Offshore edge computing network task unloading method based on game theory
By applying game theory models in the maritime edge computing system, the task offload of the maritime edge computing network is optimized, and the overload problem of the maritime edge computing system in the case of multiple users is solved, and the balanced allocation of resources and task offload efficiency is improved.
Patent Information
- Application Number
- CN202410983440.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-12
AI Technical Summary
Offshore edge computing systems may cause overload and reduce response speed when facing multiple users simultaneous unloading tasks, and the prior art fails to effectively balance the interests of users and servers.
Using game theory-based method, a task offload model for offshore edge computing network is constructed, and the utility function optimization model between underwater robots, unmanned ships and low-orbit satellites is established. , optimize the task unloading ratio and price plan to ensure balanced resource allocation.
Effectively balance the interests of users and systems, avoid resource waste, improve task unloading efficiency of mobile edge computing systems, and adapt to multi-task unloading status.
Smart Images

Figure CN118939337A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of maritime wireless communication networks, and in particular to a method for offloading maritime edge computing network tasks based on game theory. Background Art
[0002] Existing edge computing-related research is usually focused on land environments, and research on edge computing technology at sea is scarce. It is necessary to develop more efficient system architectures and edge computing solutions. At present, for offshore equipment with limited communication and computing resources, mobile edge computing (MEC) has shown great potential and can effectively solve the problem of limited data collection and computing power of marine equipment. By offloading computing tasks to edge computing servers with powerful computing power, it can significantly improve data processing speed and effectively disperse the computing burden. However, when many users choose to offload tasks at the same time, the edge computing server may face overload challenges, resulting in a significant decrease in response speed, and the interests of users and servers are not considered at the same time. Therefore, studying the offloading strategy of marine users is particularly critical in mobile edge computing. Summary of the invention
[0003] The present invention provides a method for offloading tasks in an offshore edge computing network based on game theory to overcome the above-mentioned technical problems.
[0004] In order to achieve the above object, the technical solution of the present invention is:
[0005] A method for offloading tasks in an offshore edge computing network based on game theory comprises the following steps:
[0006] S1: Based on the underwater robots, unmanned ships and low-orbit satellites that are connected in sequence, a task offloading model and a communication model of the marine edge computing network are constructed to obtain a system model of task offloading of the marine edge computing network;
[0007] S2: Obtain the time / energy consumption calculation model for task offloading based on the system model;
[0008] S3: Obtaining a task efficiency satisfaction function model of the underwater robot according to the time / energy consumption calculation model;
[0009] S4: After the underwater robot task is unloaded to the unmanned ship, the task unloading fee to the unmanned ship and the energy consumption cost of the underwater robot itself need to be paid;
[0010] Based on the task efficiency satisfaction function model of the underwater robot, a utility function of the underwater robot is constructed according to the task offloading cost of the underwater robot and its own energy consumption cost;
[0011] S5: Obtain the energy consumption cost of the unmanned ship, and construct the utility function of the unmanned ship according to the task offloading cost of the underwater robot;
[0012] Obtain the energy consumption cost of the low-orbit satellite itself, and construct the utility function of the low-orbit satellite based on the utility function of the unmanned ship;
[0013] S6: Modeling the optimization problem based on the constructed underwater robot utility function, the unmanned ship utility function and the low-orbit satellite utility function to obtain the optimization target model;
[0014] The optimization target model includes a first utility function optimization model between the lower robot and the unmanned ship, and a second utility function optimization model between the unmanned ship and the low-orbit satellite;
[0015] S7: Establish the Stackelberg game model between underwater robots and unmanned ships, and the bargaining game model between unmanned ships and low-orbit satellites;
[0016] The first utility function optimization model is solved by the Stackelberg game model, and the second utility function optimization model is solved by the bargaining game model to obtain the task unloading and price scheme corresponding to the maximum values of the first utility function optimization model and the second utility function optimization model.
[0017] Further, in S1, a system model for offloading tasks of an offshore edge computing network is obtained, including the following steps:
[0018] S11: define a set of underwater robots I for underwater tasks responsible for monitoring the marine environment and collecting marine data, where I = {1, 2, ..., i};
[0019] A set of unmanned ships M for providing edge computing services for underwater robots, where M = {1, 2, ..., m};
[0020] and a set of low-orbit satellites N for offloading computing tasks that the unmanned ship cannot undertake to a cloud server for processing, where N = {1, 2, ..., n};
[0021] S12: Use the three-dimensional Cartesian coordinate system to express the coordinates of the i-th underwater robot as v i =(x i ,y i ,z i ), the coordinates of the mth unmanned ship are represented as v m =(x m ,y m ,z m ), where x i ,y i ,z iThey represent the horizontal coordinate, vertical coordinate and position coordinate of the ith underwater robot in the three-dimensional Cartesian coordinate system respectively; m ,y m ,z m They respectively represent the horizontal coordinate, vertical coordinate and vertical coordinate of the position of the mth unmanned ship in the three-dimensional Cartesian coordinate system;
[0022] Then the distance d between the underwater robot and the unmanned ship is i,m Expressed as
[0023]
[0024] The distance d between the low-orbit satellite and the unmanned ship m,n Expressed as
[0025]
[0026] Where: R e represents the radius of the earth; h represents the distance from the satellite to the earth; φ m,n represents the geocentric angle between the unmanned ship and the satellite;
[0027] S13: Construct a task offloading model for offshore edge computing networks, specifically
[0028] The network task of underwater robot i is expressed as in Indicates the size of the network task data; Indicates the computing resources required to complete the network task;
[0029] Define the task offloading rate from underwater robot i to unmanned ship m as a i,m , and 0≤a i,m ≤1;
[0030] Then the task volume calculated by underwater robot i is expressed as
[0031] Define the mission offloading rate of the unmanned ship m to the low-orbit satellite as b m,n , and 0≤b m,n ≤1;
[0032] Then the task volume calculated by the unmanned ship m is expressed as
[0033] S14: The underwater robot and the unmanned ship communicate using NOMA, and based on the Shannon theorem according to the distance d between the underwater robot and the unmanned ship i,m With the task offloading model, the communication model between the underwater robot and the unmanned ship is obtained, which is expressed as
[0034]
[0035] Where: R i represents the communication transmission rate between the underwater robot i and the unmanned ship m; W i represents the channel bandwidth; p i Indicates the transmission power; g i represents the channel gain; H(d i,m ,f) represents the attenuation factor of the hydroacoustic signal, which is the function of the distance between the underwater robot i and the unmanned ship m; represents the transmission time of task offloading between underwater robot i and unmanned vessel m; represents the transmission energy consumption of task offloading between underwater robot i and unmanned vessel m;
[0036] S15: The unmanned ship and the low-orbit satellite communicate using FDMA, and based on the Shannon theorem, the distance d between the low-orbit satellite and the unmanned ship is m,n With the task offloading model, the distance d between the low-orbit satellite and the unmanned ship is obtained m,n The communication model is expressed as
[0037]
[0038] Where: R m represents the communication transmission rate between the unmanned ship m and the low-orbit satellite n; W represents the communication bandwidth; p m represents the transmission power of the unmanned ship m; h m,n represents channel gain; N0 represents Gaussian white noise; represents the transmission time of task offloading between the unmanned ship m and the low-orbit satellite n; represents the transmission energy consumption of task offloading between the unmanned ship m and the low-orbit satellite n.
[0039] Furthermore, the time / energy consumption calculation model of the task offloading obtained in S2 includes the time / energy consumption calculation model of the underwater robot, the time / energy consumption calculation model of the unmanned ship, and the time / energy consumption calculation model of the low-orbit satellite;
[0040] The time / energy consumption calculation model of the underwater robot is specifically:
[0041] Assume that i represents the total CPU computing power of underwater robot i, ε i Represents the energy consumption parameter of underwater robot i; then complete the network task λ i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as
[0042] The time / energy consumption calculation model of the unmanned ship is specifically:
[0043] Assume that m represents the total CPU computing amount of the unmanned ship m, ε m Represents the energy consumption parameter of the unmanned ship m, then the task λ is completed i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as
[0044] The time / energy consumption calculation model of the low-orbit satellite is specifically:
[0045] Assume that n represents the total CPU computation of low-orbit satellite n, ε n represents the energy consumption parameter of satellite n, then the task λ is completed i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as
[0046] Furthermore, the S3 specifically includes the following steps:
[0047] S31: Calculate and obtain the time it takes for the underwater robot i to complete the calculation task independently:
[0048] S32: Calculate and obtain the time taken by the underwater robot to complete the computing task after task offloading:
[0049] S33: According to steps S31 and S32, the time saved by the underwater robot to complete the task is obtained as follows: The task efficiency satisfaction function model of underwater robot i is constructed as
[0050] where ξ i Indicates adjustment parameters.
[0051] Furthermore, the S4 specifically includes the following steps:
[0052] S41: Define the network task of unloading the underwater robot i to the unmanned ship m as The task unloading fee to be paid to the unmanned ship m is where k i represents the price per bit of data paid to the unmanned ship;
[0053] And the energy consumption cost of underwater robot i is where q i represents the price of unit energy consumption of underwater robot i itself;
[0054] S42: Based on the task efficiency satisfaction function model of the underwater robot, construct the underwater robot utility function Z according to the task offloading cost of the underwater robot and its own energy consumption cost i , whose expression is
[0055]
[0056] Furthermore, the S5 specifically includes the following steps:
[0057] S51: Define the energy consumption cost of the unmanned ship m as where q m represents the price of unit energy consumption of the unmanned ship m itself;
[0058] And construct the utility function Z of the unmanned ship according to the task offloading cost of the underwater robot m , whose expression is
[0059]
[0060] Where: It means the unmanned ship needs to pay the satellite;
[0061] S52: Define the energy consumption cost of low-orbit satellite n as where q n represents the price of unit energy consumption of low-orbit satellite n itself;
[0062] And according to the fee that the unmanned ship has to pay to the satellite Construct the utility function of the low-orbit satellite, which is expressed as
[0063]
[0064] Furthermore, the optimization target model obtained in S6 is expressed as
[0065] P1:
[0066] P2:
[0067] P3:
[0068] Among them, Formula P1 and Formula P2 are the first utility function optimization model between the underwater robot and the unmanned ship; Formula P2 and Formula P3 are the second utility function optimization model between the unmanned ship and the low-orbit satellite.
[0069] Furthermore, the S7 specifically includes the following steps:
[0070] S71: Solve the first utility function optimization model, and model the utility function optimization problem of the first utility function optimization model as a Stackelberg game model, so that the underwater robot determines the proportion of its network task unloading according to the price given by the unmanned ship, and the unmanned ship adjusts the price to achieve the Stackelberg game equilibrium between the underwater robot and the unmanned ship, and then determines the unloading proportion a i,m The optimal solution of , specifically includes the following three solutions;
[0071] S711: When When , the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio are solved, that is,
[0072]
[0073] Since its second-order derivative is less than 0, let Can get the uninstall ratio a i,m The optimal solution is
[0074] S712: When When , the concavity of the utility function of the underwater robot is determined by solving the first-order derivative and the second-order derivative of the unloading ratio, that is,
[0075]
[0076] in And its second-order derivative is less than 0;
[0077] Then Can get the uninstall ratio a i,m The optimal solution is
[0078]
[0079] S713: When When , the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio are obtained, that is,
[0080]
[0081]
[0082] in
[0083] Since its second-order derivative is less than 0, let The available unloading ratio a i,m The optimal solution
[0084] for
[0085]
[0086] S72: According to the determined uninstall ratio a i,m The optimal solution of , in order to determine the maximum value of the utility function of the underwater robot and the unmanned ship;
[0087] S73: Solve the second utility function optimization model, model the utility function optimization problem of the second utility function optimization model as a bargaining game model, so as to allocate the difference profit of the unmanned ship and the low-orbit satellite according to their respective bids, solve the optimal allocation ratio to achieve the bargaining game equilibrium between the two, and then obtain the transaction price between the unmanned ship and the low-orbit satellite The specific steps include:
[0088] S731: Establish the discount factor model μ of unmanned ship m Discount factor model μ for low-orbit satellites n for
[0089]
[0090] Where: γ m With γ n They are respectively expressed as adjustment parameters of discount factors;
[0091] S732: Based on Rubinstein's bargaining model, according to the discount factor model μ of the unmanned ship m Discount factor model μ for low-orbit satellites n , solve for the profit distribution factor
[0092] S733: Establish the satisfaction function model u of the unmanned ship m Satisfaction function model u of low-orbit satellite n for
[0093] S734: Based on steps S51, S52, and S733, determine the bid price φ of the unmanned ship m (b m,n ) and the bid price of the low-orbit satellite n (b m,n ) are respectively
[0094]
[0095] S735: According to step S732 and step S734, the transaction price between the unmanned ship and the low-orbit satellite is determined Its expression is
[0096]
[0097] Beneficial effects: The present invention provides a method for offloading tasks in an offshore edge computing network based on game theory. The optimization problem is modeled according to the utility function of the constructed underwater robot, the utility function of the unmanned ship, and the utility function of the low-orbit satellite, and the first utility function optimization model between the underwater robot and the unmanned ship, and the second utility function optimization model between the unmanned ship and the low-orbit satellite are obtained; and the first utility function optimization model is solved by the Stackelberg game model, and the second utility function optimization model is solved by the bargaining game model, and the task offloading and price plan corresponding to the maximum value of the first utility function optimization model and the second utility function optimization model are obtained. This method can effectively balance the interests of users and systems, and can complete scheduling in a centralized manner. Compared with traditional scheduling methods, it is more in line with the actual task offloading state, avoids resource waste, and provides a guarantee for the implementation of multi-task offloading in mobile edge computing systems. BRIEF DESCRIPTION OF THE DRAWINGS
[0098] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0099] Figure 1 It is a flow chart of the offshore edge computing network task offloading method based on game theory of the present invention;
[0100] Figure 2 A schematic diagram of a system model for offloading tasks of an offshore edge computing network in this embodiment;
[0101] Figure 3 This is the core block diagram of the offshore edge computing network task offloading based on game theory in this embodiment. DETAILED DESCRIPTION
[0102] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0103] This embodiment provides a method for offloading tasks in an offshore edge computing network based on game theory. Figures 1 to 3 As shown, the following steps are included:
[0104] S1: Based on the underwater robots, unmanned ships and low-orbit satellites that are connected in sequence, a task offloading model and a communication model of the marine edge computing network are constructed to obtain a system model of task offloading of the marine edge computing network;
[0105] The specific steps include:
[0106] S11: define a set of underwater robots I for underwater tasks responsible for monitoring the marine environment and collecting marine data, where I = {1, 2, ..., i};
[0107] A set of unmanned ships M for providing edge computing services for underwater robots, where M = {1, 2, ..., m};
[0108] and a set of low-orbit satellites N for offloading computing tasks that the unmanned ship cannot undertake to a cloud server for processing, where N = {1, 2, ..., n};
[0109] S12: Use the three-dimensional Cartesian coordinate system to express the coordinates of the i-th underwater robot as v i =(x i ,y i ,z i ), the coordinates of the mth unmanned ship are represented as v m =(x m ,y m ,z m ), where x i ,y i ,z i They represent the horizontal coordinate, vertical coordinate and position coordinate of the ith underwater robot in the three-dimensional Cartesian coordinate system respectively; m ,y m ,z m They respectively represent the horizontal coordinate, vertical coordinate and vertical coordinate of the position of the mth unmanned ship in the three-dimensional Cartesian coordinate system;
[0110] Then the distance d between the underwater robot and the unmanned ship is i,m Expressed as
[0111]
[0112] The distance d between the low-orbit satellite and the unmanned ship m,n Expressed as
[0113]
[0114] Where: R e represents the radius of the earth; h represents the distance from the satellite to the earth; φ m,n represents the geocentric angle between the unmanned ship and the satellite;
[0115] S13: Construct a task offloading model for offshore edge computing networks, specifically
[0116] The network task of underwater robot i is expressed as in Indicates the size of the network task data; Indicates the computing resources required to complete the network task;
[0117] Define the task offloading rate from underwater robot i to unmanned ship m as a i,m , and 0≤a i,m ≤1;
[0118] Then the task volume calculated by underwater robot i is expressed as
[0119] Define the mission offloading rate of the unmanned ship m to the low-orbit satellite as b m,n , and 0≤b m,n ≤1;
[0120] Then the task volume calculated by the unmanned ship m is expressed as
[0121] S14: The underwater robot and the unmanned ship communicate using NOMA, and based on the Shannon theorem according to the distance d between the underwater robot and the unmanned ship i,m With the task offloading model, the communication model between the underwater robot and the unmanned ship is obtained, which is expressed as
[0122]
[0123]
[0124] Where: R i represents the communication transmission rate between the underwater robot i and the unmanned ship m; W i represents the channel bandwidth; p i Indicates the transmission power; g i represents the channel gain; H(d i,m ,f) represents the attenuation factor of the hydroacoustic signal, which is the function of the distance between the underwater robot i and the unmanned ship m; represents the transmission time of task offloading between underwater robot i and unmanned vessel m; represents the transmission energy consumption of task offloading between underwater robot i and unmanned vessel m;
[0125] S15: The unmanned ship and the low-orbit satellite communicate using FDMA, and based on the Shannon theorem, the distance d between the low-orbit satellite and the unmanned ship is m,n With the task offloading model, the distance d between the low-orbit satellite and the unmanned ship is obtained m,nThe communication model is expressed as
[0126]
[0127] Where: R m represents the communication transmission rate between the unmanned ship m and the low-orbit satellite n; W represents the communication bandwidth; p m represents the transmission power of the unmanned ship m; h m,n represents channel gain; N0 represents Gaussian white noise; represents the transmission time of task offloading between the unmanned ship m and the low-orbit satellite n; represents the transmission energy consumption of task offloading between the unmanned ship m and the low-orbit satellite n;
[0128] S2: Obtain the time / energy consumption calculation model for task offloading based on the system model;
[0129] Specifically, the time / energy consumption calculation model of task offloading obtained in S2 includes the time / energy consumption calculation model of the underwater robot, the time / energy consumption calculation model of the unmanned ship, and the time / energy consumption calculation model of the low-orbit satellite;
[0130] The time / energy consumption calculation model of the underwater robot is specifically:
[0131] Assume that i represents the total CPU computing power of underwater robot i, ε i Represents the energy consumption parameter of underwater robot i; then complete the network task λ i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as
[0132] The time / energy consumption calculation model of the unmanned ship is specifically:
[0133] Assume that m represents the total CPU computing amount of the unmanned ship m, ε m Represents the energy consumption parameter of the unmanned ship m, then the task λ is completed i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as
[0134] The time / energy consumption calculation model of the low-orbit satellite is specifically:
[0135] Assume that n represents the total CPU computation of low-orbit satellite n, ε n represents the energy consumption parameter of satellite n, then the task λ is completed i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as
[0136] S3: Obtaining a task efficiency satisfaction function model of the underwater robot according to the time / energy consumption calculation model; specifically comprising the following steps:
[0137] S31: Calculate and obtain the time it takes for the underwater robot i to complete the calculation task independently:
[0138] S32: Calculate and obtain the time taken by the underwater robot to complete the computing task after task offloading:
[0139] S33: According to steps S31 and S32, the time saved by the underwater robot to complete the task is obtained as follows: The task efficiency satisfaction function model of underwater robot i is constructed as
[0140] where ξ i Indicates adjustment parameters;
[0141] S4: After the underwater robot task is unloaded to the unmanned ship, the task unloading fee that needs to be paid to the unmanned ship and the underwater robot's own energy consumption cost are obtained, and based on the underwater robot's task efficiency satisfaction function model, the underwater robot utility function is constructed according to the underwater robot's task unloading fee and its own energy consumption cost, which specifically includes the following steps:
[0142] S41: Define the network task of unloading the underwater robot i to the unmanned ship m as The task unloading fee to be paid to the unmanned ship m is where k i represents the price per bit of data paid to the unmanned ship;
[0143] And the energy consumption cost of underwater robot i is where q i represents the price of unit energy consumption of underwater robot i itself;
[0144] S42: Based on the task efficiency satisfaction function model of the underwater robot, construct the underwater robot utility function Z according to the task offloading cost of the underwater robot and its own energy consumption cost i , whose expression is
[0145]
[0146] S5: Obtain the self-energy consumption cost of the unmanned ship, and construct the utility function of the unmanned ship according to the task unloading cost of the underwater robot; obtain the self-energy consumption cost of the low-orbit satellite, and construct the utility function of the low-orbit satellite according to the utility function of the unmanned ship, which specifically includes the following steps:
[0147] S51: Define the energy consumption cost of the unmanned ship m as where q m represents the price of unit energy consumption of the unmanned ship m itself;
[0148] And construct the utility function Z of the unmanned ship according to the task offloading cost of the underwater robot m , whose expression is
[0149]
[0150] Where: It indicates the fee that the unmanned ship needs to pay to the satellite, so that it can be solved through bargaining game later;
[0151] S52 defines the energy consumption cost of low-orbit satellite n as where q n represents the price of unit energy consumption of low-orbit satellite n itself;
[0152] And according to the fee that the unmanned ship has to pay to the satellite Construct the utility function of the low-orbit satellite, which is expressed as
[0153]
[0154] S6: Modeling the optimization problem according to the constructed utility function of the underwater robot, the utility function of the unmanned ship, and the utility function of the low-orbit satellite to obtain an optimization target model; the optimization target model includes a first utility function optimization model between the underwater robot and the unmanned ship, and a second utility function optimization model between the unmanned ship and the low-orbit satellite;
[0155] The optimization target model obtained in S6 is expressed as
[0156] P1:
[0157] P2:
[0158] P3:
[0159] Among them, formula P1 and formula P2 are the first utility function optimization model between the underwater robot and the unmanned ship; formula P2 and formula P3 are the second utility function optimization model between the unmanned ship and the low-orbit satellite;
[0160] S7: Establishing a Stackelberg game model between an underwater robot and an unmanned ship, and a bargaining game model between an unmanned ship and a low-orbit satellite; solving the first utility function optimization model through the Stackelberg game model, solving the second utility function optimization model through the bargaining game model, and obtaining task unloading and price solutions corresponding to the maximum values of the first utility function optimization model and the second utility function optimization model, specifically including the following steps:
[0161] S71: Solve the first utility function optimization model, and model the utility function optimization problem of the first utility function optimization model as a Stackelberg game model, so that the underwater robot determines the proportion of its network task unloading according to the price given by the unmanned ship, and the unmanned ship adjusts the price to achieve the Stackelberg game equilibrium between the underwater robot and the unmanned ship, and then determines the unloading proportion a i,m The optimal solution of , specifically includes the following three solutions;
[0162] S711: When When , the concavity of the utility function of the underwater robot with respect to the unloading ratio is determined by solving the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio (the concavity means that when the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio is less than 0, the first-order derivative of the utility function of the underwater robot with respect to the unloading ratio has an optimal solution), that is,
[0163]
[0164] Since its second-order derivative is less than 0, let Can get the uninstall ratio a i,m The optimal solution is
[0165] S712: When When , the concavity of the utility function of the underwater robot with respect to the unloading ratio is determined by solving the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio (that is, when the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio is less than 0, the first-order derivative of the utility function of the underwater robot with respect to the unloading ratio has an optimal solution), that is,
[0166]
[0167] in
[0168] Since its second-order derivative is less than 0, let Can get the uninstall ratio a i,m The most
[0169] The best solution is
[0170]
[0171] S713: When When , the concavity of the utility function of the underwater robot with respect to the unloading ratio is determined by calculating the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio (that is, when the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio is less than 0, the first-order derivative of the utility function of the underwater robot with respect to the unloading ratio has an optimal solution), that is,
[0172] in
[0173] Since its second-order derivative is less than 0; c represents the speed of light, then let The available unloading ratio a i,m The optimal solution is
[0174]
[0175] S72: According to the determined uninstall ratio a i,m The optimal solution of , in order to determine the maximum value of the utility function of the underwater robot and the unmanned ship;
[0176] S73: Solve the second utility function optimization model, and model the utility function optimization problem of the second utility function optimization model as a bargaining game model, in which the unmanned ship and the low-orbit satellite respectively propose transaction prices. When the bid of the unmanned ship is greater than the asking price of the low-orbit satellite, the game is established. The game goal is to allocate the difference profit between the unmanned ship and the low-orbit satellite based on their respective bids, and solve the optimal allocation ratio to achieve the bargaining game equilibrium between the two, so as to obtain the transaction price between the unmanned ship and the low-orbit satellite. The specific steps include:
[0177] S731: Establish the discount factor model μ of unmanned ship m Discount factor model μ for low-orbit satellites n for
[0178]
[0179] Mainly represents the patience of the unmanned ship and the satellite in waiting, γ m With γ n They are respectively expressed as adjustment parameters of the discount factor; among them, the discount factor of the unmanned ship changes with the time it takes for the satellite to complete its mission The satellite's discount factor decreases with the increase of unloading ratio b, indicating that the longer the satellite takes to complete the mission, the lower the tolerance of the unmanned ship; the satellite's discount factor and unloading ratio b m,n Positive correlation, so the larger the offloading ratio, the greater the profit earned by the satellite and the higher its tolerance;
[0180] S732: Based on Rubinstein's bargaining model, according to the discount factor model μ of the unmanned ship mDiscount factor model μ for low-orbit satellites n , solve for the profit distribution factor
[0181] S733: Establish the satisfaction function model u of the unmanned ship m Satisfaction function model u of low-orbit satellite n for For the unmanned ship, the shorter the time it takes for the low-orbit satellite to complete its unloading task, the higher its satisfaction. For the low-orbit satellite, its satisfaction will increase as the amount of unloaded data increases, indicating that it hopes to obtain higher income;
[0182] S734: Based on steps S51, S52, and S733, determine the bid price φ of the unmanned ship m (b m,n ) and the bid price of the low-orbit satellite φ n (b m,n ) are respectively
[0183]
[0184] The bid of the unmanned ship cannot exceed its total profit, and the bid of the low-orbit satellite cannot be lower than the price it pays to complete the computing task;
[0185] S735: According to step S732 and step S734, the transaction price between the unmanned ship and the low-orbit satellite is determined Its expression is
[0186]
[0187] The present invention proposes a method for offloading marine edge computing tasks based on game theory to provide edge computing services for underwater robots. The present invention solves the optimization problem of the utility function in stages. The utility function optimization problem between the underwater robot and the unmanned ship is solved by the Stackelberg game model. The utility function optimization problem between the unmanned ship and the low-orbit satellite is solved by the bargaining game model. Finally, a computing task offloading strategy and a price scheme for maximizing the utility function of the three parties are obtained.
[0188] By studying the method of offloading the marine edge computing task based on game theory, the system model includes an underwater robot, an unmanned ship and a low-orbit satellite. The underwater robot is responsible for monitoring the marine environment and collecting marine data (for example, marine water quality, marine life, geology, etc.). The unmanned ship is equipped with an edge server, which can provide edge computing services for the underwater robot. The low-orbit satellite is equipped with a powerful server, and the computing tasks that the unmanned ship cannot undertake can be offloaded to the low-orbit satellite for processing. The entire working process of the system includes the following three stages: First, in the first stage, the underwater robot performs local calculations, and at the same time, part of its workload can be offloaded to the corresponding unmanned ship for further calculation. In the second stage, the unmanned ship processes part of the workload through local calculations, and at the same time, the workload that is difficult to load can be transmitted to the low-orbit satellite for further calculation. In the third stage, the satellite is responsible for receiving the data of the unmanned ship through the radio frequency link and performing local calculations. The present invention extends the application of edge computing technology and task offloading technology to marine scenarios, such as submarine environment monitoring, which can effectively solve the problems of large task processing delay and low real-time performance in traditional submarine monitoring networks, and improve the supervision efficiency of the submarine environment.
[0189] Compared with the prior art, it has the following beneficial effects:
[0190] (1) It can effectively solve the problem of limited data collection and computing capabilities of marine equipment;
[0191] (2) Play an important role in submarine water quality monitoring, accident evidence collection, safety and pollution prevention monitoring;
[0192] (3) Provide good technical support for marine environmental safety supervision and search and rescue emergency work.
[0193] (4) It can provide more convenient, efficient and safe service support and technical guarantee for the construction of digital ports and smart ports.
[0194] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for offloading tasks in an offshore edge computing network based on game theory, characterized in that: The following steps are involved: S1: Based on the underwater robots, unmanned ships and low-orbit satellites that are connected in sequence, a task offloading model and a communication model of the marine edge computing network are constructed to obtain a system model of task offloading of the marine edge computing network; S2: Obtain the time / energy consumption calculation model for task offloading based on the system model; S3: Obtaining a task efficiency satisfaction function model of the underwater robot according to the time / energy consumption calculation model; S4: After the underwater robot task is unloaded to the unmanned ship, the task unloading fee to the unmanned ship and the energy consumption cost of the underwater robot itself need to be paid; Based on the task efficiency satisfaction function model of the underwater robot, a utility function of the underwater robot is constructed according to the task offloading cost of the underwater robot and its own energy consumption cost; S5: Obtain the energy consumption cost of the unmanned ship, and construct the utility function of the unmanned ship according to the task offloading cost of the underwater robot; Obtain the energy consumption cost of the low-orbit satellite itself, and construct the utility function of the low-orbit satellite based on the utility function of the unmanned ship; S6: Modeling the optimization problem based on the constructed underwater robot utility function, the unmanned ship utility function and the low-orbit satellite utility function to obtain the optimization target model; The optimization target model includes a first utility function optimization model between the lower robot and the unmanned ship, and a second utility function optimization model between the unmanned ship and the low-orbit satellite; S7: Establish the Stackelberg game model between underwater robots and unmanned ships, and the bargaining game model between unmanned ships and low-orbit satellites; The first utility function optimization model is solved by the Stackelberg game model, and the second utility function optimization model is solved by the bargaining game model to obtain the task offloading strategy and price plan corresponding to the maximum value of the first utility function optimization model and the second utility function optimization model.
2. According to a game theory-based offshore edge computing network task offloading method according to claim 1, it is characterized in that: The system model for offloading the offshore edge computing network task is obtained in S1, including the following steps: S11: define a set of underwater robots I for underwater tasks responsible for monitoring the marine environment and collecting marine data, where I = {1, 2, ..., i}; A set of unmanned ships M for providing edge computing services for underwater robots, where M = {1, 2, ..., m}; and a set of low-orbit satellites N for offloading computing tasks that the unmanned ship cannot undertake to a cloud server for processing, where N = {1, 2, ..., n}; S12: Use the three-dimensional Cartesian coordinate system to express the coordinates of the i-th underwater robot as v i =(x i ,y i ,z i ), the coordinates of the mth unmanned ship are represented as v m =(x m ,y m ,z m ), where x i ,y i ,z i They represent the horizontal coordinate, vertical coordinate and position coordinate of the ith underwater robot in the three-dimensional Cartesian coordinate system respectively; m ,y m ,z m They respectively represent the horizontal coordinate, vertical coordinate and vertical coordinate of the position of the mth unmanned ship in the three-dimensional Cartesian coordinate system; Then the distance d between the underwater robot and the unmanned ship is i,m Expressed as The distance d between the low-orbit satellite and the unmanned ship m,n Expressed as Where: R e represents the radius of the earth; h represents the distance from the satellite to the earth; φ m,n represents the geocentric angle between the unmanned ship and the satellite; S13: Construct a task offloading model for offshore edge computing networks, specifically The network task of underwater robot i is expressed as in Indicates the size of the network task data; Indicates the computing resources required to complete the network task; Define the task offloading rate from underwater robot i to unmanned ship m as a i,m , and 0≤a i,m ≤1; Then the task volume calculated by underwater robot i is expressed as Define the mission offloading rate of the unmanned ship m to the low-orbit satellite as b m,n , and 0≤b m,n ≤1; Then the task volume calculated by the unmanned ship m is expressed as S14: The underwater robot and the unmanned ship communicate using NOMA, and based on the Shannon theorem according to the distance d between the underwater robot and the unmanned ship i,m With the task offloading model, the communication model between the underwater robot and the unmanned ship is obtained, which is expressed as Where: R i represents the communication transmission rate between the underwater robot i and the unmanned ship m; W i represents the channel bandwidth; p i Indicates the transmission power; g i represents the channel gain; H(d i,m ,f) represents the attenuation factor of the hydroacoustic signal, which is the function of the distance between the underwater robot i and the unmanned ship m; represents the transmission time of task offloading between underwater robot i and unmanned vessel m; represents the transmission energy consumption of task offloading between underwater robot i and unmanned vessel m; S15: The unmanned ship and the low-orbit satellite communicate using FDMA, and based on the Shannon theorem, the distance d between the low-orbit satellite and the unmanned ship is m,n With the task offloading model, the distance d between the low-orbit satellite and the unmanned ship is obtained m,n The communication model is expressed as Where: R m represents the communication transmission rate between the unmanned ship m and the low-orbit satellite n; W represents the communication bandwidth; p m represents the transmission power of the unmanned ship m; h m,n represents channel gain; N0 represents Gaussian white noise; represents the transmission time of task offloading between the unmanned ship m and the low-orbit satellite n; represents the transmission energy consumption of task offloading between the unmanned ship m and the low-orbit satellite n.
3. A method for offloading tasks in an offshore edge computing network based on game theory according to claim 2, characterized in that: The time / energy consumption calculation model of task offloading obtained in S2 includes the time / energy consumption calculation model of the underwater robot, the time / energy consumption calculation model of the unmanned ship, and the time / energy consumption calculation model of the low-orbit satellite; The time / energy consumption calculation model of the underwater robot is specifically: Assume that i represents the total CPU computing power of underwater robot i, ε i Represents the energy consumption parameter of underwater robot i; then complete the network task λ i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as The time / energy consumption calculation model of the unmanned ship is specifically: Assume that m represents the total CPU computing amount of the unmanned ship m, ε m Represents the energy consumption parameter of the unmanned ship m, then the task λ is completed i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as The time / energy consumption calculation model of the low-orbit satellite is specifically: Assume that n represents the total CPU computation of low-orbit satellite n, ε n represents the energy consumption parameter of satellite n, then the task λ is completed i The computation time spent on some tasks is expressed as And the corresponding energy consumption of the calculation is expressed as 4. The method for offloading tasks of an offshore edge computing network based on game theory according to claim 3 is characterized in that: The S3 specifically includes the following steps: S31: Calculate and obtain the time it takes for the underwater robot i to complete the calculation task independently: S32: Calculate and obtain the time taken by the underwater robot to complete the computing task after task offloading: S33: According to steps S31 and S32, the time saved by the underwater robot to complete the task is obtained as T i all -T i tot , to construct the task efficiency satisfaction function model of underwater robot i as where ξ i Indicates adjustment parameters.
5. The method for offloading tasks of an offshore edge computing network based on game theory according to claim 4 is characterized in that: The S4 specifically includes the following steps S41: Define the network task of unloading the underwater robot i to the unmanned ship m as The task unloading fee to be paid to the unmanned ship m is where k i represents the price per bit of data paid to the unmanned ship; And the energy consumption cost of underwater robot i is where q i represents the price of unit energy consumption of underwater robot i itself; S42: Based on the task efficiency satisfaction function model of the underwater robot, construct the underwater robot utility function Z according to the task offloading cost of the underwater robot and its own energy consumption cost i , whose expression is 6. A method for offloading tasks in an offshore edge computing network based on game theory according to claim 4, characterized in that: The S5 specifically includes the following steps: S51: Define the energy consumption cost of the unmanned ship m as where q m represents the price of unit energy consumption of the unmanned ship m itself; And construct the utility function Z of the unmanned ship according to the task offloading cost of the underwater robot m , whose expression is Where: It means the unmanned ship needs to pay the satellite; S52: Define the energy consumption cost of low-orbit satellite n as where q n represents the price of unit energy consumption of low-orbit satellite n itself; And according to the fee that the unmanned ship has to pay to the satellite Construct the utility function of the low-orbit satellite, which is expressed as 7. A method for offloading tasks in an offshore edge computing network based on game theory according to claim 6, characterized in that: The optimization target model obtained in S6 is expressed as P1: P2: P3: Among them, Formula P1 and Formula P2 are the first utility function optimization model between the underwater robot and the unmanned ship; Formula P2 and Formula P3 are the second utility function optimization model between the unmanned ship and the low-orbit satellite.
8. The method for offloading tasks in an offshore edge computing network based on game theory according to claim 7, characterized in that: The S7 specifically comprises the following steps: S71: Solve the first utility function optimization model, and model the utility function optimization problem of the first utility function optimization model as a Stackelberg game model, so that the underwater robot determines the proportion of its network task unloading according to the price given by the unmanned ship, and the unmanned ship adjusts the price to achieve the Stackelberg game equilibrium between the underwater robot and the unmanned ship, and then determines the unloading proportion a i,m The optimal solution of , specifically includes the following three solutions; S711: When When , the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio are solved, that is, Since its second-order derivative is less than 0, let Can get the uninstall ratio a i,m The optimal solution is S712: When When , the concavity of the utility function of the underwater robot is determined by solving the first-order derivative and the second-order derivative of the unloading ratio, that is, in And its second-order derivative is less than 0; Then Can get the uninstall ratio a i,m The optimal solution is S713: When When , the first-order derivative and the second-order derivative of the utility function of the underwater robot with respect to the unloading ratio are obtained, that is, in Since its second-order derivative is less than 0, let The available unloading ratio a i,m The optimal solution is S72: According to the determined uninstall ratio a i,m The optimal solution of , in order to determine the maximum value of the utility function of the underwater robot and the unmanned ship; S73: Solve the second utility function optimization model, model the utility function optimization problem of the second utility function optimization model as a bargaining game model, so as to allocate the difference profit of the unmanned ship and the low-orbit satellite according to their respective bids, solve the optimal allocation ratio to achieve the bargaining game equilibrium between the two, and then obtain the transaction price between the unmanned ship and the low-orbit satellite The specific steps include: S731: Establish the discount factor model μ of unmanned ship m Discount factor model μ for low-orbit satellites n for Where: γ m With γ n They are respectively expressed as adjustment parameters of discount factors; S732: Based on Rubinstein's bargaining model, according to the discount factor model μ of the unmanned ship m Discount factor model μ for low-orbit satellites n , solve for the profit distribution factor S733: Establish the satisfaction function model u of the unmanned ship m Satisfaction function model u of low-orbit satellite n for S734: Based on steps S51, S52, and S733, determine the bid price φ of the unmanned ship m (b m,n ) and the bid price of the low-orbit satellite φ n (b m,n ) are respectively S735: According to step S732 and step S734, the transaction price between the unmanned ship and the low-orbit satellite is determined Its expression is
