CN116366127A - Task completion rate maximization method for unmanned aerial vehicle auxiliary multi-MEC server - Google Patents
Task completion rate maximization method for unmanned aerial vehicle auxiliary multi-MEC server Download PDFInfo
- Publication number
- CN116366127A CN116366127A CN202310223494.9A CN202310223494A CN116366127A CN 116366127 A CN116366127 A CN 116366127A CN 202310223494 A CN202310223494 A CN 202310223494A CN 116366127 A CN116366127 A CN 116366127A
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- user
- energy consumption
- task
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000005265 energy consumption Methods 0.000 claims abstract description 93
- 238000004364 calculation method Methods 0.000 claims abstract description 65
- 238000005457 optimization Methods 0.000 claims abstract description 32
- 230000005540 biological transmission Effects 0.000 claims abstract description 19
- 238000004422 calculation algorithm Methods 0.000 claims description 22
- 230000009977 dual effect Effects 0.000 claims description 10
- 230000001364 causal effect Effects 0.000 claims description 4
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 238000006467 substitution reaction Methods 0.000 claims description 3
- 238000005562 fading Methods 0.000 abstract description 3
- 238000004088 simulation Methods 0.000 description 29
- 238000004891 communication Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 230000008878 coupling Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 4
- 238000005859 coupling reaction Methods 0.000 description 4
- 230000007423 decrease Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013468 resource allocation Methods 0.000 description 2
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 description 1
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 description 1
- 241000277284 Salvelinus fontinalis Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/185—Space-based or airborne stations; Stations for satellite systems
- H04B7/18502—Airborne stations
- H04B7/18504—Aircraft used as relay or high altitude atmospheric platform
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/18—Network planning tools
-
- 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
- H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
- H04W52/02—Power saving arrangements
- H04W52/0209—Power saving arrangements in terminal devices
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Astronomy & Astrophysics (AREA)
- Aviation & Aerospace Engineering (AREA)
- General Physics & Mathematics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a task completion rate maximization method for an unmanned aerial vehicle auxiliary multi-MEC server, which mainly solves the problems of serious channel fading and limited calculation resources of a single MEC server in the prior art. Comprising the following steps: 1) Building a network system model of a multi-MEC server assisted by the unmanned aerial vehicle; 2) The method comprises the steps that a time division multiple access protocol is adopted, a user unloads a task input data part to an unmanned aerial vehicle, and the unmanned aerial vehicle unloads a task input data part which cannot be completed by the unmanned aerial vehicle to a plurality of base stations on the ground for calculation; 3) Searching the optimal task completion rate, CPU calculation frequency, transmission power, unloading decision and unmanned plane track by maximizing the user task completion rate and minimizing the system energy consumption; 4) And setting working parameters of the system according to the optimizing result to realize optimization. The method and the system can effectively improve network computing resources, realize stability under a high-speed computing task environment, reduce energy consumption of the unmanned aerial vehicle and a user, and can be used for an edge computing system.
Description
Technical Field
The invention belongs to the technical field of wireless communication, and further relates to an edge computing technology, in particular to a task completion rate maximization method of an unmanned aerial vehicle auxiliary multi-mobile edge computing MEC server, which can be used for improving the task completion rate by unmanned aerial vehicle auxiliary multi-edge server scene user unloading tasks in an edge computing system.
Background
As the number of wireless access devices continues to grow, the large amounts of data collected from these devices need to be transferred from one place to another for intelligent decision making, which places a tremendous burden on the wireless communication infrastructure, which is limited in the radio spectrum. In addition, many wireless access devices are implementing many attractive new applications such as real-time video analysis, augmented or virtual reality and artificial intelligence. However, these computational tasks are computationally intensive and delay sensitive, relying on our ability to process data and extract useful information quickly.
The mobile edge computing can transfer heavy computing tasks from users with limited power to the edge, and process data in a place close to the users, so that flow bottlenecks in a core network and a backhaul network are reduced, rapid processing of the data is realized, and the problem of network delay is effectively solved. In general, this decentralized cloud architecture constitutes a post technology for 5G systems, changing the traditional cloud-based data processing paradigm by providing cloud computing capabilities and network-edge service environments.
However, in conventional mobile edge computing networks, mobile edge computing servers are typically deployed on the ground in a fixed manner. Effective terrestrial wireless communication cannot be established with terrestrial users due to signal blockage and shadowing caused by the common occlusion of the mobile edge computing server by terrestrial obstructions. In view of this disadvantage, with the advantages of an unmanned aerial vehicle that is easy to deploy, flexible to move, and line-of-sight link connection, a mobile edge computing network supported by the unmanned aerial vehicle has recently been proposed as a promising solution to improve the reliable connection for ground users.
Hu Q, cai Y, yu G et al in their published paper "Joint offloading and trajectory design for unmanned aerial vehicle-enabled mobile edge computing systems" (IEEE Internet of Things Journal, 2019:1879-1892) consider deploying unmanned aerial vehicles to provide computing services for users, and by optimizing the offloading rate, user scheduling, and unmanned aerial vehicle trajectories, the problem of minimizing the sum of maximum delays between users is addressed. Hua M, wang Y, li C et al consider a drone in its published paper "drone-aided mobile edge computing systems with one by one access scheme" (IEEE Transactions on Green Communications and Networking, 2019:664-678) to help the user offload computing tasks, either locally or to the drone. Unlike the above study, hu X, wong KK, yang K et al in their published paper "unmanned aerial vehicle-assisted relaying and edge computing: scheduling and trajectory optimization" (IEEE transactions on wireless communications, 2019:4738-4752) utilize unmanned aerial vehicles as relay stations to assist in the offloading of computational tasks by users, the unmanned aerial vehicles not only can act as mobile edge computational servers, but also can provide communication services to users by forwarding received computational tasks to a base station for remote computation. Therefore, the stability of the system in a high-speed computing task environment can be improved, and the task processing time delay of the unmanned aerial vehicle is reduced. The above-mentioned research on the unmanned aerial vehicle-assisted mobile edge computing network mainly considers the unmanned aerial vehicle as a mobile edge computing server or only considers one base station as a mobile edge computing server, but in a practical application scenario, the unmanned aerial vehicle has limited energy and computing resources, and a plurality of base stations may exist on the ground, so that the above-mentioned method is difficult to solve practical problems.
Disclosure of Invention
The invention aims to provide a task completion rate maximization method of an unmanned aerial vehicle auxiliary multi-MEC server aiming at the defects of the prior art. Searching the optimal task completion rate, CPU calculation frequency, transmission power, unloading decision and unmanned plane track by maximizing the user task completion rate and minimizing the system energy consumption; the method solves the problems that the traditional mobile edge computing network has serious channel fading and the computing resources of a single mobile edge computing server are limited, effectively improves the stability of the edge computing network in a computing task high-speed environment, and reduces the energy consumption of the unmanned aerial vehicle and the user.
The basic idea for realizing the invention is as follows: under the scene of an unmanned aerial vehicle auxiliary multi-mobile-edge computing server, a user adopts a time division multiple access protocol to offload a task input data part to the unmanned aerial vehicle, and the unmanned aerial vehicle and the user share computing resources to help the user to compute tasks; the unmanned aerial vehicle also adopts a time division multiple access protocol to further partially unload the task input data which cannot be completed to a plurality of base stations on the ground for calculation, so that the energy consumption of the unmanned aerial vehicle is saved, and the calculation efficiency is improved.
In order to achieve the above object, the technical scheme of the present invention includes the following steps:
(1) Building an edge computing network model:
an edge computing network model consisting of one unmanned plane, K users and M base stations carrying mobile edge computing MEC servers is built; order theAnd->Respectively representing a set of users and base stations, wherein k and m respectively represent a kth user and an mth base station on which a mobile edge computing server is mounted;
(2) Dividing the time slot structures of users and unmanned aerial vehicles:
when the limited task is completedThe T is discretized into N equal time slots, so thatRepresenting a set of N time slots; the duration of each time slot is τ=t/N, assuming that the position of the drone is unchanged during each time slot; a user adopts a time division multiple access protocol to divide each time slot into K 'equal sub-time slots, K' =K, the duration of each sub-time slot is delta=T/(NK '), and the user K unloads task input data of the user K in the K' sub-time slot corresponding to the user K; the unmanned aerial vehicle also adopts K' time division multiple access protocols with equal time division, each sub-time slot is divided into M small sub-time slots, and the size of each small sub-time slot is determined by an unloading decision variable alpha U,k,m [n]Determining;
(3) Obtaining an optimal optimization variable under the scene of an unmanned aerial vehicle auxiliary multi-mobile-edge computing server:
(3.1) using a three-dimensional Euclidean coordinate system, assuming that the positions of the base station and all users are fixed on the ground at zero altitude, the horizontal positions of the base station m and the user k are respectively b m And w k The method comprises the steps of carrying out a first treatment on the surface of the Assuming that the unmanned aerial vehicle flies at a fixed height H within a task completion time T, wherein H is more than 0; and the flight start point and the flight end point are respectively set as q I And q F The method comprises the steps of carrying out a first treatment on the surface of the Based on the discrete path planning method, the horizontal position q [ n ] of the unmanned aerial vehicle in the nth time slot is obtained]=q[nτ]Wherein q [0 ]]=q I ,q[N]=q F The method comprises the steps of carrying out a first treatment on the surface of the Assuming that the wireless channels between the unmanned aerial vehicle and the base station and between the unmanned aerial vehicle and the user are all dominated by the line-of-sight link, obtaining the channel power gain h between the unmanned aerial vehicle and the user k and between the unmanned aerial vehicle and the base station m in the time slot n U,k [n]And g U,m [n];
(3.2) assuming each user has a defined computing task to be performed, the computing task is composed of tripletsRepresentation, wherein I k Representing the size of the input data of the computing task, C k Representing the computational resources required to input 1-bit data, T k Maximum tolerable delay for user k; maximum delay tolerance for all usersIdentical, i.e. T k =t; let the task completion rate per user be μ k ;
(3.3) unloading one part of the calculation tasks to the unmanned aerial vehicle for calculation by each user by adopting a time division multiple access protocol, and carrying out local calculation by the other part; the unmanned aerial vehicle collects calculation tasks unloaded by users through optimizing flight trajectories, and also adopts a time division multiple access protocol to unload one part of the collected calculation tasks to a plurality of base stations on the ground for calculation, and the other part of the collected calculation tasks carries out unmanned aerial vehicle local calculation; obtaining the task data volume of the local calculation execution of the kth user in the time slot nAnd energy consumption->Task data amount offloaded to unmanned plane by kth user +.>And transmission energy consumption->The unmanned plane locally calculates the task data quantity of executing the kth user +.>And energy consumption->Unmanned aerial vehicle offloads task data volume of kth user to base station m +.>And transmission energy consumption->Flight energy consumption of unmanned aerial vehicle +.>
(3.4) obtaining the total energy consumption E of user k in each time slot n according to the following formula k [n]And the total energy consumption E of the unmanned aerial vehicle in each time slot n U [n]:
(3.5) constructing an optimal task completion Rate mu max User and unmanned aerial vehicle weighted energy consumption E min Is represented by the expression:
wherein the optimization variable is a user offloading decision α= { α U,k,m [n]User task completion rate μ= { μ } k Local calculation CPU frequency F= { F of } and user and unmanned plane k [n],f U,k [n]Transmit power p= { P of } user and unmanned plane k [n],p U,k,m [n]Unmanned aerial vehicle trajectory q= { Q [ n }]};ω U And omega E Respectively representing the energy consumption weight and the total energy consumption weight of the unmanned aerial vehicle;
setting the following constraint conditions of a user and the unmanned aerial vehicle:
representing task completion constraints; />Representing information causal constraints, wherein +.>Wherein (1)>0≤μ k ≤1,0≤f k [n]≤f k,max ,/>0≤p k [n]≤p k,max ,p k [N-1]=p k [N]=0,0≤f U,k [n]≤f U,max ,/>f U,k [1]=f U,k [N]=0,0≤p U,k,m [n]≤p U,max ,/>p U,k,m [1]=p U,k,m [N]=0,0≤α U,k,m [n]≤1,α U,k,m [1]=α U,k,m [N]=0,q[0]=q I ,q[N]=q F Wherein f k,max ,f U,max ,p k,max And p U,max The maximum available CPU frequency and maximum transmit power for the kth user and drone respectively,q[n]-q[n-1]||≤τV max ,/>representing a maximum speed constraint of the unmanned aerial vehicle;
(3.6) maximizing the task completion rate and minimizing the weighted energy consumption of the unmanned aerial vehicle and the user, and obtaining the task completion rate mu through an alternate optimization algorithm max Weighted energy consumption E of user and unmanned aerial vehicle min The corresponding optimal optimization variables;
(4) Optimal task completion Rate μ according to step (3.6) max Weighted energy consumption E of user and unmanned aerial vehicle min And setting working parameters of the system by the corresponding optimal optimization variables, and enabling the system to run under the parameters to realize the optimization of the system task completion rate and the energy consumption.
Compared with the prior art, the invention has the following advantages:
firstly, because the unmanned aerial vehicle is reliably connected with the user and the unmanned aerial vehicle and the base station, the unmanned aerial vehicle not only can be used as a mobile edge computing server, but also can provide communication service for the user by forwarding a received computing task to the base station for remote computing, thereby improving the stability of the system in a computing task high-speed environment and reducing the task processing time delay of the unmanned aerial vehicle.
Secondly, in a real scene, a network system is often a multi-ground base station scene, and most of existing unmanned aerial vehicle auxiliary mobile edge computing systems consider a single-base station scene, so that the actual problem is difficult to solve; compared with the prior art, the method has the advantages that on the basis that the unmanned aerial vehicle can forward the calculation tasks to the base stations for remote calculation, the scene of the multi-ground base stations is considered, and under the fixed time delay requirement, the calculation tasks are distributed to different base stations to complete more tasks, and the minimum system energy consumption is realized.
Drawings
FIG. 1 is a schematic view of an application scenario of the method of the present invention;
FIG. 2 is a schematic diagram of a slot structure according to the present invention;
FIG. 3 is a flow chart of an implementation of the method of the present invention;
FIG. 4 is a graph of unmanned aerial vehicle trajectories corresponding to different task completion times in the method of the present invention;
FIG. 5 is a graph of unmanned aerial vehicle speed corresponding to different task completion times in the method of the present invention;
FIG. 6 is a diagram of simulation results of the influence of the number of tasks and the number of base stations on the weighted sum of the task completion rate and the energy consumption in the method of the invention;
FIG. 7 is a diagram of simulation results of the influence of the number of tasks and the number of base stations on the task completion rate and the energy consumption, respectively, in the method of the present invention;
FIG. 8 is a simulation result diagram of the influence of the weight of the unmanned aerial vehicle and the system bandwidth on the weighted sum of the task completion rate and the energy consumption in the method of the invention;
FIG. 9 is a simulation result diagram of the influence of the weight of the unmanned aerial vehicle and the system bandwidth on the task completion rate and the energy consumption respectively in the method of the invention;
FIG. 10 is a graph showing the influence of the weight of the unmanned aerial vehicle and the weight of the energy consumption on the weighted sum of the task completion rate and the energy consumption in the method of the invention;
FIG. 11 is a simulation result diagram of the influence of the unmanned aerial vehicle weight and the energy consumption weight on the task completion rate and the energy consumption respectively in the method of the invention;
Detailed Description
The following describes the implementation process of the technical scheme in detail with reference to the accompanying drawings:
embodiment one: referring to fig. 3, the invention provides an energy efficiency optimization method in a multi-user cognitive edge computing network, which comprises the following specific implementation steps:
step 1: referring to fig. 1, an edge computing network constructed by the present invention is composed of one unmanned plane, K users, and M base stations on which mobile edge computing servers are mounted; order theAnd->Respectively represent a set of users and base stations, and k and m represent a kth user and an mth base station on which a mobile edge calculation server is mounted, respectively.
Step 2: dividing the time slot structures of users and unmanned aerial vehicles:
referring to FIG. 2, the present invention discretizes a limited task completion time T into N equal time slots, toRepresenting a set of N time slots. The duration of each time slot is τ=t/N, where τ is small enough that the position of the drone may be assumed to be unchanged during each time slot. To avoid use in unloadingInterference between users, users employ a time division multiple access protocol, each time slot is further divided into K equal sub-slots, each sub-slot has a duration of δ=t/(NK), and user K offloads its task input data in the kth sub-slot. In order to better distinguish between the offloading signals from the different users, the drone also employs K equal time division multiple access protocols. Because the unmanned aerial vehicle needs to further partially offload the task input data of each user to a plurality of ground base stations carrying the mobile edge computing server, the unmanned aerial vehicle divides each sub-time slot into M small sub-time slots, and the size of each small sub-time slot is determined by an offloading decision variable alpha U,k,m [n]And (5) determining.
Step 3: obtaining an optimal optimization variable under the scene of an unmanned aerial vehicle auxiliary multi-mobile-edge computing server:
(3.1) using a three-dimensional Euclidean coordinate system, the coordinates of which are measured in meters in this example; assuming that the positions of the base station and all users are fixed on the ground at zero altitude, the horizontal positions of the base station m and the user k are respectively b m And w k The method comprises the steps of carrying out a first treatment on the surface of the Assuming that the unmanned aerial vehicle flies at a fixed height H within a task completion time T, wherein H is more than 0; and the flight start point and the flight end point are respectively set as q I And q F The method comprises the steps of carrying out a first treatment on the surface of the Based on the discrete path planning method, the horizontal position q [ n ] of the unmanned aerial vehicle in the nth time slot is obtained]=q[nτ]Wherein q [0 ]]=q I ,q[N]=q F The method comprises the steps of carrying out a first treatment on the surface of the Assuming that the wireless channels between the unmanned aerial vehicle and the base station and between the unmanned aerial vehicle and the user are all dominated by the line-of-sight link, obtaining the channel power gain h between the unmanned aerial vehicle and the user k and between the unmanned aerial vehicle and the base station m in the time slot n U,k [n]And g U,m [n];
(3.2) assuming each user has a defined computing task to be performed, the computing task is composed of tripletsRepresentation, wherein I k Representing the size of the input data of the computing task, C k Representing the computational resources required to input 1-bit data, T k Maximum tolerable delay for user k; to make the maximum tolerance time delay of all users the same, i.e. T k =t; let the task completion rate per user be μ k ;
(3.3) unloading one part of the calculation tasks to the unmanned aerial vehicle for calculation by each user by adopting a time division multiple access protocol, and carrying out local calculation by the other part; the unmanned aerial vehicle collects calculation tasks unloaded by users through optimizing flight trajectories, and also adopts a time division multiple access protocol to unload one part of the collected calculation tasks to a plurality of base stations on the ground for calculation, and the other part of the collected calculation tasks carries out unmanned aerial vehicle local calculation; obtaining the task data volume of the local calculation execution of the kth user in the time slot nAnd energy consumption->Task data amount offloaded to unmanned plane by kth user +.>And transmission energy consumption->The unmanned plane locally calculates the task data quantity of executing the kth user +.>And energy consumption->Unmanned aerial vehicle offloads task data volume of kth user to base station m +.>And transmission energy consumption->Flight energy consumption of unmanned aerial vehicle +.>
The amount of task data performed by the local calculation of the kth user in time slot nAnd energy consumption->Task data amount offloaded to unmanned plane by kth user +.>And transmission energy consumption->The unmanned plane locally calculates the task data quantity of executing the kth user +.>And energy consumption->Unmanned aerial vehicle offloads task data volume of kth user to base station mAnd transmission energy consumption->Flight energy consumption of unmanned aerial vehicle +.>The calculation is respectively carried out according to the following steps:
wherein B represents the bandwidth of user k to the drone and the bandwidth of the drone to the base station, σ 2 Representing noise power, κ, of any node in the system k And kappa (kappa) U Representing the chip capacitance coefficients of user k and unmanned aerial vehicle, C U Representing the computational resources required by the unmanned aerial vehicle to input 1 bit of data, P (|v [ n ])]I) represents the flight power consumption of the rotor unmanned plane, and is specifically as follows:
wherein P is 0 And P H Respectively the leaf type power and the induced power in a hovering state; u (U) tip 、v 0 、d 0 G, s and A are aerodynamically related, respectivelyThe rotor blade tip speed, average rotor induced speed when hovering, fuselage resistance ratio, air density, rotor solidity, and rotor area are expressed.
(3.4) obtaining the total energy consumption E of the scene user k in each time slot n according to the following formula k [n]And the total energy consumption E of the unmanned aerial vehicle in each time slot n U [n]:
(3.5) constructing the optimal task completion Rate μ for the scene max Weighted energy consumption E of user and unmanned aerial vehicle min The expression:
wherein the optimization variable is a user offloading decision α= { α U,k,m [n]User task completion rate μ= { μ } k Local calculation CPU frequency F= { F of } and user and unmanned plane k [n],f U,k [n]Transmit power p= { P of } user and unmanned plane k [n],p U,k,m [n]Unmanned aerial vehicle trajectory q= { Q [ n }]};ω U And omega E Respectively representing the energy consumption weight and the total energy consumption weight of the unmanned aerial vehicle;
setting that the user and the unmanned aerial vehicle in the scene meet the following constraint conditions:
representing task completion constraints; />Representing information causal constraints, wherein +.>Wherein (1)>0≤μ k ≤1,0≤f k [n]≤f k,max ,/>0≤p k [n]≤p k,max ,/>p k [N-1]=p k [N]=0,0≤f U,k [n]≤f U,max ,/>f U,k [1]=f U,k [N]=0,0≤p U,k,m [n]≤p U,max ,p U,k,m [1]=p U,k,m [N]=0,0≤α U,k,m [n]≤1,/> α U,k,m [1]=α U,k,m [N]=0,q[0]=q I ,q[N]=q F Wherein f k,max ,f U,max ,p k,max And p U,max Maximum available CPU frequency and maximum transmit power for kth user and drone, respectively, < >>||q[n]-q[n-1]||≤τV max ,/>Representing a maximum speed constraint of the unmanned aerial vehicle;
(3.6) atUnder the scene, the task completion rate is maximized, the weighted energy consumption of the unmanned aerial vehicle and the user in the scene is minimized, the highly complex problem is decoupled into the task unloading and resource allocation problem and the unmanned aerial vehicle track design problem through an alternate optimization algorithm, and the task completion rate mu is obtained max Weighted energy consumption E of user and unmanned aerial vehicle min The corresponding optimal optimization variables comprise the following steps:
(3.6.1) fixing the unmanned aerial vehicle track, and converting the task unloading and resource allocation problems into convex problems by using variable substitution, namely converting the expression constructed in the step (3.5) into a convex expression by using variable substitution;
(3.6.2) iteratively solving the transformed convex expression using Lagrangian dual decomposition algorithm and sub-gradient algorithm to obtain an optimal user unloading decision α * Optimal user task completion rate μ * Local calculation CPU frequency F of optimal user and unmanned plane * Optimal user and unmanned aerial vehicle's transmit power P * ;
(3.6.3) fixing alpha * 、μ * 、F * And P * Solving the unmanned aerial vehicle track through a continuous convex approximation algorithm to obtain an optimal unmanned aerial vehicle track Q * 。
Step 4: optimal task completion Rate μ according to step (3.6) max Weighted energy consumption E of user and unmanned aerial vehicle min And setting working parameters of the system by the corresponding optimal optimization variables, and enabling the system to run under the parameters to realize the optimization of the system task completion rate and the energy consumption.
Embodiment two: the method for optimizing the task completion rate and the energy consumption in the unmanned aerial vehicle auxiliary multi-mobile-edge computing server network according to the embodiment has the same overall implementation steps as those in the first embodiment, and further describes the unmanned aerial vehicle auxiliary multi-mobile-edge computing server scene:
step a: an edge computing network consisting of one unmanned plane, K users and M base stations carrying mobile edge computing servers is built; order theAnd->Respectively representing a set of users and base stations, k and m respectively representing a kth user and an mth base station on which a mobile edge computing server is mounted;
step b: discretizing a limited task completion time T into N equal time slots to enableRepresenting a set of N time slots. The duration of each time slot is τ=t/N, where τ is small enough that the position of the drone may be assumed to be unchanged during each time slot. To avoid interference between users during offloading, users employ a time division multiple access protocol. Each time slot is further divided into K equal sub-slots, each sub-slot having a duration δ=t/(NK), and user K offloads its task input data in the kth sub-slot. In order to better distinguish between the offloading signals from the different users, the drone also employs K equal time division multiple access protocols. Because the unmanned aerial vehicle needs to further partially offload the task input data of each user to a plurality of ground base stations carrying the mobile edge computing server, the unmanned aerial vehicle divides each sub-time slot into M small sub-time slots, and the size of each small sub-time slot is determined by an offloading decision variable alpha U,k,m [n]Determining;
step c: for ease of illustration, a three-dimensional Euclidean coordinate system is used, the coordinates of which are measured in meters. Assuming that the positions of the base station and all users are fixed on the ground at zero altitude, the horizontal positions of the base station m and the user k are respectively b m =(x m ,y m ) And w k =(x k ,y k ) The method comprises the steps of carrying out a first treatment on the surface of the Suppose that the unmanned aerial vehicle flies at a fixed height H (H > 0) within a mission completion time T, and the flight start point and the flight end point are set to q, respectively I =(x I ,y I ) And q F =(x F ,y F ) The method comprises the steps of carrying out a first treatment on the surface of the Based on discrete path planning method, the horizontal position of the unmanned aerial vehicle in the nth time slot is q [ n ]]=q[nτ]=(x[n],y[n]) Wherein q [0 ]]=q I ,q[N]=q F The method comprises the steps of carrying out a first treatment on the surface of the Suppose unmanned aerial vehicleThe wireless channel between the base station and the unmanned aerial vehicle and the user is led by the line-of-sight link, so that the channel power gains between the unmanned aerial vehicle and the user k and between the unmanned aerial vehicle and the base station m in the time slot n are respectively h U,k [n]And g U,m [n]:
Wherein ρ is the channel power gain when the reference distance is 1m, d k [n]And d m [n]Respectively representing the distance between the drone and the user k and between the drone and the base station m.
Step d: assuming each user has a defined computing task to be performed, the computing task is composed of triplesRepresentation, wherein I k Representing the size of the input data of the computing task, C k Representing the computational resources required to input 1-bit data, T k For the maximum tolerable delay for user k, where T k T is not more than. Assuming that all users have the same maximum tolerated time delay, i.e. T k =t. Let the task completion rate per user be μ k ;
Step e: each user adopts a time division multiple access protocol to unload one part of the calculation tasks to the unmanned aerial vehicle for calculation, and the other part of the calculation tasks carries out local calculation of the users; the unmanned aerial vehicle collects calculation tasks unloaded by users through optimizing flight trajectories, and also adopts a time division multiple access protocol to unload one part of the collected calculation tasks to a plurality of base stations on the ground for calculation, and the other part of the collected calculation tasks are locally calculated by the unmanned aerial vehicle.
Order theAnd->The task data quantity and the energy consumption which are respectively executed by the local calculation of the kth user in the time slot n are respectively represented by the following specific expression:
wherein f k [n]Representing the CPU clock frequency, κ, of user k in slot n k The chip capacitance coefficient for user k.
Order theAnd->The task data amount and the transmission energy consumption of the kth user unloaded to the unmanned plane in the time slot n are represented as follows:
wherein p is k [n]Representing the transmission power of user k in slot n, B represents the bandwidth of user k to the drone, σ 2 Representing the noise power at the drone. Without loss of generality, it is assumed that the user k to drone bandwidth is the same as the drone to base station bandwidth, and that the noise power of any node in the system is considered to be σ 2 The same applies.
Order theAnd->The method represents the task data quantity and the energy consumption of the kth user executed by the unmanned aerial vehicle through local calculation, and the specific expression is as follows:
wherein f U,k [n]Representing CPU clock frequency allocated by unmanned aerial vehicle to user k in time slot n, C U Representing the computational resources, κ, required for a drone to input 1 bit of data U Is the chip capacitance coefficient of the unmanned plane.
Order theAnd->The method indicates that the unmanned aerial vehicle unloads the task data volume and the transmission energy consumption of the kth user to the base station m, and specifically comprises the following steps:
wherein p is U,k,m [n]And the unmanned aerial vehicle is used for unloading the transmission power of the user k task to the base station m in the time slot n.
The specific expression of the flight energy consumption of the unmanned aerial vehicle is as follows:
wherein, P (||v [ n ] |) represents the flight power consumption of the rotor unmanned plane, and the method is as follows:
wherein P is 0 And P H Respectively, leaf power and induced power in hover state, U tip 、v 0 、d 0 G, s and A are aerodynamically related to rotor blade tip speed, average rotor induced speed when hovering, fuselage resistance ratio, air density, rotor solidity, rotor area, respectively.
Step f: obtaining the total energy consumption E of user k in each time slot n according to k [n]And the total energy consumption E of the unmanned aerial vehicle in each time slot n U [n]:
Step g: the method comprises the following steps of establishing an optimization problem of an unmanned aerial vehicle auxiliary multi-mobile-edge computing server scene, wherein the expression is as follows:
wherein, the optimization variable is alpha= { alpha U,k,m [n]},μ={μ k },F={f k [n],f U,k [n]},P={p k [n],p U,k,m [n]},Q={q[n]};ω U And omega E Respectively representing the energy consumption weight and the total energy consumption weight of the unmanned aerial vehicle; task completion constraints areAnd->Information causal constraint is +.>In (1) the->Wherein (1)>Task completion rate constraint is 0.ltoreq.mu. k Is less than or equal to 1; the calculation task constraint of the user is 0.ltoreq.f k [n]≤f k,max ,/>0≤p k [n]≤p k,max ,p k [N-1]=p k [N]=0, wherein>f k,max And p k,max The maximum available CPU frequency and the maximum transmission power of the kth user are respectively; the calculation task constraint of the unmanned aerial vehicle is 0-f U,k [n]≤f U,max ,/>f U,k [1]=f U,k [N]=0,0≤p U,k,m [n]≤p U,max ,/>p U,k,m [1]=p U,k,m [N]=0, where f U,max And p U,max The maximum available CPU frequency and the maximum transmission power of the unmanned aerial vehicle are respectively; the unloading decision variable constraint of the unmanned aerial vehicle for further unloading to different base stations is 0-alpha U,k,m [n]≤1,/>α U,k,m [1]=α U,k,m [N]=0; the initial and final horizontal position constraints of the unmanned aerial vehicle are q [0 ]]=q I ,q[N]=q F The method comprises the steps of carrying out a first treatment on the surface of the Maximum speed constraint of unmanned aerial vehicle is q [ n ]]-q[n-1]||≤τV max ,/>
Step h: and solving the formula <1.1> by using an alternate iteration, variable replacement, lagrange dual decomposition algorithm, a sub-gradient algorithm and a continuous convex approximation algorithm to obtain the optimal parameters.
(8.1) due to the variable α in the objective function U,k,m [n]And p is as follows U,k,m [n]There is a nonlinear coupling between them, and the variable is also related to the trajectory q [ n ] of the drone]Strong coupling, formula<1.1>Is a complex non-convex expression. To solve this, a two-step alternating optimization algorithm is proposed to solve the expression. The method comprises the steps that firstly, under the condition of a given unmanned aerial vehicle track Q, an unloading decision variable alpha, a task completion rate variable mu and calculation and communication resource scheduling variables F and P in an expression are solved;
(8.1.1) considering that there is coupling between the objective function and the constraint variable, formula<1.1>Is a non-convex expression. To solve it, an auxiliary variable beta is introduced U,k,m [n]So that is beta U,k,m [n]=α U,k,m [n]p U,k,m [n]Thereby eliminating variable coupling, and then the formula<1.1>The method comprises the following steps of:
wherein p= { p k [n]},β={β U,k,m [n]-a }; constraintBecomes intoConstraintBecomes into Constraint 0 is less than or equal to p U,k,m [n]≤p U,max ,/>Becomes 0.ltoreq.beta U,k,m [n]≤α U,k,m [n]p U,max ,/>Constraint p U,k,m [1]=p U,k,m [N]=0 becomes beta U,k,m [1]=β U,k,m [N]=0, wherein%> />
(8.1.2) orderObviously phi U,k,m [i]Zeta is zeta U,k,m [i]Perspective function of (c). Zeta at this time U,k,m [i]As a concave function, then phi U,k,m [i]Also a concave function. Note that use phi U,k,m [i]After approximation, constraintOn the left is a reference to beta U,k,m [i]Is not convex constrained. Thus, the concave function of the left logarithmic term of the inequality is converted to a locally convex approximation using a continuous convex approximation algorithm. Specifically, < +.>Wherein j represents a continuous convex approximationThe number of iterations of the algorithm, the constraint left logarithmic term is expressed approximately as:
wherein,,
thus, equation <1.2> translates to:
(8.1.3) equation <1.3> is a convex expression, which is solved by using the Lagrange dual method. The Lagrangian function of equation <1.3> is:
wherein the constraint is 0.ltoreq.mu. k ≤1,0≤f k [n]≤f k,max ,0≤p k [n]≤p k,max ,/>p k [N-1]=p k [N]=0,0≤f U,k [n]≤f U,max ,/>f U,k [1]=f U,k [N]=0,0≤β U,k,m [n]≤α U,k,m [n]p U,max ,/>β U,k,m [1]=β U,k,m [N]=0,0≤α U,k,m [n]≤1,α U,k,m [1]=α U,k,m [N]=0。
(8.1.4) decomposing the formula <1.4> into a plurality of sub-formulas by a dual decomposition method, specifically as follows:
for the variable f k [n]The corresponding sub-formula is:
wherein the constraint is 0.ltoreq.f k [n]≤f k,max ,Formula (VI)<1.4.1>For a convex expression, the optimal solution can be obtained using the KKT condition:
for the variable f U,k [n]The corresponding sub-formula is:
wherein the constraint is f U,k [1]=f U,k [N]=0,0≤β U,k,m [n]≤α U,k,m [n]p U,max ,Formula (VI)<1.4.2>For a convex expression, the optimal solution can be obtained using the KKT condition: />
For the variable p k [n]The corresponding sub-formula is:
wherein the constraint is 0.ltoreq.p k [n]≤p k,max ,p k [N-1]=p k [N]=0. Formula (VI)<1.4.3>For a convex expression, the optimal solution can be obtained using the KKT condition:
for the variable alpha U,k,m [n]And beta U,k,m [n]The corresponding sub-formula is:
wherein the constraint is 0.ltoreq.beta U,k,m [n]≤α U,k,m [n]p U,max ,β U,k,m [1]=β U,k,m [N]=0,0≤α U,k,m [n]≤1,/>α U,k,m [1]=α U,k,m [N]=0. Formula (VI)<1.4.4>It is difficult to give a closed expression, which can be solved by a block iterative algorithm, and a is initialized first U,k,m [n]Is a feasible solution, then the formula<1.4.4>The process is as follows:
wherein the constraint is 0.ltoreq.beta U,k,m [n]≤α U,k,m [n]p U,max ,β U,k,m [1]=β U,k,m [N]=0. Formula (VI)<1.4.4.1>For a convex expression, the optimal solution can be obtained using the KKT condition:
wherein the constraint is 0.ltoreq.alpha U,k,m [n]≤1,α U,k,m [1]=α U,k,m [N]=0. Formula (VI)<1.4.4.2>It is still difficult to give a closed form solution, using standardsSolution of alpha by convex optimization tool U,k,m [n]Is a solution to the optimization of (3).
Acquisition ofAfter that, let->Returning to solving problem formula<1.4.4.1>The steps are sequentially circulated until convergence can obtain the final optimal solution +.>And->
(8.1.5) solving the dual problem of the requirement solution formula <1.4>, wherein the dual problem is as follows:
wherein the constraint isSince the Salter condition is satisfied, the formula<1.5>Optimum value and formula<1.4>The optimum values are the same. Therefore, the secondary gradient method can be adopted to solve the dual problem, optimize the dual variable and aim at the dual variable omicron k ,/>η k The secondary gradients are respectively:
(8.2) a second step, focusing on designing the unmanned trajectory Q with the optimized variables α, μ, F and P.
(8.2.1) by fixing previously optimized offloading decisions, task completion rates, and computing and communication resources. Thus, the unmanned trajectory design problem can be expressed as follows:
(8.2.2) obviously, formula<1.5>Is a non-convex expression. For this purpose, the formula is applied to the equation by using a continuous convex approximation algorithm<1.5>The non-convex terms in (a) are optimized to obtain the approximate expression. First, for the objective functionNon-convex term P (||v [ n ])]||) introduces variable v 1 [n]And v 2 [n]The method comprises the following steps:
v 1 [n]≥||v[n]||
thus, it is possible to obtain:
using a continuous convex approximation algorithm, give v 1 [n],v 2 [n]Arbitrary feasible solutions of (3)And->Formula (VI)<1.6>Approximated with the following convex constraint:
wherein,,
thus, the non-convex term P (||v [ n ] |) is replaced by the following convex approximation
(8.2.3) in the formula<1.5>In the constraint of (a), ζ 1 [n]And xi 2 [n]For q [ n ]]Is not convex, but is opposite to q n]-b m || 2 And q n]-w k || 2 The whole is a convex function. Based on this, q [ n ] is given using a continuous convex approximation algorithm]Any feasible point q of (2) (j) [n]Has the following components
however, constraintZeta on the left 1 [n]By means ofAfter approximation, it is still a function of q [ n ]]Is not convex constrained. Thus let p U,k,m [n]g U,m [n]|/σ 2 ≥1/Q[n]The first-order Taylor expansion is carried out on the material
(8.2.4) based on a continuous convex approximation algorithm, the suboptimal solution of the original problem may be obtained by solving the following convex approximation expression:
wherein the constraint is q[0]=q I ,q[N]=q F ,||q[n]-q[n-1]||≤τV max ,/>v 1 [n]≥||v[n]||,/> In the method, in the process of the invention,wherein (1)>
Equation <1.7> is a standard convex expression. However, the positions of the drones in the different time slots are shown coupled to each other, so it is difficult to get a closed-form solution for q [ n ]. In this case, the approximation of equation <1.7> is solved by resorting to a standard convex optimization tool.
(8.3) repeating steps (8.1) - (8.2) until the algorithm converges to obtain the optimal optimization variable, i.e., the optimal user offloading decision α * Optimal user task completion rate μ * Local calculation CPU frequency F of optimal user and unmanned plane * Optimal user and unmanned aerial vehicle's transmit power P * And an optimal unmanned aerial vehicle trajectory Q * 。
Step i: and the system selects working parameters according to the optimal optimization variables so as to ensure that the system performance is optimal.
The invention solves the problems of serious channel fading in the traditional mobile edge computing network and limited computing resources of a single mobile edge computing server. Firstly, reliable channel connection with ground users is established by utilizing the advantages of easiness in deployment, flexible movement, line-of-sight link connection and the like of the unmanned aerial vehicle; and secondly, the multi-mobile edge computing server is utilized to promote network computing resources so as to realize stability under a high-transmission environment of computing tasks, and meanwhile, the energy consumption of the unmanned aerial vehicle and a user is reduced, so that the system can be used for an edge computing system to improve the task completion rate.
The effects of the present invention are further described below in conjunction with simulation experiments:
A. simulation conditions
The simulation is carried out by using computer simulation software, the flight start point and the flight end point of the unmanned aerial vehicle are respectively (0, 0) and (40, 50), and three scenes that the number of base stations is respectively 1, 2 and 3 are considered. Setting the base station position as (0, 0) when the number of the base stations is 1; setting the base station positions as (0, 0) and (40, 50) when the number of the base stations is 2; setting the base station positions as (0, 0), (20, 25) and (40, 50) when the number of base stations is 3; the simulation will analyze the impact of key parameters, including the user's computational effort I k The number M of base stations, the system bandwidth B, the task completion time T and the energy consumption weight omega of the unmanned aerial vehicle U And the total energy consumption weight omega E . The basic simulation parameters are listed in table 1, unless otherwise indicated.
Table 1 simulation parameters
B. Emulation content
Simulation 1: unmanned aerial vehicle tracks corresponding to different task completion times, and simulation results are shown in fig. 4;
simulation 2: the unmanned plane speeds corresponding to different task completion times are shown in fig. 5 as simulation results;
simulation 3: the influence of the task quantity and the number of the base stations on the weighted sum of the task completion rate and the energy consumption, and the simulation result is shown in fig. 6;
simulation 4: the influence of the task quantity and the number of the base stations on the task completion rate and the energy consumption respectively, and the simulation result is shown in fig. 7;
simulation 5: the influence of the weight of the unmanned aerial vehicle and the system bandwidth on the weighted sum of the task completion rate and the energy consumption, and the simulation result is shown in fig. 8;
simulation 6: the influence of the weight of the unmanned aerial vehicle and the system bandwidth on the task completion rate and the energy consumption respectively, and the simulation result is shown in fig. 9;
simulation 7: the influence of the unmanned plane weight and the energy consumption weight on the weighted sum of the task completion rate and the energy consumption, and the simulation result is shown in fig. 10;
simulation 8: the influence of the unmanned plane weight and the energy consumption weight on the task completion rate and the energy consumption respectively, and the simulation result is shown in fig. 11;
C. simulation results
As can be seen from fig. 4, as the task completion time T increases, the drone may take advantage of its mobility to more approach the user's location. This is because unmanned aerial vehicles fly closer to the user can reduce path loss. It can also be observed in connection with fig. 5 that for longer task completion times (e.g., t=7.5 s and t=10.5 s), the drone trajectory will tend to stabilize, fly at maximum speed, then slow down, and even tend to hover over a fixed point that can optimally cause the user to perform task offload transmissions, thereby reducing the user's computational power consumption. Further, it can be seen in fig. 4, 11 and 12, that as the drone gets closer to the base station, the speed begins to drop or even hover. This is because the drone can reduce its own calculation and offloading energy consumption by being close to the base station.
As can be seen from fig. 6, as the user's task volume increases, the weighted sum of task completion rate and energy consumption also increases. In addition, as the number of the base stations increases, the weighted sum of the task completion rate and the energy consumption is reduced, and as the task amount increases, the effect is more remarkable.
As can be seen from fig. 7, when the user's calculation task amount is small (e.g., I k =12 Mbit), in the case of three base stations, the energy consumption sum of the drone and the user is instead increased. This is because the drone needs to offload user's task input data to more base stations, which increases the communication energy consumption of the drone. However, when the task amount is large, the unmanned aerial vehicle unloads task input data of users to more base stations to reduce calculation energy consumption of the unmanned aerial vehicle. Therefore, the unmanned aerial vehicle needs to calculate the two factors of energy consumption and communication energy consumption on its ownBalance points are found between the elements. In addition, compared with a single base station scene, the multi-base station scene can obviously improve the completion rate of users.
As can be seen from fig. 8, the weighted sum of the task completion rate and the energy consumption increases with increasing weight of the drone and decreases with increasing system bandwidth. As can be seen from fig. 9, the unmanned aerial vehicle weight increases and the energy consumption of the unmanned aerial vehicle and the user decreases continuously. This is because unmanned aerial vehicle energy consumption is mainly flight energy consumption, and the increase of unmanned aerial vehicle weight makes the optimization objective function more focus on unmanned aerial vehicle's flight energy consumption. Thus, the drone may reduce the flight and hover near the user, which also causes the task completion rate to decrease as the drone weight increases. In addition, the system bandwidth is increased, a user can unload more task input data for the unmanned aerial vehicle, and the calculation and communication energy consumption of the unmanned aerial vehicle is increased. However, as the system bandwidth increases, the unmanned aerial vehicle can offload more data to the ground base station, so that the task completion rate is improved.
As can be seen from fig. 10, as the weight of the unmanned aerial vehicle and the weight of the energy consumption increase, the task completion rate and the weight sum of the energy consumption increase. As can be seen from fig. 11, the increase of the energy consumption weight makes the optimization objective function pay attention to the unmanned energy consumption and also pay more attention to the energy consumption of the user. Because the bandwidth is increased, the user and the unmanned aerial vehicle have more decision space in terms of power, and the unmanned aerial vehicle reduces the flight energy consumption, and simultaneously, the user also can reduce own calculation and communication energy consumption, so the task completion rate is reduced along with the increase of the energy consumption weight. Thus, the objective function needs to find a balance point between two factors, the energy consumption weight and the task completion rate.
The simulation analysis proves the correctness and effectiveness of the method provided by the invention.
The non-detailed description of the invention is within the knowledge of a person skilled in the art.
The foregoing description of the preferred embodiment of the invention is not intended to be limiting, but it will be apparent to those skilled in the art that various modifications and changes in form and detail may be made without departing from the principles and construction of the invention, but these modifications and changes based on the idea of the invention are still within the scope of the appended claims.
Claims (3)
1. The unmanned aerial vehicle assisted multi-MEC server task completion rate maximization method is characterized by comprising the following steps of:
(1) Building an edge computing network model:
an edge computing network model consisting of one unmanned plane, K users and M base stations carrying mobile edge computing MEC servers is built; order theAnd->Respectively representing a set of users and base stations, wherein k and m respectively represent a kth user and an mth base station on which a mobile edge computing server is mounted;
(2) Dividing the time slot structures of users and unmanned aerial vehicles:
discretizing a limited task completion time T into N equal time slots to enableRepresenting a set of N time slots; the duration of each time slot is τ=t/N, assuming that the position of the drone is unchanged during each time slot; a user adopts a time division multiple access protocol to divide each time slot into K 'equal sub-time slots, K' =K, the duration of each sub-time slot is delta=T/(NK '), and the user K unloads task input data of the user K in the K' sub-time slot corresponding to the user K; the unmanned aerial vehicle also adopts K' time division multiple access protocols with equal time division, each sub-time slot is divided into M small sub-time slots, and the size of each small sub-time slot is determined by an unloading decision variable alpha U,k,m [n]Determining;
(3) Obtaining an optimal optimization variable under the scene of an unmanned aerial vehicle auxiliary multi-mobile-edge computing server:
(3.1) Using a three-dimensional Euclidean coordinate System, assume that the locations of the base station and all users are fixed at zeroOn the ground at altitude, the horizontal positions of the base station m and the user k are respectively b m And w k The method comprises the steps of carrying out a first treatment on the surface of the Assuming that the unmanned aerial vehicle flies at a fixed height H within a task completion time T, wherein H is more than 0; and the flight start point and the flight end point are respectively set as q I And q F The method comprises the steps of carrying out a first treatment on the surface of the Based on the discrete path planning method, the horizontal position q [ n ] of the unmanned aerial vehicle in the nth time slot is obtained]=q[nτ]Wherein q [0 ]]=q I ,q[N]=q F The method comprises the steps of carrying out a first treatment on the surface of the Assuming that the wireless channels between the unmanned aerial vehicle and the base station and between the unmanned aerial vehicle and the user are all dominated by the line-of-sight link, obtaining the channel power gain h between the unmanned aerial vehicle and the user k and between the unmanned aerial vehicle and the base station m in the time slot n U,k [n]And g U,m [n];
(3.2) assuming each user has a defined computing task to be performed, the computing task is composed of tripletsRepresentation, wherein I k Representing the size of the input data of the computing task, C k Representing the computational resources required to input 1-bit data, T k Maximum tolerable delay for user k; to make the maximum tolerance time delay of all users the same, i.e. T k =t; let the task completion rate per user be μ k ;
(3.3) unloading one part of the calculation tasks to the unmanned aerial vehicle for calculation by each user by adopting a time division multiple access protocol, and carrying out local calculation by the other part; the unmanned aerial vehicle collects calculation tasks unloaded by users through optimizing flight trajectories, and also adopts a time division multiple access protocol to unload one part of the collected calculation tasks to a plurality of base stations on the ground for calculation, and the other part of the collected calculation tasks carries out unmanned aerial vehicle local calculation; obtaining the task data volume of the local calculation execution of the kth user in the time slot nAnd energy consumptionTask data amount offloaded to unmanned plane by kth user +.>And transmission energy consumption->The unmanned plane locally calculates the task data quantity of executing the kth user +.>And energy consumption->Unmanned aerial vehicle offloads task data volume of kth user to base station m +.>And transmission energy consumption->Flight energy consumption of unmanned aerial vehicle +.>
(3.4) obtaining the total energy consumption E of user k in each time slot n according to the following formula k [n]And the total energy consumption E of the unmanned aerial vehicle in each time slot n U [n]:
(3.5) constructing an optimal task completion Rate mu max User and unmanned aerial vehicle weighted energy consumption E min Is represented by the expression:
wherein the optimization variable is a user offloading decision α= { α U,k,m [n]User task completion rate μ= { μ } k Local calculation CPU frequency F= { F of } and user and unmanned plane k [n],f U,k [n]Transmit power p= { P of } user and unmanned plane k [n],p U,k,m [n]Unmanned aerial vehicle trajectory q= { Q [ n }]};ω U And omega E Respectively representing the energy consumption weight and the total energy consumption weight of the unmanned aerial vehicle;
setting the following constraint conditions of a user and the unmanned aerial vehicle:
representing task completion constraints; />Representing information causal constraints, wherein +.>Wherein,,0≤μ k ≤1,0≤f k [n]≤f k,max ,/>0≤p k [n]≤p k,max ,/>p k [N-1]=p k [N]=0,0≤f U,k [n]≤f U,max ,/>f U,k [1]=f U,k [N]=0,0≤p U,k,m [n]≤p U,max ,p U,k,m [1]=p U,k,m [N]=0,0≤α U,k,m [n]≤1,/> α U,k,m [1]=α U,k,m [N]=0,q[0]=q I ,q[N]=q F wherein f k,max ,f U,max ,p k,max And p U,max Maximum available CPU frequency and maximum transmit power for kth user and drone, respectively, < >>||q[n]-q[n-1]||≤τV max ,/>Representing a maximum speed constraint of the unmanned aerial vehicle;
(3.6) maximizing the task completion rate and minimizing the weighted energy consumption of the unmanned aerial vehicle and the user, and obtaining the task completion rate mu through an alternate optimization algorithm max Weighted energy consumption E of user and unmanned aerial vehicle min The corresponding optimal optimization variables;
(4) Optimal task completion Rate μ according to step (3.6) max Weighted energy consumption E of user and unmanned aerial vehicle min And setting working parameters of the system by the corresponding optimal optimization variables, and enabling the system to run under the parameters to realize the optimization of the system task completion rate and the energy consumption.
2. The method according to claim 1, characterized in that: step (3)3) the amount of task data performed by the local calculation of the kth user in time slot nAnd energy consumption->Task data volume offloaded to unmanned aerial vehicle by kth userAnd transmission energy consumption->The unmanned plane locally calculates the task data quantity of executing the kth user +.>And energy consumptionUnmanned aerial vehicle offloads task data volume of kth user to base station m +.>And transmission energy consumption->Flight energy consumption of unmanned aerial vehicle +.>Respectively according to the following calculation
Wherein B represents the bandwidth of user k to the drone and the bandwidth of the drone to the base station, σ 2 Representing noise power, κ, of any node in the system k And kappa (kappa) U Representing the chip capacitance coefficients of user k and unmanned aerial vehicle, C U Representing the computational resources required by the unmanned aerial vehicle to input 1 bit of data, P (|v [ n ])]I) represents the flight power consumption of the rotor unmanned plane, and is specifically as follows:
wherein P is 0 And P H Respectively, leaf power and induced power in hover state, U tip 、v 0 、d 0 G, s and A represent rotor blade tip speed, average rotor induced speed when hovering, fuselage resistance ratio, air density, rotor solidity and rotor area, respectively.
3. The method according to claim 1, characterized in that: the task completion rate mu is obtained through an alternate optimization algorithm in the step (3.6) max Weighted energy consumption E of user and unmanned aerial vehicle min The corresponding optimal optimization variables are realized as follows:
(3.6.1) fixing the unmanned aerial vehicle track, and converting the expression constructed in the step (3.5) into a convex expression by using variable substitution;
(3.6.2) iteratively solving the transformed convex expression using Lagrangian dual decomposition algorithm and sub-gradient algorithm to obtain an optimal user unloading decision α * Optimal user task completion rate μ * Local calculation CPU frequency F of optimal user and unmanned plane * Optimal user and unmanned aerial vehicle's transmit power P * ;
(3.6.3) fixing alpha * 、μ * 、F * And P * Solving the unmanned aerial vehicle track through a continuous convex approximation algorithm to obtain an optimal unmanned aerial vehicle track Q * 。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310223494.9A CN116366127A (en) | 2023-03-09 | 2023-03-09 | Task completion rate maximization method for unmanned aerial vehicle auxiliary multi-MEC server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310223494.9A CN116366127A (en) | 2023-03-09 | 2023-03-09 | Task completion rate maximization method for unmanned aerial vehicle auxiliary multi-MEC server |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116366127A true CN116366127A (en) | 2023-06-30 |
Family
ID=86911856
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310223494.9A Pending CN116366127A (en) | 2023-03-09 | 2023-03-09 | Task completion rate maximization method for unmanned aerial vehicle auxiliary multi-MEC server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116366127A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117177306A (en) * | 2023-11-03 | 2023-12-05 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle MEC network system based on NFV and SDN |
-
2023
- 2023-03-09 CN CN202310223494.9A patent/CN116366127A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117177306A (en) * | 2023-11-03 | 2023-12-05 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle MEC network system based on NFV and SDN |
CN117177306B (en) * | 2023-11-03 | 2024-01-02 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle MEC network system based on NFV and SDN |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Cao et al. | Mobile edge computing for cellular-connected UAV: Computation offloading and trajectory optimization | |
Hu et al. | Joint offloading and trajectory design for UAV-enabled mobile edge computing systems | |
CN110730031B (en) | Unmanned aerial vehicle track and resource allocation joint optimization method for multi-carrier communication | |
CN114051254B (en) | Green cloud edge collaborative computing unloading method based on star-ground fusion network | |
CN113377533B (en) | Dynamic computing unloading and server deployment method in unmanned aerial vehicle assisted mobile edge computing | |
CN110730495A (en) | Unmanned aerial vehicle data distribution optimization method under energy constraint | |
CN111953407A (en) | Unmanned aerial vehicle video relay system and energy consumption minimizing method thereof | |
Hua et al. | Energy optimization for cellular-connected UAV mobile edge computing systems | |
CN116366127A (en) | Task completion rate maximization method for unmanned aerial vehicle auxiliary multi-MEC server | |
CN114880046B (en) | Low-orbit satellite edge computing and unloading method combining unloading decision and bandwidth allocation | |
CN113627013B (en) | System throughput maximization method based on unmanned aerial vehicle binary unloading edge calculation | |
CN116634544A (en) | MEC energy efficiency maximization method for enabling back-scattering RIS auxiliary unmanned aerial vehicle | |
CN116436512A (en) | Multi-objective optimization method, system and equipment for RIS auxiliary communication | |
Tang et al. | Completed tasks number maximization in UAV-assisted mobile relay communication system | |
CN114866979A (en) | Unmanned aerial vehicle-assisted edge calculation time delay minimization method | |
CN112996121B (en) | U2U distributed dynamic resource allocation method for intra-cluster communication | |
CN115664486B (en) | Energy efficiency optimization method for wireless energy supply in RIS auxiliary UAV edge computing system | |
CN116737391A (en) | Edge computing cooperation method based on mixing strategy in federal mode | |
Cho et al. | IRS-aided energy efficient UAV communication | |
CN114979135B (en) | Unmanned aerial vehicle flight height and task allocation control method based on mobile edge calculation | |
Fu et al. | Joint speed and bandwidth optimized strategy of UAV-assisted data collection in post-disaster areas | |
Zhang et al. | Joint optimization of uav trajectory and relay ratio in uav-aided mobile edge computation network | |
CN114980205A (en) | QoE (quality of experience) maximization method and device for multi-antenna unmanned aerial vehicle video transmission system | |
Song et al. | Personalized Federated Deep Reinforcement Learning-based Trajectory Optimization for Multi-UAV Assisted Edge Computing | |
CN114500533B (en) | Unmanned aerial vehicle auxiliary mobile edge computing system optimization method based on user cooperation |
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 |