Time delay minimization unloading method and system in multi-user industrial Internet of Things scene

Abstract

The invention discloses a time delay minimization unloading method and system in a multi-user industrial Internet of Things scene. The method comprises the following steps: S1, initializing; s2, establishing a joint optimization model; s3, modeling is carried out at a user side, and an optimal partial unloading strategy of each user is solved; s4, modeling is carried out on the cloud, and the optimal computing resource allocation of the MEC server is solved through an optimal partial unloading strategy; and S5, calculating the minimum time delay of the whole system through the solved optimal partial unloading strategy and the optimal computing resource allocation of the MEC server. According to the method, the original optimization model is modeled at the user side and the server side respectively, so that the calculation complexity is greatly reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of mobile edge computing networks, and in particular relates to a delay minimization unloading method and system in a multi-user industrial Internet of Things scenario. Background Art

[0002] The Industrial Internet of Things is an important network for modern industrial enterprises to move towards automation and digitalization in Industry 5.0, bringing value-added services to multiple fields such as smart manufacturing, smart grid, and smart digital supply chain. The surge in connected devices poses a challenge to the Industrial Internet of Things network that relies on frequent communications. To overcome these problems, mobile edge computing has been proposed and widely used. Mobile edge computing improves the computing power at the edge of the network by deploying high-performance servers closer to users, solves the problem of limited resources of mobile devices and the problem of excessive load pressure on cloud computing networks, and can achieve ultra-low latency and ultra-low energy consumption requirements. However, the MEC-based offloading system still faces some challenges. MEC servers are located closer to user devices than cloud servers, but MEC servers have fewer resources than central clouds. Usually edge servers provide services to nearby users. In addition, due to the mobility of users, channel conditions in MEC systems change, especially in the case of multiple users. Therefore, it is crucial to design efficient computing offloading strategies.

[0003] In the prior art, computing tasks are usually set to support binary computing offloading. This means that computing tasks can only be calculated locally on the user device, or completely offloaded to the MEC server. However, in some specific scenarios, the computing offloading strategy should support partial offloading according to the characteristics of the computing task to make full use of the limited resources of the MEC server. Based on this, the present invention coordinates the calculation between the edge server and the user's local side, and unloads the task to the local or MEC server with the total computing resources and offloading strategy of the system as constraints. The model is modeled with the purpose of minimizing the total system delay, and then a minimum-maximum algorithm is used to solve it, and finally the minimum delay of the entire system is calculated. Summary of the invention

[0004] In view of the above-mentioned status of the prior art, the present invention discloses a delay-minimizing unloading method and system in a multi-user industrial Internet of Things scenario. In a wireless network of a multi-user single MEC server, the present invention uses the user's unloading decision and the server's computing resources as constraints, minimizes the total system delay as the goal of modeling, and proposes an algorithm for joint computing unloading and resource allocation. The algorithm first models the user side to solve the optimal partial unloading strategy for each user, and then models the server side. The optimal computing resource allocation of the MEC server is solved by the solved optimal partial unloading strategy, and finally the minimum delay of the entire system is solved by the solved optimal partial unloading strategy and the optimal computing resource allocation of the MEC server.

[0005] In order to achieve the purpose of the present invention, the present invention adopts the following technical scheme:

[0006] A delay-minimizing offloading method in a multi-user industrial Internet of Things scenario comprises the following steps:

[0007] Step 1: Initialization phase: Each user node obtains basic configuration information of the network through information interaction;

[0008] Step 2: Establishing the optimization model: With the goal of minimizing the total system delay, a joint optimization model of offloading decision and resource allocation is established based on task offloading constraints and user resource allocation constraints.

[0009] Step 3: Model the user end and solve the optimal uninstallation strategy for each user;

[0010] Step 4: Model the server side and solve the optimal computing resource allocation of the MEC server by solving the optimal partial offloading strategy;

[0011] Step 5: Determine the minimum delay of the entire system by solving the optimal partial offloading strategy and the optimal computing resource allocation of the MEC server.

[0012] Furthermore, in step one, each node obtains basic configuration information of the network through information interaction, and the basic configuration information includes topology information, transmission power, link distance, task data size of user equipment, local computing resources and MEC computing resources, etc.

[0013] Furthermore, in step 2, the optimization model established is:

[0014]

[0015] θ n ∈[0,1](2)

[0016]

[0017]

[0018] Where θ represents the set of partial offloading decisions of the user, f represents the set of computing resources allocated to the user by the MEC server; Formula (2) represents the task offloading decision of the user; Formula (3) indicates that the computing resources allocated by the MEC server are positive and should not exceed the total computing resources of the MEC server; Formula (4) indicates that the computing resources allocated by the MEC server to all offloaded tasks should not exceed the total computing resources of the MEC server.

[0019] Furthermore, in step three, modeling is first performed on the user side:

[0020]

[0021] st(2)

[0022] Wherein, formula (2) represents the user's task offloading decision. When , the optimal offloading strategy that minimizes the delay can be solved.

[0023] Proof: First, when When , we can solve:

[0024]

[0025] When the local computation delay is large, the derivative of the objective function is:

[0026]

[0027] It can be found that the objective function T n At 0≤θ n ≤θ mec When the offloading execution delay is large, the derivative of the objective function is:

[0028]

[0029] It can be found that the objective function T n In θ mec ≤θ n ≤1 when it increases monotonically. Therefore, it can be proved that when When , there is a minimum value. Therefore, the optimal uninstallation decision for each user n can be solved as:

[0030]

[0031] Furthermore, in step 4, modeling is performed on the server side, and the optimal computing resource allocation of the MEC server is solved by solving the optimal partial offloading strategy. Then the minimized total delay of the entire system is:

[0032]

[0033] st(3)(4)

[0034] If and only if T 1 =T 2 =T 3 =…=T n When , we can obtain the solution of (10). The proof is the same as above.

[0035] Furthermore, in step five, the minimum delay of the entire system is solved by solving the optimal partial offloading strategy and the optimal computing resource allocation of the MEC server.

[0036] The present invention also discloses a time delay minimization unloading system in a multi-user industrial Internet of Things scenario, which is used to execute the above method and includes the following modules:

[0037] Initialization module: Each user node obtains basic configuration information of the network through information interaction;

[0038] Joint optimization model building module: with the goal of minimizing the total system delay, a joint optimization model of offloading decision and resource allocation is established according to task offloading constraints and user resource allocation constraints;

[0039] User-side modeling module: Models the user side and solves the optimal uninstallation decision for each user;

[0040] Server-side modeling module: Model the server side and solve the optimal computing resource allocation of the MEC server through the optimal offloading strategy;

[0041] Minimum delay solution module: solves the minimum delay through the optimal offloading strategy and the optimal computing resource allocation of the MEC server.

[0042] The present invention has the following advantages:

[0043] (1) The present invention realizes a delay minimization offloading method in a multi-user industrial Internet of Things scenario. Experiments show that the method proposed by the present invention is smaller than the total system delay calculated by the existing algorithm.

[0044] (2) The most prominent advantage of the present invention is that the original optimization model is modeled on the user side and the server side respectively, and an optimal offloading algorithm is proposed to simultaneously solve the offloading problem on the user side and the computing resource allocation problem on the server side, thereby greatly reducing the computational complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1The present invention is a flowchart of a method for minimizing delay in a multi-user industrial Internet of Things scenario according to a preferred embodiment of the present invention.

[0046] Figure 2 This is a system model diagram of a multi-user single MEC server.

[0047] Figure 3 This is a graph showing how the number of users changes with the total system delay under different methods.

[0048] Figure 4 The figure shows the change of the computing power of the MEC server to the total system delay under different methods.

[0049] Figure 5 This is a block diagram of a delay-minimizing offloading system in a multi-user industrial Internet of Things scenario in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0050] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0051] like Figure 1-2 As shown, this embodiment provides a delay minimization unloading method in a multi-user industrial Internet of Things scenario, including the following steps:

[0052] Step 1, initialization phase: Each node obtains basic configuration information of the network through information exchange, and the basic configuration information includes topology information, transmission power, link distance, task data size of user equipment, local computing resources and MEC computing resources, etc.

[0053] This embodiment considers a multi-user industrial IoT mobile edge computing system. The system consists of a base station, multiple user devices, and an MEC server. The MEC server is deployed together with the base station. The set of user devices is represented by N = {1,2,...,N}. The user device can perceive the production environment in the Industry 5.0 scenario, and each user device has an AI-based computationally intensive task to be calculated. However, due to the limitation of user device resources, the user device may not be able to meet the computing requirements of such computationally intensive tasks. The user device can migrate the task to the MEC server, thereby reducing the computing delay of the task. Compared with the traditional cloud, the MEC server is closer to the user, which can reduce the communication cost between the user device and the MEC server. However, the computing power and storage capacity of the MEC server are relatively low, and the MEC server may not be able to meet all tasks at the same time. Assume that the task priority of each user device is the same. When the tasks are offloaded at the same time, the wireless bandwidth resources will be evenly allocated to each user device to upload their tasks. The upload data rate of the nth user device is:

[0054]

[0055] Where W is the bandwidth of the wireless channel, K is the number of user devices being offloaded, and p n is the data transmission power of the nth user equipment uploading, h n is the wireless channel gain between the nth user equipment and the base station, N 0 is the noise power spectral density.

[0056] R n =(D n ,C n ) is defined as the task that the nth user device needs to perform. This task can be calculated locally or on the MEC server through computation offloading. n represents the amount of data required to complete this task, C n Indicates completion of task R n The number of CPU cycles required. User equipment tasks can be offloaded in any proportion. This is the substantial difference between the model of the present invention and other existing binary computing offloading solutions.

[0057] The present invention uses θ n ∈[0,1] as the computation offloading strategy for the nth user device, that is, the nth user device will n Some tasks are offloaded to the MEC server, and (1-θ n ) is calculated locally. The computation offloading strategy θ of the user device set N consists of all user devices, expressed as θ = [θ 1 ,θ 2 ,...,θ n ]. When θ n =0, indicating the nth task R n is calculated locally. When θ n =1, indicating R n Completely offloaded to the MEC server. When 0<θ n When <1, it means that the task R n θ n Part of the task is offloaded to the MEC server, and the (1-θ n ) is calculated locally.

[0058] Step 2: Establish optimization model: With the goal of minimizing the total system delay, a joint optimization model of offloading decision and resource allocation is established based on task offloading constraints and user resource allocation constraints.

[0059] When the user device performs tasks locally, the task completion process has nothing to do with the MEC server and completely uses local computing resources. n The delay T in local computation n lIt can be expressed as:

[0060]

[0061] Among them, f n l Indicates the computing power of the nth user device (the number of CPU cycles per second of the nth user device).

[0062] When the user device requests to offload the computation, task R n θ n Part of it is offloaded to the MEC server. Task R n θ n The total delay of the part includes data upload delay, MEC calculation delay and data download delay. n (1-θ n ) is calculated locally. Therefore, task R n θ n Upload latency when partially offloading to MEC server It can be expressed as:

[0063]

[0064] Task R n θ n Computation latency when partially offloading to MEC servers It can be expressed as:

[0065]

[0066] Among them, f n Represents the task R assigned by MEC n The computing resources allocated by the MEC server to all offloaded tasks should not exceed the total computing resources, which means that the condition

[0067] Task R n θ n Data download delay when partially offloading to MEC server It can be expressed as:

[0068]

[0069] Among them, B n Indicates the data size of the execution result returned by the MEC server, r' n Indicates the download rate of the nth user equipment downloading the execution result from the MEC server. Due to the size of the uploaded data D n Much larger than Bn , the download rate is much higher than the upload rate. Therefore, the present invention does not consider the data download delay problem.

[0070] For task R n (1-θ n ) is calculated locally, and the process does not need to consider the MEC server. Therefore, we can get R n (1-θ n ) part of the computational delay. n (1-θ n ) part of the calculation delay It can be expressed as:

[0071]

[0072] Due to the task R n The total time delay of the nth user device to execute the partial calculation offloading strategy should be task R n θ n Partial Delay and Task R n (1-θ n ) part of the delay. According to formula (3)(4)(5)(6), define T n Task R performs a partial computation offloading strategy for the nth user device n The total delay, T n The calculation method is:

[0073]

[0074] In formula (7), when θ n = 0, we can get When θ n = 1, we can get Therefore, the time delay of the three strategies of local computing, partial computing offloading, and complete computing offloading can be comprehensively considered using formula (7). In the present invention, the user device side is first considered, and it is hoped that the delay of each user device of the present invention is minimized. Secondly, the present invention minimizes the task execution time of all user devices in the entire industrial Internet of Things mobile edge system as the optimization goal of the present invention. According to the above network model, computing task model, local computing model and edge computing model, the optimization problem can be expressed as:

[0075]

[0076] θ n ∈[0,1](9)

[0077]

[0078] Where θ represents the set of partial offloading decisions of the user, f represents the set of computing resources allocated to the user by the MEC server; Formula (2) represents the task offloading decision of the user; Formula (3) indicates that the computing resources allocated by the MEC server are positive and should not exceed the total computing resources of the MEC server; Formula (4) indicates that the computing resources allocated by the MEC server to all offloaded tasks should not exceed the total computing resources of the MEC server.

[0079] The original optimization model is modeled on the user side and the cloud side respectively.

[0080] Step 3: First, model the model on the user side, which is specifically expressed as follows:

[0081]

[0082] st(9)

[0083] Wherein, formula (9) represents the user's task offloading decision. When , the optimal offloading strategy that minimizes the delay can be solved.

[0084] Proof: First, when When , we can solve:

[0085]

[0086] When the local computation delay is large, the derivative of the objective function is:

[0087]

[0088] It can be found that the objective function T n At 0≤θ n ≤θ mec When the offloading execution delay is large, the derivative of the objective function is:

[0089]

[0090] It can be found that the objective function T n In θ mec ≤θ n ≤1 when it increases monotonically. Therefore, it can be proved that when When , there is a minimum value. Therefore, the optimal uninstallation decision for each user n can be solved as:

[0091]

[0092] Step 4: Model the cloud and solve the optimal computing resource allocation of the MEC server by solving the optimal partial offloading strategy θ. Then the minimized total delay of the entire system is:

[0093]

[0094] st(10)(11)

[0095] If and only if T 1 =T 2 =T 3 =…=T n When , we can obtain the solution of (17). The proof is the same as above.

[0096] Step 5: Substitute the optimal partial offloading strategy θ and the optimal computing resource allocation f of the MEC server into equation (8) to solve the minimum delay of the entire system.

[0097] exist Figure 3 In the paper, the influence of the number of user devices on the total delay of the system is studied. The horizontal axis represents the number of user devices, and the vertical axis represents the total delay for the completion of tasks of all user devices. As can be seen from the figure, the total delay of all unloading increases with the increase in the number of user devices. This is because as the number of user devices increases, the computing resources allocated by the server to each device will decrease, resulting in an increase in the delay of each user device, and the total delay of the entire system will increase. The total delay of partial unloading of the present invention will also increase slowly. When the number of user devices is large enough, the total delay of partial unloading will approach the total delay of local calculation. This is because when the number of users is large enough, the computing resources allocated by the server will be very small. It can almost be considered that the user's computing tasks are all calculated locally, so it will approach the total delay of local calculation.

[0098] exist Figure 4 In the paper, the influence of the computing power of the MEC server on the total delay of the system is studied. The horizontal axis represents the computing power of the MEC server, and the vertical axis represents the total delay of the task completion of all user devices. As can be seen from the figure, the total delay of all unloading decreases with the increase of the computing power of the MEC server. This is because as the computing power of the MEC server increases, the server will allocate more computing resources to each device, resulting in a reduction in the delay of each user device, and the total delay of the entire system will be reduced. At the same time, it will be found that the total delay of partial unloading is always less than the total delay of full unloading, and when the computing power of the MEC server is infinite, the curve will be infinitely close but never intersect. This is because the partial unloading of the present invention is unloading locally and the server at the same time, while the full unloading is just the server performing tasks alone. When the computing power of the MEC server is infinite, most of the tasks in the partial unloading will be handed over to the server to complete, and only a few tasks will be handed over to local calculations, so they will be infinitely close but never intersect.

[0099] like Figure 5As shown, this embodiment discloses a time delay minimization unloading system in a multi-user industrial Internet of Things scenario, which is used to execute the above method and includes the following modules:

[0100] Initialization module: Each user node obtains basic configuration information of the network through information interaction;

[0101] Joint optimization model building module: with the goal of minimizing the total system delay, a joint optimization model of offloading decision and resource allocation is established according to task offloading constraints and user resource allocation constraints;

[0102] User-side modeling module: Models the user side and solves the optimal uninstallation decision for each user;

[0103] Server-side modeling module: Model the server side and solve the optimal computing resource allocation of the MEC server through the optimal offloading strategy;

[0104] Minimum delay solution module: solves the minimum delay through the optimal offloading strategy and the optimal computing resource allocation of the MEC server.

[0105] For other contents of this embodiment, please refer to the above method embodiment.

[0106] The preferred embodiments and principles of the present invention are described in detail above. For those skilled in the art, according to the ideas provided by the present invention, there may be changes in the specific implementation methods, and these changes should also be regarded as the protection scope of the present invention.

Claims

1. A delay minimization offloading method in a multi-user industrial Internet of Things scenario, characterized in that: The specific steps include: Step 1: Each user node obtains basic configuration information of the network through information interaction; Step 2: With the goal of minimizing the total system delay, a joint optimization model of offloading decision and resource allocation is established according to the task offloading constraints and user resource allocation constraints; Step 3: Model the user end and solve the optimal uninstallation decision for each user; Step 4: Model the server side and solve the optimal computing resource allocation of the MEC server through the optimal offloading strategy; Step 5: Determine the minimum delay through the optimal offloading strategy and the optimal computing resource allocation of the MEC server.

2. The delay minimization unloading method in a multi-user industrial Internet of Things scenario according to claim 1 is characterized in that: In step 1, the basic configuration information includes topology information, transmit power, link distance, task data size of user equipment, local computing resources and MEC computing resources.

3. The delay minimization unloading method in a multi-user industrial Internet of Things scenario according to claim 1 or 2 is characterized in that: In step 2, the joint optimization model is as follows: i n ∈[0,1](2) Where θ represents the set of partial offloading decisions of the user, f represents the set of computing resources allocated to the user by the MEC server; Formula (2) represents the task offloading decision of the user; Formula (3) indicates that the computing resources allocated by the MEC server are positive and do not exceed the total computing resources of the MEC server; Formula (4) indicates that the computing resources allocated by the MEC server to all offloaded tasks do not exceed the total computing resources of the MEC server.

4. The delay minimization unloading method in a multi-user industrial Internet of Things scenario according to claim 3 is characterized in that: In step 3, the user end is modeled, and the minimum delay for each user n is: Among them, formula (2) represents the user's task offloading decision; when When , the optimal offloading strategy that minimizes the delay is obtained.

5. The delay minimization unloading method in a multi-user industrial Internet of Things scenario according to claim 4 is characterized in that: Step 4: Model the server side and solve the optimal computing resource allocation of the MEC server by solving the optimal partial offloading strategy θ. The minimized total delay of the entire system is: When T1=T2=T3=…=T n When , we obtain the solution of (10).

6. A time-delay minimization unloading system in a multi-user industrial Internet of Things scenario, used to execute the method according to any one of claims 1 to 5, characterized in that: The system includes the following modules: Initialization module: Each user node obtains basic configuration information of the network through information interaction; Joint optimization model building module: with the goal of minimizing the total system delay, a joint optimization model of offloading decision and resource allocation is established according to task offloading constraints and user resource allocation constraints; User-side modeling module: Models the user side and solves the optimal uninstallation decision for each user; Server-side modeling module: Model the server side and solve the optimal computing resource allocation of the MEC server through the optimal offloading strategy; Minimum delay solution module: solves the minimum delay through the optimal offloading strategy and the optimal computing resource allocation of the MEC server.

Images