An edge computing resource scheduling optimization method for scalable video

Abstract

This invention discloses an edge computing resource scheduling optimization method for scalable video. The method includes: setting user video task parameters, modeling the user, and constructing a calculation model of the latency and energy consumption of the user executing tasks locally; constructing a calculation model of the signal-to-noise ratio of task offloading and the resulting uplink latency based on the computing power of the video terminal; setting MEC server parameters, and constructing a calculation model of video task T. u A computational model for execution time on MEC servers; a computational offloading model for edge computing systems oriented towards scalable video coding technology. u Construct a joint offloading and resource optimization model; solve the joint offloading and resource optimization model using KKT conditions and annealing simulation algorithm to obtain the optimized resource allocation strategy F. * and unloading decision χ * .

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology and relates to an edge computing resource scheduling optimization method for scalable video. Background Technology

[0002] In recent years, with the rapid development of mobile communication technology and multimedia terminals, countless video traffic has been injected into existing networks. According to Cisco's forecast, by the end of 2022, mobile video traffic will account for 79% of total mobile network traffic. This massive amount of video traffic poses a significant challenge to existing mobile networks. One important solution to address the problem of ever-increasing video traffic overload is the introduction of mobile edge computing, which offloads video tasks from the cloud to the edge for computation. Mobile-Edge Computing (MEC) technology provides distributed computing capabilities and localized cloud services by offloading highly complex and energy-intensive video tasks to the edge of the mobile network closer to the user.

[0003] To fully utilize dynamically changing bandwidth and improve the Quality of Experience (QoE) for mobile video users, Scalable Video Coding (SVC) technology has been introduced. SVC can provide mobile video at different bitrates based on available bandwidth. However, most existing video multicasting is based on Adaptive Modulation and Coding (AMC), treating the video content as a large file and choosing a lower bitrate to transmit the video stream to accommodate users with poor channel conditions. This results in multicast users with good channel conditions being limited by those with poor channel conditions, unable to effectively utilize valuable bandwidth resources. SVC technology can divide the video stream into a layered structure of a base layer and multiple enhancement layers. Therefore, SVC-based collaborative video multicast technology can flexibly adjust the video layer rate based on channel feedback to cope with different fading channels between multicast groups. Furthermore, for users with poor channel links, as long as they can correctly receive the base layer, even if the picture is no longer clear, normal playback is still possible.

[0004] To date, research on mobile edge computing video transmission scenarios based on SVC remains very limited. Some studies have utilized the energy consumption and quality of experience (QoE) of video streams over software-defined mobile networks (SDCs), jointly considering buffer dynamics, video quality adaptation, edge caching, video transcoding, and transmission issues to achieve energy savings and improved QoE. Other literature proposes a novel adaptive geographic routing scheme to establish single-path transmission routes in urban environments and then select the optimal connectivity route, which can reduce packet loss and optimize latency. However, existing MEC cooperative SVC video transmission schemes primarily rely on downlink transmission and do not consider the user's uplink transmission methods. Summary of the Invention

[0005] Objective: To overcome the shortcomings of existing technologies, this invention addresses the optimization of 5G edge computing resource scheduling for scalable video. Specifically, it solves the problem of resource scheduling optimization for uploading real-time video to the server for parsing, given the overload limitations of server video traffic. This invention provides a method for optimizing edge computing resource scheduling for scalable video.

[0006] This application considers the following video transmission scenario: multiple mobile video users in a group of geographically located locations upload the same real-time video, and each user's user equipment (UE) can access multiple base stations through a network interface to upload video content. For this scenario, a video upload scheme is designed by combining scalable video coding technology and utilizing the transmission characteristics of small base stations to improve transmission efficiency, thereby optimizing latency and resource allocation.

[0007] Technical solution: To solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0008] In a first aspect, the present invention provides an edge computing resource scheduling optimization method for scalable video, comprising:

[0009] S1. Set the user's video task parameters, model the user, and construct the latency of the user executing tasks locally. and energy consumption The computational model;

[0010] S2. Based on the computing power of the video terminal, construct the signal-to-noise ratio for task offloading. and the resulting uplink latency The computational model;

[0011] S3. Configure MEC server parameters and build video task T. u Execution time on the MEC server The computational model;

[0012] S4, based on and The computational model is used to construct a computational offloading model for edge computing systems oriented towards scalable video coding technology. u ;

[0013] S5, Based on the computational unloading model J u Construct a joint unloading and resource optimization model;

[0014] S6. Solve the joint unloading and resource optimization model using KKT conditions and annealing simulation algorithm to obtain the optimized resource allocation strategy F. * and unloading decision χ * .

[0015] In a second aspect, the present invention provides an edge computing resource scheduling optimization device for scalable video, including a processor and a storage medium;

[0016] The storage medium is used to store instructions;

[0017] The processor is configured to operate according to the instructions to perform the steps of the method according to the first aspect.

[0018] Thirdly, the present invention provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0019] Beneficial Effects: This invention addresses the shortcomings in resource optimization research for video transmission systems in traditional edge computing scenarios by proposing a resource scheduling optimization method for scalable video. Based on existing video transmission methods, the proposed solution considers the layered structure of SVC video and utilizes mobile edge computing collaboration to address the limited bandwidth resources of multi-MEC systems, enabling users to experience smoother video uploads. This solution constructs a mobile edge computing system for scalable video encoding and proposes an optimization scheme to address the latency and energy consumption consumed by video terminal task offloading. This scheme introduces SVC technology to solve the resource scheduling problem during video uploads. Using the KKT conditional and annealing simulation algorithms, the solution measures the weighted sum of task completion time and energy consumption, aiming to maximize the user's task offloading benefits. It jointly optimizes task offloading decisions, uplink transmission power for video users, and computing resource allocation on MEC servers to calculate the optimal decision for each video terminal's service task offloading, thereby maximizing system benefits. Through modeling and analysis, while meeting the requirements of network fluctuations and real-time video applications, this solution can allocate rates to different video layers based on the channel conditions of the path. Attached Figure Description

[0020] Figure 1This is a scenario model diagram of an edge computing resource scheduling optimization method for scalable video according to an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram illustrating a scene according to an embodiment of the present invention. Detailed Implementation

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and should not be used to limit the scope of protection of the present invention.

[0023] In the description of this invention, "several" means one or more, "multiple" means two or more, "greater than," "less than," and "exceeding" are understood to exclude the stated number, while "above," "below," and "within" are understood to include the stated number. The use of "first" and "second" in the description is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or implicitly indicating the order of the indicated technical features.

[0024] In the description of this invention, the terms "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0025] To address the aforementioned issues, this invention proposes a 5G edge computing resource scheduling optimization method based on scalable video coding. By introducing scalable video coding technology as an aid, and considering the different needs of video terminal users, it constructs an optimal decision-making scheme for MEC resource scheduling, addressing video traffic overload and combining the characteristics of latency-sensitive tasks. Based on the concept of scalable video coding and fully considering the SVC layered coding structure, this patent proposes a scheme for collaborative SVC video stream transmission in mobile edge computing. In the proposed video transmission scheme, through analysis and derivation, the optimization problem of different SVC video layer allocations in this scenario is presented, and a heuristic algorithm is proposed for solving it, jointly optimizing the system's latency and resource allocation. The scenario model for 5G edge computing resource scheduling optimization based on scalable video coding is as follows: Figure 1 As shown.

[0026] In the video transmission scenario discussed in this invention, the video tasks of the video terminal device can be offloaded to a small base station containing an MEC server for task computation. This patent, based on edge computing and combined with the layered characteristics of scalable video coding technology, primarily addresses the problem of optimizing the scheduling of transmission resources for video uploads. Figure 2 As shown, in the problem studied in this invention, the decision-maker located at the edge service node pre-estimates the computation offloading strategy for video tasks, and then decides on the optimal resource allocation of the wireless transmission network and the offloading location of the terminal's video tasks. The scenario discussed in this invention involves multiple users and multiple edge servers.

[0027] Example 1

[0028] An edge computing resource scheduling optimization method for scalable video includes:

[0029] S1. Set the user's video task parameters, model the user, and construct the latency of the user executing tasks locally. and energy consumption The computational model;

[0030] S2. Based on the computing power of the video terminal, construct the signal-to-noise ratio for task offloading. and the resulting uplink latency The computational model;

[0031] S3. Configure MEC server parameters and build video task T. u Execution time on the MEC server The computational model;

[0032] S4, based on and The computational model is used to construct a computational offloading model for edge computing systems oriented towards scalable video coding technology. u ;

[0033] S5, Based on the computational unloading model J u Construct a joint unloading and resource optimization model;

[0034] S6. Solve the joint unloading and resource optimization model using KKT conditions and annealing simulation algorithm to obtain the optimized resource allocation strategy F. * and unloading decision χ * .

[0035] In some embodiments, the scenarios considered are as follows Figure 1As shown, in a multi-unit, multi-server MEC system, each base station (BS) is equipped with an MEC server to provide computational offloading services for resource-constrained mobile video users (such as live webcasting, video surveillance, and wearable video devices). Generally, each MEC server can be either a physical server provided by the network operator or a virtual machine with moderate computing power, communicating with mobile devices via the wireless channel provided by the server. Each mobile user can choose to offload video tasks from nearby servers (s) to the MEC server. This paper uses U = {1,2,3,...,U} to represent the set of users and MEC servers in the mobile video system, and M = {1,2,3,...,m}. The modeling of user video tasks, task upload and transmission, MEC computing resources, and offloading utilities is described below.

[0036] Step 1: Set the user's video task parameters and model the user.

[0037] This patent assumes that each user u∈U, and each user has only one video task, denoted as T. u This task is atomic and cannot be divided into subtasks. Each video task T u The feature is determined by two parameters <d u ,c u >A tuple consisting of d u [bits] refers to the amount of input data required to be transferred from the local mobile device to the MEC server. u [cycles] refers to workload, that is, the amount of computation required to complete the task. And d u and c u The value can be obtained based on the amount of data executed by the user task. In this system, each task can be executed either locally on the user's end or offloaded to the MEC server. By offloading video tasks to the MEC server, video users save the energy required for video tasks; however, sending task input to the uplink consumes additional time and energy.

[0038] use This represents the local computing power of user u. The time is measured in CPU cycles per second. Therefore, if user u executes its task locally, the task completion time is:

[0039]

[0040] This patent calculates the user's energy consumption while the local device performs its tasks. It uses an energy consumption model that assumes each calculation cycle is ε = αf. 2Where κ represents the energy coefficient based on the chip architecture, and f is the CPU frequency.

[0041] Therefore, it is calculated that user u performs task T locally. u The energy consumption at that time is:

[0042]

[0043] Step 2: Based on the parameters given in Step 1 and the computing power of the video terminal, analyze the signal-to-noise ratio and latency of task offloading.

[0044] This patent sets the system as a multi-user, multi-MEC scenario, so for each user u∈U, the video task can be offloaded to any MEC server within the system. During the offloading process, the following delays will occur: (1) the time it takes to offload the task to the MEC server on the uplink. Where u represents the user and up represents the uplink; (2) the time to execute the video task on the MEC server; (3) the time for the MEC server to transmit the calculation results back to the user on the downlink. Since the output size is usually much smaller than the input, and the downlink data rate is much higher than the uplink data rate, we ignore the transmission delay of the output in the calculation.

[0045] In this work, the OFDMA system is configured as an uplink multiple access scheme, dividing the transmission frequency band B into N equal sub-bands of size W, i.e., W = B / N [Hz]. To ensure the orthogonality of uplink transmissions for users under the same server, each user can only transmit video data on one sub-band. Therefore, each server can serve a maximum of N users simultaneously. Assume that the set of available sub-bands for each server is N = {1, 2, 3, ..., N}. Considering the scheduling of uplink sub-bands, this patent defines the offload variable as... Where u∈U,m∈M,j∈N,l∈L, this formula indicates that user u unloads the l-th layer of the video to server m through subband j. Indicates task T u User u unloads the l-th layer of the video onto m via sub-band j. Conversely, this patent sets the unloading decision as χ, then... Since each task can be executed locally or offloaded to at most one MEC server, a viable offload strategy must meet the following constraints:

[0046]

[0047] Furthermore, this patent specifies that each user and server has a single antenna for uplink transmission. This patent assumes that the power of user u transmitting a task to the server is p. u[W], then P = {p u |0<p u ≤P u Since the uplink uses OFDMA technology, users on the same server upload tasks to the MEC using different subbands, effectively suppressing uplink intra-cell interference. However, interference still exists between mobile devices. In this case, the signal-to-noise ratio (SINR) from user u to server m on subband j is...

[0048]

[0049] Where, σ 2 The variance is the background noise, and the first term in the denominator is the cumulative intra-cell interference of all users associated with other servers m on the same subband j. This represents the channel gain coefficient between server m and user u, taking into account channel characteristics such as path loss and shadow fading. Here, the path loss model used in this patent is... R represents the distance (in kilometers) between server m and user u. Each user can only transmit on one subband, therefore the rate at which user u sends data to server m is R. um (χ)[bits / s] can be expressed as

[0050] R um (χ)=Wlog2(1+γ um (5)

[0051] in, χ represents the offloading decision. Therefore, user u sends video task d on the uplink. u The transmission time can be calculated as follows:

[0052]

[0053] in,

[0054] Step 3: Configure MEC server parameters.

[0055] In this system, each mini-server's MEC server can simultaneously provide compute offloading services to multiple users, which can be achieved using f. s The computing resources provided by each MEC server to associated users are quantified using CPU cycles / s. Users offload video tasks to the MEC server, which performs the computation and returns the output to the user upon completion. This patent defines the computing resource allocation strategy as F = {f...} um |u∈U,m∈M},fum For server m, task T will be u The amount of computing resources unloaded from user u, f um >0. Furthermore, the strategy for allocating computing resources must satisfy the computing resource constraints of the MEC server, therefore:

[0056]

[0057] Therefore, video task T u Execution time on the MEC server for:

[0058]

[0059] Step 4: Construct a computing model for edge computing systems oriented towards scalable video coding technology.

[0060] Based on the above descriptions of each module, the total latency t experienced by user u when unloading the task is... u for:

[0061]

[0062] Based on the system model description, the uplink energy consumption E of user u processing tasks is... u Represented as:

[0063]

[0064] Where, ξ u This represents the power amplifier efficiency for user u. In this system, assume ξ... u =1, then the uplink energy consumption model can be simplified to

[0065]

[0066] In mobile edge computing systems, this patent primarily considers user QoE, and the main characteristics of QoE are task completion time and energy consumption. In the video transmission scenario of this system, the characteristics of task completion time and energy consumption are set as follows:

[0067]

[0068]

[0069] Therefore, the computational unloading model for user u can be expressed as:

[0070]

[0071] in, and These represent user preferences for latency and energy consumption, respectively, where t represents time preference and e represents energy consumption preference. and In practical applications, and It can be set according to the battery level, and users can set different temple modes.

[0072] The main work of this patent is to jointly optimize video task offloading and resource allocation. For a given offloading decision χ and computational resource allocation F, this patent defines the system benefit as the weighted sum of the offloading utility of all users:

[0073]

[0074] λ u In (0,1), λ represents the resource provider's preference for user u. For example, in video tasks on mobile devices used for operations with high security requirements, λ can be used to represent the preference of the resource provider for user u. u Set the value higher. Then, the overall system model is:

[0075]

[0076] In the above formula, constraints (16a) and (16b) indicate that each task can either be executed locally or can be offloaded to a server in a subband at most; constraints (16c) and (16d) stipulate that each MEC server must allocate a positive computing resource to each user associated with it, and the total computing resources allocated to all associated users cannot exceed the server's computing power; constraint (16e) is for scalable video transmission, avoiding receiving only upper-layer data without receiving lower-layer data.

[0077] Step 5: Based on the joint unloading and resource optimization model proposed in Step 4, a low-complexity and suboptimal solution is proposed.

[0078] This paper attempts to find a low-complexity, suboptimal solution to the joint unloading and resource optimization model proposed in (16). Since this joint optimization problem is a mixed-integer nonlinear programming (MINLP), finding its optimal solution typically requires exponential time complexity. Based on the problem modeling in (16), a temporary binary variable {x} can be fixed. um Problem (16) is decomposed into subproblems with a separate objective function and multiple constraints. Thus, the original high-complexity problem can be transformed into a main problem and a set of low-complexity subproblems, and (16) can be re-expressed as...

[0079]

[0080] Since the Task Offloading (TO) and Computing Resource Allocation (CRA) problems are decoupled, solving the problem in (17) is equivalent to solving the offloading problem:

[0081]

[0082] Among them, J * (χ) is the optimal solution to the resource allocation problem, which is...

[0083]

[0084] Step 6: Utilize KKT conditions to determine the optimal resource allocation strategy for the wireless network.

[0085] First, given a feasible task unloading decision χ that satisfies the objective function, the objective function in (15) can be rewritten as:

[0086]

[0087] in,

[0088]

[0089] It can be seen that the first term of equation (20) is a constant for this system. Therefore, V(χ,F) is equivalent to the total uninstallation cost of all uninstalling users. That is, the above problem can be transformed into minimizing V(χ,F). Substituting (9) and (10) into (21), we get:

[0090]

[0091] in, as well as Currently only for f um To optimize, fix p u Then the calculation of resource allocation can be expressed by only the second term of equation (22) as follows:

[0092]

[0093] It can be seen that the Hessian matrix of the objective function is positive definite, therefore (23) is a convex optimization problem, which can be solved using the KKT conditions, then:

[0094]

[0095]

[0096] Step 7: Jointly design task offloading and resource allocation strategies, and finally obtain the optimal decision for task offloading and resource allocation of video terminals.

[0097] In the previous section, a computational resource optimization scheme was given. Combining with (20), we can obtain

[0098]

[0099] The task offloading joint resource allocation model constructed in this paper can be expressed as follows:

[0100]

[0101] As can be seen, the above TO problem is combinable. A simple way to solve this problem is to use an exhaustive search method for all possible task unloading decisions. However, when n = M * U * N, the complexity of the task unloading decision will be 2^n.

[0102] To overcome the above shortcomings, this patent employs a low-complexity annealing simulation algorithm that can find a local optimum of problem (27) in polynomial time. Specifically, this algorithm starts with an empty set, and if it improves set J... * If (χ), then a local operation, either a deletion or a swap operation, is repeated. Since this patent deals with two matroid constraints, the swap operation involves adding an element from outside the current set and deleting at most two elements from the set to satisfy the constraints.

[0103] Example 2

[0104] Secondly, this embodiment provides an edge computing resource scheduling optimization device for scalable video, including a processor and a storage medium;

[0105] The storage medium is used to store instructions;

[0106] The processor is configured to operate according to the instructions to perform the steps of the method according to Embodiment 1.

[0107] Example 3

[0108] Thirdly, this embodiment provides a storage medium on which a computer program is stored, which, when executed by a processor, implements the steps of the method described in Embodiment 1.

[0109] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0110] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0111] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0112] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0113] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for optimizing edge computing resource scheduling for scalable video, characterized in that, include: S1. Set the user's video task parameters, model the user, and construct the latency of the user executing tasks locally. and energy consumption The computational model; S2. Based on the computing power of the video terminal, construct the signal-to-noise ratio for task offloading. and the resulting uplink latency The computational model; S3. Configure MEC server parameters and build video tasks. Execution time on the MEC server The computational model; S4, based on , , and The computational model is used to construct a computational offloading model for edge computing systems oriented towards scalable video coding technology. ; S5, Based on computational unloading model Construct a joint unloading and resource optimization model, including: ； in, For the unloading decision, F is the resource allocation strategy; This indicates that the resource provider is to the user Preferences; Indicates user via sub-band The video's first Layer unloaded to server superior; Indicates task From the user via sub-band The video's first Layer offloaded to server superior, Conversely; For server The task From the user The amount of computing resources to be unloaded. Indicates server Computing resources; Indicates user via sub-band The video's first Layered server The data transmission rate; M is the set of MEC servers, N is the set of subbands available for each server, and U is the set of users in the mobile video system; S6. Solve the joint unloading and resource optimization model using KKT conditions and annealing simulation algorithm to obtain the optimized resource allocation strategy. and uninstallation decision ,include: S6.1, System Benefits Defined as the weighted sum of uninstall utilities for all users: ； By fixing the binary variables The joint unloading and resource optimization model is decomposed into subproblems with separate objective functions and multiple constraints, as follows: ； The unloading decision and resource allocation problems are decoupled, which is equivalent to solving the unloading decision problem: ； in, The optimal solution to the resource allocation problem is: ； S6.2 Utilize KKT conditions to determine the optimal wireless network resource allocation strategy ; For a given task unloading decision that satisfies the objective function The objective function is rewritten as: ； in, , and These represent the user's preferences for latency and energy consumption, respectively. For users The total delay experienced when unloading the task. For users Energy consumption of the uplink for processing tasks; The objective function is transformed into Minimization problem ； in, Indicates user The power required to transmit tasks to the server. , Indicates user To sub-belt Go to server Signal-to-noise ratio; , as well as , Indicates user Local computing power User The amount of input data for uplink video transmission tasks, where W represents the size of the subband; right Optimize and fix The resource allocation problem can be represented as: ； Using the KKT conditions, the optimal strategy for computational resource scheduling in wireless transmission networks is found. : ； ； S6.

3. Jointly design task offloading and resource allocation strategies, and finally obtain the optimal decision for task offloading and resource allocation of video terminals; ； The constructed task offloading joint resource allocation model is represented as follows: ； The annealing simulation algorithm is used to solve the problem, and the final offloading decision for each video task in each device under resource allocation constraints is obtained. .

2. The edge computing resource scheduling optimization method for scalable video according to claim 1, characterized in that, S1 includes: Assume each user There is a video task , The set of MEC servers represents the users in a mobile video system. Each video task The feature is determined by two parameters The tuples formed, in which User The amount of input data for uplink video transmission tasks. Workload refers to the amount of computation required to complete a task; user Latency of executing tasks locally : ； in Indicates user Local computing power; user Perform tasks locally Energy consumption per hour : ； in, This represents the energy coefficient.

3. The edge computing resource scheduling optimization method for scalable video according to claim 1, characterized in that, S2 includes: user To sub-belt Go to server signal-to-noise ratio for: ； in, Indicates user The power required to transmit tasks to the server. Indicates server With users Channel gain coefficient for transmission between them For the background noise variance, Indicates user via sub-band The video's first Layer unloaded to server superior, Indicates user The power required to transmit tasks to the server. Indicates server With users Channel gain coefficient for transmission between them The set of MEC servers represents the users in a mobile video system. The set of subbands available to each server is: .

4. The edge computing resource scheduling optimization method for scalable video according to claim 3, characterized in that, S2 further includes: user Input data volume for uplink video transmission tasks The resulting delay for: ； in, ; Indicates user via sub-band The video's first Layer unloaded to server Above, users to server Data transmission rate Represented as: ； in, , Indicates user To sub-belt Go to server Signal-to-noise ratio; For the decision of unloading, It satisfies the following constraints: 。 5. The edge computing resource scheduling optimization method for scalable video according to claim 2, characterized in that, S3 includes: Computing resource allocation strategy , For server The task From the user The amount of computing resources to be unloaded. ; The computing resource allocation strategy must meet the computing resource constraints of the MEC server: in Indicates server MEC server computing resources; Video Task Execution time on the MEC server for: , , Indicates user via sub-band The video's first Layer unloaded to server superior.

6. The edge computing resource scheduling optimization method for scalable video according to claim 1, characterized in that, S4 includes: user Calculation unloading model : ； in, and These represent the user's preferences for latency and energy consumption, respectively. Indicates time preference, This indicates energy consumption preference, while ,and ; Task completion time for: ； in For users Latency of executing tasks locally; user Total delay experienced when unloading tasks for: ; Energy consumption for: ； in For users Perform tasks locally Energy consumption per hour; user Energy consumption of the uplink for processing tasks for: ； in, Indicates user The power required to transmit tasks to the server. Indicates user Input data volume for uplink video transmission tasks The resulting delay; This refers to the user. The power amplifier efficiency.

7. An edge computing resource scheduling optimization device for scalable video, characterized in that, Including processor and storage media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1 to 6.

8. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images