A method for computing offloading and resource allocation under a relay-assisted MEC architecture

Abstract

The application discloses a kind of computing unloading and resource allocation method under relay auxiliary MEC architecture, belong to wireless communication and computing resource management technical field.The application provides multiple task processing modes by introducing relay device, including local execution, relay unloading and edge computing, in order to minimize execution cost from user angle, joint optimization method of task division strategy, computing resource allocation strategy, relay selection strategy and uplink spectrum allocation strategy is proposed, can reduce network congestion while guaranteeing quality of service, improve user experience;In order to solve joint optimization problem, using block coordinate descent, reconstruction linearization technology and convex optimization method, with fast convergence speed and lower computational complexity, and relay auxiliary MEC architecture and optimization method can adapt to different types of intelligent devices and application scenarios, with wide applicability and flexibility, meet the needs of future mobile edge computing development.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of under the relay auxiliary MEC architecture computing offloading and resource allocation method, belong to wireless communication and computing resource management technical field. BACKGROUND

[0002] The popularity of smart devices (SD) in modern communication and computing fields has led to a sharp increase in demand for computing power and storage capacity. However, due to hardware limitations, a single smart device cannot efficiently handle complex computing tasks. Therefore, mobile edge computing (MEC) has emerged as a key technology to solve this problem, by deploying computing resources at the network edge to help smart devices complete computing tasks.

[0003] The introduction of relay devices in the MEC architecture can effectively make up for the lack of resources in traditional MEC systems. Relay devices not only provide communication services, but also assist in completing computing tasks, thereby reducing the computing burden of smart devices and improving the overall system's computing power and resource utilization efficiency.

[0004] In traditional MEC systems, resource management schemes usually only consider a single computing offloading mode, failing to fully utilize various resources in the network, resulting in low resource utilization, severe network congestion, and affecting the Quality of Service (QoS). Therefore, designing an optimization scheme that can fully utilize multiple resources is a pressing problem. SUMMARY

[0005] To solve the problem of low resource utilization and high execution cost in current mobile edge computing (MEC) systems, the present application provides a computing offloading and resource allocation method under a relay-assisted MEC architecture. The technical solution is as follows:

[0006] The computing offloading and resource allocation method under the relay-assisted MEC architecture of the present application includes:

[0007] Step 1: Construct a relay-assisted MEC system model, define the functions and task processing modes of relay devices, smart devices, and edge servers;

[0008] Step 2: Jointly formulate the task division strategy, computing resource allocation strategy, relay selection strategy, and uplink spectrum allocation strategy in the relay-assisted MEC system model to form a joint optimization problem, with the goal of minimizing system execution cost while ensuring service quality;

[0009] Step 3: Design a preliminary solution method. The block coordinate descent method is adopted to decompose the joint optimization problem into multiple sub-problems. Each sub-problem optimizes the task partitioning strategy, computing resource allocation strategy, relay selection strategy, and uplink spectrum allocation strategy, respectively.

[0010] Step 4: Further optimize the preliminary solution results by using the reconstruction linearization technique to approximate the non-convex problem into a convex optimization problem, and then apply the convex optimization method to solve it.

[0011] Optionally, step 1 includes:

[0012] Using decision matrix Describe the relationship between the i-th smart device and the j-th relay device, where x i,j ∈{0,1} is a binary choice variable, x i,j =1 indicates that the i-th smart device selects the j-th relay device as the relay, while x i,j = 0 indicates that the i-th smart device and the j-th relay device have no matching relationship. According to the general physical meaning of relay selection, the variable x is selected. i,j The restrictions are as follows:

[0013]

[0014] Because the system adopts a partial offloading model, the task is divided into three parts: local computation, relay offloading, and edge offloading.

[0015]

[0016] in, and Describe the number of computation bits allocated to local computation, relay offloading, and edge offloading, respectively, L i This indicates the total amount of data for this computation task;

[0017] For the i-th smart device, latency is divided into three parts: local computation, relay offload, and edge offload. This indicates that, for relay equipment and base stations, the offloading phase is divided into two states: transmission and computation. This indicates that the communication process is divided into two stages, each corresponding to two different transmission delays. The first stage is when the i-th smart device... Within the time slot The unloading data is sent to the selected j-th relay device; the second stage is... The j-th relay device will receive within the time slot Data is forwarded to the MEC server; using h i,j B represents the power gain of the sub-channel. iis the sub-channel bandwidth allocated to the ith smart device by the jth relay, p i,j is the transmit power of the ith smart device, while N i,j is the noise power; if using α i to represent the proportion of the uplink transmission sub-bandwidth to the total bandwidth, then B i = Bα i , α i ∈ {0, 1}, ∑ i α i ≤ 1;

[0018] In order to complete the allocated task within the maximum tolerable delay T, the time limit of the local computation and relay offloading process is:

[0019]

[0020] where C i is the number of CPU cycles required per bit, and represent the CPU frequency of the ith smart device and the jth relay device serving the allocated task, respectively, satisfying and and represent the maximum computing capacity that the smart device and the relay device can allocate to the task, respectively, represents the delay of relay computation;

[0021] The total energy consumption of the ith smart device associated with the jth relay device for local computation and task offloading is represented as:

[0022]

[0023] where, is the power consumption per CPU cycle, κ i is the energy consumption coefficient depending on the chip architecture, E comm represents the energy consumption of task offloading to the relay in the first transmission phase, E comp represents the energy consumption that the SD needs to consume to complete the local computation task.

[0024] Optionally, the step 2 comprises:

[0025] The unit cost of the relay receiving, storing and forwarding data supply to a smart device task is m r , and the supply price of the computing unit bit data is m c ; using M i to represent the service cost of supplying the ith smart device task, when the task is successfully collected by the demanded relay, it gets:

[0026]

[0027] for a given task partitioning policy a computation resource allocation policy a relay selection policy and an uplink spectrum allocation policy The joint optimization problem is formulated as follows using equation (8):

[0028]

[0029] where, is the solution of P1, ω1 represents the unit cost of using computation capability service for the i-th smart device, and ω2 is the weight of offloading service represents the maximum CPU frequency of the i-th smart device for local computation.

[0030] Optionally, the step 3 comprises:

[0031] a given relay selection policy and an uplink spectrum allocation policy jointly optimize a task partitioning policy a computation resource allocation policy Two sub-problems are formulated as follows:

[0032]

[0033] The objective function and the constraints in the sub-problem P2 are both convex functions, and P2 is solved using convex optimization techniques to obtain a local optimal solution at the current iteration number;

[0034] a given task partitioning policy a computation resource allocation policy jointly optimize a relay selection policy and an uplink spectrum allocation policy Two sub-problems are formulated as follows:

[0035]

[0036] where,

[0037] Optionally, the step 4 comprises:

[0038] The discrete variable x i,j is relaxed to 0≤x i,j ≤1 using the slack variable method;

[0039] Let where the value θ is approximately 0;

[0040] For the second order term an auxiliary variable is introduced, where xi,j and respectively, are restricted to 0≤x i,j ≤1 and

[0041] transform P3 into:

[0042]

[0043] Solving P3.1 by using convex optimization technique, a local optimal solution under the current iteration number is obtained, and the solution is substituted into and the rounding method is used to restore the optimization variable x i,j .

[0044] Optionally, the channel access in the two stages from the intelligent device to the relay device and from the relay device to the base station adopts orthogonal frequency division multiplexing.

[0045] Optionally, the convex optimization method used in the step 4 is a CVX solver.

[0046] Optionally, the task processing mode of the step 1 includes local execution, relay offloading and edge computing.

[0047] The present application provides an electronic device, comprising a memory and a processor;

[0048] The memory is used for storing a computer program;

[0049] The processor is used for implementing the method according to any one of the above when executing the computer program.

[0050] The present application provides a computer readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of the above is implemented.

[0051] The present application has the following beneficial effects:

[0052] The application optimizes the calculation offloading and resource allocation strategy, maximally reduces the user's overhead on communication and calculation service, thereby reducing the overall execution cost of edge users. By introducing a relay device and reasonably selecting a task processing mode (local execution, relay offloading, edge calculation), efficient use of resources is achieved, and resource waste is avoided. By jointly optimizing calculation offloading, relay selection, calculation resource allocation and spectrum resource allocation, network congestion can be reduced while ensuring quality of service (QoS), and user experience can be improved. The two-stage iterative algorithm designed in the application uses block coordinate descent, reconstruction linearization technology (RLT) and convex optimization method, which can effectively solve non-convex optimization problems, has fast convergence speed and low calculation complexity. The relay-assisted MEC architecture and optimization method can adapt to different types of intelligent devices and application scenarios, have wide applicability and flexibility, and meet the needs of future mobile edge calculation development. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0054] Figure 1 is the convergence performance curve diagram of the calculation offloading and resource allocation method under the relay-assisted MEC architecture in the application.

[0055] Figure 2 is the relationship diagram between user location distribution and relay selection in the task offloading scenario in the second embodiment of the application.

[0056] Figure 3 is the total cost change diagram of the distance-based offloading scheme (D-JOA), the random offloading scheme (R-JOA), the order-based offloading scheme (O-JOA) and the technical scheme of the application with different mobile device quantities under the condition that the number of relay devices is 4 and the maximum tolerable delay is 1s.

[0057] Figure 4 is the total cost change diagram of the above technical schemes with different maximum tolerable delays under the condition that the number of mobile devices is 20 and the number of relay devices is 4.

[0058] Figure 5 is the convergence performance curve diagram of the calculation offloading and resource allocation method under the relay-assisted MEC architecture in the application. r m eWhen T = 1, the present application compares the two benchmark algorithms of the MEC system of a single base station (SBS) and the MEC system assisted by D2D communication to evaluate the overall execution cost curve under different relay quantities.

[0059] Figure 6 is the total cost curve of the above technical solutions of the present application varying with different maximum tolerable time delays.

[0060] Figure 7 is the average execution cost, tolerable delay, and relay quantity relationship curve of the present application when the task size is L i = 200 bits.

[0061] Figure 8 is the task size and relay device quantity relationship curve when the maximum tolerable time T = 1 s. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions, and advantages of the present application clearer, the embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0063] Embodiment One:

[0064] The present embodiment provides a computing offloading and resource allocation method under a relay-assisted MEC architecture, comprising:

[0065] Step 1: Construct a relay-assisted MEC system model, define the functions and task processing modes of relay devices, intelligent devices, and edge servers;

[0066] Step 2: Jointly formulate the task division strategy, computing resource allocation strategy, relay selection strategy, and uplink spectrum allocation strategy in the relay-assisted MEC system model to form a joint optimization problem, the goal of which is to minimize the system execution cost while ensuring the quality of service;

[0067] Step 3: Design a preliminary solving method, use the block coordinate descent method to decompose the joint optimization problem into multiple sub-problems, each of which optimizes the task division strategy, computing resource allocation strategy, relay selection strategy, and uplink spectrum allocation strategy;

[0068] Step 4: Further optimize the preliminary solving result, approximate the non-convex problem to a convex optimization problem using the reconstruction linearization technology, and apply a convex optimization method to solve it.

[0069] Embodiment Two

[0070] The embodiment provides a computing offloading and resource allocation method under a relay-assisted MEC architecture. The system model to which the method is applicable is as follows: a relay-assisted MEC system composed of intelligent devices SDs, relay devices relays and a base station BS powered by an edge server. The set of intelligent devices and relay devices in the system can be represented by and respectively. Specifically, the relay devices and the base station can simultaneously provide communication and computing services for the intelligent devices, both of which have certain data storage and processing capabilities and can assist the intelligent devices in task forwarding and offloading.

[0071] The tasks generated by the application have three execution modes, including local execution, relay offloading and edge computing. In the relay device-MEC scenario, the intelligent devices cannot directly communicate with the base station and must be forwarded through the relay devices. In addition, in order to encourage the relay devices to serve more resource-hungry users and solve the resource competition problem among users, the embodiment proposes a user charging system, which sells the resources and services of the relay devices and the base station to users through pricing. Each intelligent device will generate a computing-intensive and delay-urgent random task, which can be represented by a data tuple , where C i represents the computing resources required in CPU cycles for processing a unit of bit task, L i represents the total data volume of the computing task, and T i represents the maximum tolerable delay for completing the task. In the embodiment, for simplicity, all tasks are assumed to be completed within a constant time block T.

[0072] Due to the limitations of link quality and computing resources, it is assumed that the intelligent devices can be associated with the relay devices through some selection strategies. Due to the task divisibility, these relay devices will help offload part of the task to the MEC server and also execute the remaining task within the capability range. In this way, the intelligent devices can offload tasks without consuming a large amount of transmission energy. The quality of the relay device selection strategy directly affects the energy consumption of the intelligent devices in the system and also affects the cost of resource acquisition of the users.

[0073] The computing offloading and resource allocation method of the embodiment comprises the following steps.

[0074] Step 1: Construct a relay-assisted MEC system model and define the functions and task processing modes of the relay devices, the intelligent devices and the edge server. The task processing modes include three modes of local execution, relay offloading and edge computing.

[0075] A decision matrix is used to describe the association relationship between the i th intelligent device and the j th relay device, where x i,j∈{0, 1} is a binary selection variable, x i,j = 1 indicates that the ith smart device selects the jth relay device as a relay, while x i,j = 0 indicates that the ith smart device and the jth relay device have no matching relationship. According to the general physical meaning of relay selection, the constraint condition of the selection variable x i,j is shown below

[0076]

[0077] Since the system adopts a partial offloading model, the task can be divided into three parts: local computing, relay offloading, and edge offloading, which satisfy the following relationship:

[0078]

[0079] wherein, and represent the number of computing bits allocated to local computing, relay offloading, and edge offloading, respectively, L i represents the total data volume of the current computing task.

[0080] For the ith smart device, the delay can be divided into three parts: local computing, relay offloading, and edge offloading, which are simply represented by In particular, for the relay device and the base station, the offloading stage is further divided into two states: transmission and computing, that is,

[0081] The communication model is divided into two stages, corresponding to two transmission delays. In the first stage, the ith smart device transmits the offloaded data of to the selected jth relay device within the time slot Since the relay adopts a simplex communication mode, it needs to complete all data reception before performing the computing and forwarding tasks; in the time slot , the jth relay device forwards the received data to the MEC server in the second stage. The channel access of the two stages from the smart device to the relay device and from the relay device to the base station adopts the orthogonal frequency division multiplexing (OFDM) mode. h i,j represents the power gain of the subchannel, B i is the subchannel bandwidth allocated to the jth relay device for the ith smart device, p i,j is the transmission power of the ith smart device, and N i,j is the noise power. If α i represents the proportion of the uplink transmission subbandwidth to the total bandwidth, then B i = Bα i , α i ∈{0, 1},∑i α i ≤1.

[0082] Since data cannot be directly offloaded from the local to the base station, the relay needs to receive data including the part of data processed by the base station computing power in addition to the data it needs to perform itself. In order to complete the assigned task within the maximum tolerable latency T, the time limit of local computing and relay offloading process is:

[0083]

[0084] where, denotes the total latency of the local computing phase under the partial offloading model, denotes the computing latency in the local computing phase, C i the number of CPU cycles required per bit, and represent the CPU frequency of the ith smart device and the jth relay device serving the assigned task, respectively, satisfying and denotes the total latency of the relay offloading phase, denotes the transmission latency in the relay offloading phase, denotes the total latency of the relay computing phase, and denote the maximum computing capacity that the smart device and the relay device can assign to the task, respectively.

[0085] The total energy consumption of the ith smart device associated with the jth relay device for local computing and task offloading can be represented as:

[0086]

[0087] where, E comm denotes the energy consumption of the task offloading to the relay in the first transmission phase, E comp denotes the energy consumption that the SD needs to consume to complete the local computing task, is the power consumption per CPU cycle, κ i is the energy consumption coefficient depending on the chip architecture.

[0088] Step 2: Task partitioning strategy Computing resource allocation strategy Relay selection strategy and uplink spectrum allocation strategy The problem is jointly formulated as a joint optimization problem (JORA) of computation offloading, relay selection strategy, computation resource allocation, and spectrum resource allocation, and its objective is to minimize the system execution cost while guaranteeing the quality of service (QoS).

[0089] The unit cost of relaying, storing, and forwarding data for one smart device task is m r , and the supply price of one unit of computation bit data is m c ; the service cost of supplying the i-th smart device task is denoted by M i , which can be obtained when the task is successfully collected by the demanded relay:

[0090]

[0091] Considering the limitations of communication, computing power, and other aspects, for a given task partition strategy computation resource allocation strategy relay selection strategy and uplink spectrum allocation strategy The JORA optimization problem is expressed by the following formula (8):

[0092]

[0093] wherein, is the solution of P1, ω1 can be understood as the unit cost of using computing power services for the i-th smart device, and ω2 is the weight of offloading services to ensure fairness.

[0094] Step 3: Design a preliminary solution algorithm, use the block coordinate descent method to decompose the joint optimization problem (JORA) into multiple sub-problems, each of which optimizes the task partition strategy computation resource allocation strategy relay selection strategy and uplink spectrum allocation strategy

[0095] temporarily fix the relay selection strategy and uplink spectrum allocation strategy jointly optimize the task partition strategy computation resource allocation strategy Two sub-problems:

[0096]

[0097] The objective function and constraints in the subproblem are convex, so P2 is a convex optimization problem. P2 can be solved by many convex optimization techniques, such as CVX, to obtain the local optimal solution at the current iteration.

[0098] For a given task partition strategy Computing resource allocation strategy Relay selection strategy of problem P1 And uplink spectrum allocation strategy P3 can be optimized by the problem:

[0099]

[0100] Where Obviously, due to the non-convex objective function, problem P3 is non-convex.

[0101] Step 4: Approximate the non-convex problem to a convex optimization problem by using the reformulation linearization technique (RLT), and apply the convex optimization method to solve it.

[0102] Considering the coupling of the relay selection strategy x i,j and the spectrum allocation α i , first relax the variable method to facilitate the design of low complexity algorithm, and then use the constant adding method and reformulation-linearization technique (RLT) to eliminate the quadratic term. x i,j is a 0-1 integer variable, and the discrete variable x i,j is relaxed to 0≤x i,j ≤1. At the same time, since there may be a case of division by zero, a constant θ is added to the original variable α i , and let where the value of θ is approximately 0. Since there is also a product term , it is transformed by the RLT technique. For the second order term introduce auxiliary variables where x i,j and are limited to 0≤x i,j ≤1 and At this time, the transformed P3 can be formulated as:

[0103]

[0104] The transformed problem P3.1 is convex, so the solution of the problem can be obtained by many solvers (such as CVX). Substitute the solution into and use the rounding method to restore the optimization variable x i,j .

[0105] The iterative optimization process is as follows:

[0106]

[0107] In order to make the object, technical solutions and advantages of the present application clearer, some classical detection algorithms will be compared with the proposed algorithm below, and the superiority of the calculation offloading and resource allocation method under the relay assisted MEC architecture of the present application in performance will be shown. Figure 2 The relationship between user location distribution and relay selection in the task offloading scenario is shown from the perspective of relay selection strategy.

[0108] The baseline algorithms used for simulation are distance-based offloading scheme (D-JOA), random offloading scheme (R-JOA) and order-based offloading scheme (O-JOA). The experimental results are shown in Figure 3 、 4 From the simulation curves, it can be found that the calculation offloading and resource allocation method of the present application can effectively optimize the total cost of the system by comparing the total cost under different numbers of intelligent devices and different maximum tolerable delays.

[0109] The MEC system with a single base station (SBS) and the MEC system assisted by D2D communication are used as benchmark algorithms to evaluate the performance of the JORA algorithm, and the convex optimization technique is used for iterative optimization. Figure 5 The performance of the JORA joint optimization method of the present application is the best, and the overall execution cost shows an upward trend with the increase of the number of relays. Figure 6 It shows that as the maximum tolerable delay increases, the total cost of the system decreases, and the JORA joint optimization method of the present application shows the most economical performance.

[0110] When the task size is set to L i = 200 bits, Figure 7 The relationship between the average execution cost, tolerable delay and the number of relays is shown. The relationship between the task size and the number of relay devices is shown in Figure 8 , where the maximum tolerable time T = 1 second.

[0111] In summary, the simulation results verify the performance of the JORA joint optimization method of the present application in different scenarios, and compared with the baseline algorithm, the algorithm can greatly reduce the execution cost. Therefore, the calculation offloading and resource allocation method of the present application can complete the low delay task of SDs while effectively improving the performance of the network system.

[0112] Some steps in the embodiments of the present application can be implemented by software, and the corresponding software program can be stored in a readable storage medium, such as an optical disc or a hard disk.

[0113] The above merely describes preferred embodiments of the present application and is not used to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for computation offloading and resource allocation under a relay-assisted MEC architecture, characterized in that, The method includes: Step 1: Construct a relay-assisted MEC system model, defining the functions and task processing modes of relay devices, intelligent devices, and edge servers; Step 2: Jointly formulate the task partitioning strategy, computing resource allocation strategy, relay selection strategy and uplink spectrum allocation strategy in the relay-assisted MEC system model to form a joint optimization problem. The goal of the joint optimization problem is to minimize the system execution cost while ensuring the quality of service. Step 3: Design a preliminary solution method. The block coordinate descent method is adopted to decompose the joint optimization problem into multiple sub-problems. Each sub-problem optimizes the task partitioning strategy, computing resource allocation strategy, relay selection strategy, and uplink spectrum allocation strategy, respectively. Step 4: Further optimize the preliminary solution results by using the reconstruction linearization technique to approximate the non-convex problem into a convex optimization problem, and then apply the convex optimization method to solve it; Step 1 includes: Using decision matrix Describe the relationship between the i-th smart device and the j-th relay device, where x i，j ∈{0,1} is a binary choice variable, x i，j =1 indicates that the i-th smart device selects the j-th relay device as the relay, while x i，j = 0 indicates that the i-th smart device and the j-th relay device have no matching relationship. According to the general physical meaning of relay selection, the variable x is selected. i，j The restrictions are as follows: Because the system adopts a partial offloading model, the task is divided into three parts: local computation, relay offloading, and edge offloading. in, and Describe the number of computation bits allocated to local computation, relay offloading, and edge offloading, respectively, L i This indicates the total amount of data for this computation task; For the i-th smart device, latency is divided into three parts: local computation, relay offload, and edge offload. This indicates that, for relay equipment and base stations, the offloading phase is divided into two states: transmission and computation. This indicates that the communication process is divided into two stages, each corresponding to a different transmission delay. The first stage is when the i-th smart device... Within the time slot The offloading data is sent to the selected j-th relay device, and the second stage is in The j-th relay device will receive within the time slot Data is forwarded to the MEC server; using h i，j B represents the power gain of the sub-channel. i p is the sub-channel bandwidth allocated to the j-th relay device for the i-th smart device. i，j It is the transmit power of the i-th smart device, and N is the transmit power of the i-th smart device. i，j It is noise power; if α is used i Let B represent the proportion of the uplink transmission sub-bandwidth of the i-th smart device to the total bandwidth. i =Bα i α i ∈{0,1},∑ i α i ≤1, B represents the total bandwidth; In order to complete the assigned task within the maximum tolerable latency T, the time constraints for the local computation and relay offloading process are: Where C i The number of CPU loops required per bit. and The CPU frequencies of the i-th smart device and the j-th relay device, respectively, for the assigned tasks, satisfy the following conditions: and and These represent the maximum computing power that intelligent devices and relay devices can allocate to the task, respectively. Indicates the latency of relay computing; The total energy consumption of the i-th intelligent device associated with the j-th relay device for local computing and task offloading is expressed as: in, It is the power consumption per CPU cycle, κ i It depends on the chip architecture's power consumption coefficient, E comm E represents the energy consumption of offloading the task to the relay during the first transmission phase. comp This indicates the energy required for a smart device to complete a local computing task.

2. The method according to claim 1, characterized in that, Step 2 includes: The unit cost of relaying, receiving, storing, and forwarding data to a smart device is m. r The supply price of a unit bit of data is m. c ; Use M i This represents the service cost of supplying the i-th smart device task, which is obtained when the task is successfully collected by the relay of the demand: For a given task partitioning strategy Computing resource allocation strategy Relay selection strategy and uplink spectrum allocation strategy The joint optimization problem can be expressed using the following equation (8): in, For the solution P1, ω1 represents the unit cost of the computing power service used by the i-th smart device, and ω2 is the weight of the offloading service. This represents the maximum CPU frequency used by the i-th smart device for local computation.

3. The method according to claim 2, characterized in that, Step 3 includes: Given a relay selection strategy and uplink spectrum allocation strategy Joint optimization of task partitioning strategy Computing resource allocation strategy The two subproblems are represented as follows: The objective function and constraints in subproblem P2 are both convex functions. P2 is solved using convex optimization techniques to obtain the local optimal solution for the current iteration number. Given a task partitioning strategy Computing resource allocation strategy Joint optimization relay selection strategy and uplink spectrum allocation strategy The two subproblems are represented as follows: in 4. The method according to claim 3, characterized in that, Step 4 includes: Using the slack variable method to slack off discrete variables x i，j Relax to 0≤x i，j ≤1; make The numerical value of θ is approximately 0; For second-order terms Introducing auxiliary variables Where x i，j and Each is subject to 0 ≤ x i，j ≤1 and Transform P3 into: Using convex optimization techniques to solve P3.1, we obtain the local optimum at the current iteration number and substitute the solution into... The optimization variable x is rounded off. i，j To restore.

5. The method according to claim 1, characterized in that, The channel access from the smart device to the relay device and from the relay device to the base station adopts orthogonal frequency division multiplexing.

6. The method according to any one of claims 1-5, characterized in that, The convex optimization method used in step 4 is the CVX solver.

7. The method according to claim 1, characterized in that, The task processing modes in step 1 include: local execution, relay unloading, and edge computing.

8. An electronic device, characterized in that, Including memory and processor; The memory is used to store computer programs; The processor is configured to implement the method as described in any one of claims 1 to 7 when executing the computer program.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images