Multi-access edge computing
Shi et al. Multi-UAV-assisted computation offloading in DT-based networks: A distributed deep reinforcement learning approach
Jiang et al. Age-of-Information Minimization for UAV-Based Multi-View Sensing and Communication
Tang et al. A cooperative MEC framework based on multi-UAV and AP to minimize weighted energy consumption
Linpei et al. Energy-efficient computation offloading assisted by RIS-based UAV
Gao et al. MO-AVC: Deep Reinforcement Learning Based Trajectory Control and Task Offloading in Multi-UAV enabled MEC Systems
Dong et al. Deep Progressive Reinforcement Learning-Based Flexible Resource Scheduling Framework for IRS and UAV-Assisted MEC System
Wu et al. UAV-Assisted Data Synchronization for Digital-Twin-Enabled Vehicular Networks
CN113825145B (en) Unmanned aerial vehicle system service method and system for user experience
CN114598721B (en) High-energy-efficiency data collection method and system based on joint optimization of track and resources

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