Service delay optimization method, device, equipment and storage medium

Abstract

The present invention relates to the field of communication technology, and discloses a service delay optimization method, device, equipment and storage medium. The method comprises: when network congestion is detected, obtaining the unloading factor of various application services transmitted uplink to a base station, the various application services comprising delay-sensitive computing power application services, delay-sensitive non-computing power application services and non-delay-sensitive application services, calculating the minimum guaranteed bandwidth of various application services through a preset optimization model according to the unloading factor, and optimizing the service delay of various application services according to the minimum guaranteed bandwidth; because the present invention takes into account the delay requirements of delay-sensitive computing power application services and delay-sensitive non-computing power application services at the same time when optimizing the service delay, it can meet the delay requirements of delay-sensitive computing power application services and optimize the delay of computing services.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a service delay optimization method, device, equipment and storage medium. Background Art

[0002] At present, the service demand under 5G network is gradually increasing, and application services such as low latency and large bandwidth are emerging in an endless stream. In order to meet the existing service needs, the current wireless resource scheduling often adopts the low-latency queuing (LLQ) method, which is mainly a combination of priority queuing (PQ) (or strict priority (SP)) scheduling and class-based weighted fair queuing (CBWFQ) scheduling. For each class, there is a certain bandwidth guarantee value, which may be a proportion or a fixed value. When congestion occurs, control is performed at this time, and the high priority queue cannot exceed its bandwidth allocation value. When there is no congestion, no control is performed, and the high priority will obtain a bandwidth exceeding the fixed value.

[0003] LLQ is a more advanced Quality of Service (QoS) scheduling algorithm that can provide one or more strict priority queues with bandwidth guarantees for traffic of delay- and jitter-sensitive applications. The biggest advantage of LLQ is that it can ensure that high-priority traffic is processed in a timely manner, thereby ensuring the responsiveness and reliability of sensitive applications.

[0004] However, the LLQ algorithm in the related technology focuses on providing dedicated queues for delay-sensitive non-computing application services such as voice to reduce the air interface delay of such applications. It has the defect of not considering the delay requirements of delay-sensitive computing application services. Summary of the invention

[0005] The main purpose of the present invention is to provide a service delay optimization method, device, equipment and storage medium, aiming to solve the technical problem that the LLQ algorithm in the related technology focuses on providing dedicated queues for delay-sensitive non-computing application services such as voice, reducing the air interface delay of this type of application, and has the defect of not considering the delay requirements of delay-sensitive computing application services.

[0006] To achieve the above object, the present invention provides a service delay optimization method, the service delay optimization method comprising:

[0007] When network congestion is detected, obtaining unloading factors of various application services transmitted uplink to the base station, the various application services including delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services, the unloading factors are used to represent parameters for unloading services to the base station and user equipment for processing;

[0008] The minimum guaranteed bandwidth of each type of application service is calculated according to the offloading factor through a preset optimization model, and the preset optimization model is used to calculate the minimum guaranteed bandwidth of each type of application service according to the offloading factor under the condition that the priority of the delay-sensitive computing power application service is higher than the delay-sensitive non-computing power application service, and the task processing delay of the delay-sensitive computing power application service is not greater than the task processing delay of the delay-sensitive non-computing power application service;

[0009] The service delay of each type of application service is optimized according to the minimum guaranteed bandwidth.

[0010] Optionally, before calculating the minimum guaranteed bandwidth of each type of application service by using a preset optimization model according to the offloading factor, the method further includes:

[0011] Calculate the uplink transmission delay of tasks for latency-sensitive computing application services and latency-sensitive non-computing application services;

[0012] Constructing a computing power application service delay optimization algorithm according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service;

[0013] A preset optimization model is constructed based on the computing power application service delay optimization algorithm, and the preset optimization model is a neural network model.

[0014] Optionally, the calculation of the task uplink transmission delay of the latency-sensitive computing power application service includes:

[0015] Calculate the processing delay of the local offload task of the delay-sensitive computing power application service according to the offload factor of the delay-sensitive computing power application service, the task length, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task;

[0016] Calculate the transmission delay of the uplink offload task of the delay-sensitive computing power application service according to the offload factor of the delay-sensitive computing power application service, the task length, the signal to interference plus signal to noise ratio, the minimum guaranteed bandwidth, the computing power processing capability of the base station, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task;

[0017] The processing delay of the local offload task is compared with the transmission delay of the uplink offload task, and the task uplink transmission delay of the delay-sensitive computing power application service is determined according to the comparison result.

[0018] Optionally, the calculation of the task uplink transmission delay of the latency-sensitive non-computing power application service includes:

[0019] Obtain the task length, signal-to-interference plus signal-to-noise ratio, and minimum guaranteed bandwidth of latency-sensitive non-computing application services;

[0020] The task uplink transmission delay of the delay-sensitive non-computing power application service is calculated according to the task length, signal-to-interference-plus-signal-to-noise ratio and minimum guaranteed bandwidth of the delay-sensitive non-computing power application service.

[0021] Optionally, constructing a computing power application service delay optimization algorithm according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service includes:

[0022] On the premise that the transmission power and channel conditions of various types of application services are consistent and the minimum guaranteed bandwidth of similar services is the same, a computing power application service delay optimization algorithm is constructed according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service.

[0023] Optionally, when network congestion is detected, obtaining offloading factors of various application services includes:

[0024] When network congestion is detected, the computing processing capacity of the base station and the processing capacity of the user equipment are obtained;

[0025] The offloading factors of various application services are obtained according to the computing processing capability of the base station and the processing capability of the user equipment.

[0026] Optionally, before obtaining the offloading factors of various application services transmitted uplink to the base station when network congestion is detected, the method further includes:

[0027] Collect user data of various application services transmitted uplink to the base station in real time and pre-process the user data;

[0028] The service demand is determined according to the preprocessed user data, and whether the network is congested is detected according to the service demand and the currently used bandwidth.

[0029] In addition, to achieve the above-mentioned purpose, the present invention further proposes a service delay optimization device, the service delay optimization device comprising:

[0030] An acquisition module, configured to acquire, when network congestion is detected, an unloading factor of various application services transmitted uplink to a base station, wherein the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services, and the unloading factor is used to represent a parameter for unloading the service to the base station and the user equipment for processing;

[0031] A calculation module, used to calculate the minimum guaranteed bandwidth of each type of application service through a preset optimization model according to the offloading factor, wherein the preset optimization model is used to calculate the minimum guaranteed bandwidth of each type of application service under the condition that the priority of the delay-sensitive computing power application service is higher than that of the delay-sensitive non-computing power application service, and the task processing delay of the delay-sensitive computing power application service is not greater than the task processing delay of the delay-sensitive non-computing power application service;

[0032] The optimization module is used to optimize the service delay of each type of application service according to the minimum guaranteed bandwidth.

[0033] In addition, to achieve the above-mentioned objectives, the present invention also proposes a service delay optimization device, which includes a memory, a processor, and a service delay optimization program stored in the memory and executable on the processor, and the service delay optimization program is configured to implement the service delay optimization method described above.

[0034] In addition, to achieve the above-mentioned purpose, the present invention further proposes a storage medium, on which a service delay optimization program is stored, and when the service delay optimization program is executed by a processor, the service delay optimization method as described above is implemented.

[0035] In the present invention, it is disclosed that when network congestion is detected, an offloading factor of various application services that are uplinked to a base station is obtained, and the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services. The offloading factor is used to represent parameters for offloading services to base stations and user equipment for processing. The minimum guaranteed bandwidth of various application services is calculated according to the offloading factor through a preset optimization model. The preset optimization model is used to calculate the minimum guaranteed bandwidth of various application services according to the offloading factor under the condition that the priority of delay-sensitive computing power application services is higher than that of delay-sensitive non-computing power application services, and the task processing delay of delay-sensitive computing power application services is not greater than the task processing delay of delay-sensitive non-computing power application services. The service delay of various application services is optimized according to the minimum guaranteed bandwidth. Since the present invention takes into account the delay requirements of delay-sensitive computing power application services and delay-sensitive non-computing power application services when optimizing service delay, it is possible to meet the delay requirements of delay-sensitive computing power application services and optimize the delay of computing services. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a structural diagram of a service delay optimization device for a hardware operating environment involved in an embodiment of the present invention;

[0037] Figure 2 This is a flow chart of a first embodiment of a method for optimizing service delay according to the present invention;

[0038] Figure 3 A schematic diagram of various application services in an embodiment of a service delay optimization method of the present invention;

[0039] Figure 4 This is a flow chart of a second embodiment of the service delay optimization method of the present invention;

[0040] Figure 5 This is a flow chart of a third embodiment of the service delay optimization method of the present invention;

[0041] Figure 6 This is a structural block diagram of the first embodiment of the service delay optimization device of the present invention.

[0042] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0043] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0044] Reference Figure 1 , Figure 1 A schematic diagram of the structure of a service delay optimization device for a hardware operating environment involved in an embodiment of the present invention.

[0045] like Figure 1As shown, the service delay optimization device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The wired interface of the user interface 1003 may be a USB interface in the present invention. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) or a stable memory (Non-volatile Memory, NVM), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0046] Those skilled in the art will understand that Figure 1 The structure shown in does not constitute a limitation on the service delay optimization device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0047] like Figure 1 As shown, the memory 1005 identified as a computer storage medium may include an operating system, a network communication module, a user interface module, and a service delay optimization program.

[0048] exist Figure 1 In the service delay optimization device shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the user device; the service delay optimization device calls the service delay optimization program stored in the memory 1005 through the processor 1001, and executes the service delay optimization method provided in the embodiment of the present invention.

[0049] Based on the above hardware structure, an embodiment of the service delay optimization method of the present invention is proposed.

[0050] Reference Figure 2 , Figure 2 The flowchart of the first embodiment of the service delay optimization method of the present invention is provided.

[0051] It should be understood that, at present, the business demand under 5G network is gradually increasing, and application services such as low latency and large bandwidth are emerging in an endless stream. In order to meet the existing business needs, the current wireless resource scheduling often adopts the low-latency queue (Low-Latency Queuing, LLQ) method, which is mainly a combination of priority queue (Priority Queuing, PQ) (or strict priority (Strict Priority, SP)) scheduling and class-based weighted fair queue (Class-Based Weighted Fair Queuing, CBWFQ) scheduling. For each class, there is a certain bandwidth guarantee value, which may be a proportion or a fixed value. When congestion occurs, control is carried out at this time, and the high priority queue cannot exceed its bandwidth allocation value. When there is no congestion, no control is carried out, and the high priority will obtain a bandwidth exceeding the fixed value.

[0052] LLQ is a more advanced Quality of Service (QoS) scheduling algorithm that can provide one or more strict priority queues with bandwidth guarantees for traffic of delay- and jitter-sensitive applications. The biggest advantage of LLQ is that it can ensure that high-priority traffic is processed in a timely manner, thereby ensuring the responsiveness and reliability of sensitive applications.

[0053] The LLQ algorithm has the following advantages: 1. It can provide real-time priority control: LLQ can provide one or more high-priority dedicated queues with guaranteed bandwidth for delay- and jitter-sensitive applications, thereby ensuring that high-priority traffic is processed in a timely manner. 2. More flexible configuration options: LLQ allows the application of WRED (Random Early Detection) on each queue, allowing traffic to use bandwidth resources more efficiently. 3. It can be applied to a wide range of network platforms and transmission media: LLQ is not limited to specific platforms or transmission media and can be applied to various network environments and network devices.

[0054] However, the LLQ algorithm has the following disadvantages: 1. It is only applicable to delay- and jitter-sensitive applications: The LLQ algorithm focuses on providing dedicated queues for delay-sensitive applications such as voice to reduce the air interface delay of such applications, but other types of delay-sensitive application services are not considered. 2. For other types of delay-sensitive services, the user transmission delay is not considered from the perspective of the overall uplink transmission and calculation of the service. For example, for the service requirements of delay-sensitive computing power applications (XR, games), for the user's QoE, the user-perceived delay mainly includes the uplink air interface transmission delay and computing power service calculation delay, as well as the downlink transmission delay. Therefore, for similar delay-sensitive application services, optimization only from the perspective of air interface delay is not comprehensive. 3. For the delay optimization of the entire system, the air interface delay and service processing delay are not completely in the same relationship. When the transmission service volume reaches a certain threshold, the user can start the terminal's own processing capability to reduce the packet loss and delay caused by excessive transmission volume.

[0055] This embodiment innovatively proposes a latency optimization algorithm for latency-sensitive computing application services based on traditional low-latency queues, and utilizes a variety of unloading methods of user terminals to optimize the latency of computing services.

[0056] In this embodiment, the service delay optimization method includes:

[0057] Step S10: When network congestion is detected, obtain the unloading factors of various application services transmitted uplink to the base station, wherein the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services. The unloading factors are used to represent parameters for offloading services to the base station and user equipment for processing.

[0058] It can be understood that the executor of this embodiment can be a service delay optimization device with data processing, network communication and program running functions, such as a base station or an edge server, or other electronic devices that can achieve the same or similar functions. This embodiment does not impose any restrictions on this.

[0059] For ease of understanding, refer to Figure 3 This is for illustration only, but not for limitation of the present invention. Figure 3 This is a schematic diagram of various types of application services in an embodiment of the service delay optimization method of the present invention. In the figure, in a certain cell environment, multiple types of application services are uplinked to the base station. It is assumed that the application services include: delay-sensitive computing power application services (such as XR, game services, etc.), delay-sensitive non-computing power application services (video services, voice services, etc.) and non-delay-sensitive application services (cache services, etc.), which are defined here as the following three types of application services: type 1, type 2 and type 3.

[0060] Among them, each user equipment (User Equipment, UE) adopts partial offloading unloading mode, that is, each user's task can be partially offloaded to its own CPU for local processing, and the other part of the task can be offloaded to the base station for processing through the offloading unit (Offloading Unit, OU). The offloading factor is used to represent the parameters for offloading the service to the base station and the user equipment for processing.

[0061] Furthermore, in order to accurately detect whether the network is congested, before step S10, it also includes: collecting user data of various application services that are uplinked to the base station in real time, and preprocessing the user data; determining the service demand according to the preprocessed user data, and detecting whether the network is congested according to the service demand and the currently used bandwidth. The specific steps may be to collect user data of various application services that are uplinked to the base station in real time through a centralized unit (CU) and a distributed unit (DU), and preprocess the data, determine the service demand according to the preprocessed user data, and detect whether the network is congested according to the service demand and the currently used bandwidth, wherein the preprocessing may be a processing method such as data cleaning, which is not limited in this embodiment.

[0062] Step S20: Calculate the minimum guaranteed bandwidth of each type of application service according to the unloading factor through a preset optimization model. The preset optimization model is used to calculate the minimum guaranteed bandwidth of each type of application service according to the unloading factor under the condition that the priority of delay-sensitive computing power application service is higher than that of delay-sensitive non-computing power application service, and the task processing delay of delay-sensitive computing power application service is not greater than the task processing delay of delay-sensitive non-computing power application service.

[0063] It should be understood that the preset optimization model can obtain the minimum guaranteed bandwidth according to the unloading factor, or vice versa, the user's unloading factor can also be obtained according to the minimum guaranteed bandwidth. Therefore, as long as the above formula is satisfied, the minimum guaranteed bandwidth of various application services can be calculated according to the unloading factor, while ensuring that the priority of delay-sensitive computing power application services is higher than that of delay-sensitive non-computing power application services, and that the task processing delay of delay-sensitive computing power application services is not greater than the task processing delay of delay-sensitive non-computing power application services, thereby improving the delay processing capability of computing power services.

[0064] Step S30: Optimizing the service delay of each type of application service according to the minimum guaranteed bandwidth.

[0065] It is understandable that optimizing the service delay of each application service according to the minimum guaranteed bandwidth may be adjusting the bandwidth of uplink transmission of each application service to the base station according to the minimum guaranteed bandwidth to optimize the service delay of each application service.

[0066] In this embodiment, this type of queue delay optimization algorithm is applicable to latency-sensitive computing services to ensure the latency requirements of the computing services.

[0067] In this embodiment, the processing capability of the actual user itself is coordinated with the processing capability of the base station (or edge server), and an offloading factor is introduced to reduce the processing delay of the user task.

[0068] In this embodiment, based on the LLQ algorithm, a minimum bandwidth guarantee for delay-sensitive computing services and other services is proposed, and a computing application service delay optimization algorithm based on LLQ is proposed to ensure that the task processing and unloading delays of delay-sensitive computing services in congestion scenarios are minimized.

[0069] In this embodiment, it is disclosed that when network congestion is detected, an offloading factor of various application services that are uplinked to a base station is obtained, and the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services. The offloading factor is used to represent parameters for offloading services to base stations and user equipment for processing. The minimum guaranteed bandwidth of various application services is calculated according to the offloading factor through a preset optimization model. The preset optimization model is used to calculate the minimum guaranteed bandwidth of various application services according to the offloading factor under the condition that the priority of delay-sensitive computing power application services is higher than that of delay-sensitive non-computing power application services, and the task processing delay of delay-sensitive computing power application services is not greater than the task processing delay of delay-sensitive non-computing power application services. The service delay of various application services is optimized according to the minimum guaranteed bandwidth. Since this embodiment takes into account the delay requirements of delay-sensitive computing power application services and delay-sensitive non-computing power application services when optimizing service delay, it is possible to meet the delay requirements of delay-sensitive computing power application services and optimize the delay of computing services.

[0070] Reference Figure 4 , Figure 4 This is a flow chart of the second embodiment of the service delay optimization method of the present invention. Figure 2 The first embodiment shown provides a second embodiment of the service delay optimization method of the present invention.

[0071] In the second embodiment, before step S10, the method further includes:

[0072] Step S01: Calculate the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service.

[0073] It should be understood that in this embodiment, a computing power application service delay optimization algorithm is first constructed based on the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service, and then a preset optimization model is constructed based on the computing power application service delay optimization algorithm, thereby being able to pre-construct an accurate and reliable preset optimization model, thereby facilitating subsequent service delay optimization.

[0074] Furthermore, in order to more accurately calculate the task uplink transmission delay of the delay-sensitive computing power application service, the calculation of the task uplink transmission delay of the delay-sensitive computing power application service includes: calculating the processing delay of the local unloading task of the delay-sensitive computing power application service according to the unloading factor of the delay-sensitive computing power application service, the task length, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task; calculating the transmission delay of the uplink unloading task of the delay-sensitive computing power application service according to the unloading factor of the delay-sensitive computing power application service, the task length, the signal to interference plus signal-to-noise ratio, the minimum guaranteed bandwidth, the computing power processing capability of the base station, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task; comparing the processing delay of the local unloading task with the transmission delay of the uplink unloading task, and determining the task uplink transmission delay of the delay-sensitive computing power application service according to the comparison result.

[0075] Furthermore, in order to more accurately calculate the task uplink transmission delay of delay-sensitive non-computing power application services, the calculation of the task uplink transmission delay of delay-sensitive non-computing power application services includes: obtaining the task length, signal to interference plus signal to noise ratio and minimum guaranteed bandwidth of the delay-sensitive non-computing power application services; calculating the task uplink transmission delay of the delay-sensitive non-computing power application services according to the task length, signal to interference plus signal to noise ratio and minimum guaranteed bandwidth of the delay-sensitive non-computing power application services.

[0076] Step S02: construct a computing power application service delay optimization algorithm according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service.

[0077] It can be understood that constructing a computing power application service delay optimization algorithm based on the task uplink transmission delay of delay-sensitive computing power application services and the task uplink transmission delay of delay-sensitive non-computing power application services can be based on the assumption that the transmission power and channel conditions of various application services are consistent and the minimum guaranteed bandwidth of the same type of services is the same.

[0078] Step S03: construct a preset optimization model based on the computing power application service delay optimization algorithm, and the preset optimization model is a neural network model.

[0079] For ease of understanding, refer to Figure 3 This is for illustration only, but not for limitation of the present invention. Figure 3 This is a schematic diagram of various application services in an embodiment of a service delay optimization method of the present invention. In the figure, building a preset optimization model includes the following steps:

[0080] 1. System modeling:

[0081] In a certain cell environment, multiple types of application services are uplinked to the base station. It is assumed that the application services include: delay-sensitive computing application services (such as XR, gaming services, etc.), delay-sensitive non-computing application services (video services, voice services, etc.) and non-delay-sensitive application services (cache services, etc.), which are defined here as the following three types of application services: type 1, type 2 and type 3.

[0082] Among them, each UE adopts partial offloading unloading mode, that is, each user's task can be partially offloaded to its own CPU for local processing, and the other part of the task can be offloaded to the base station for processing through the offloading unit OU. The offloading factor is used to represent the parameters for offloading the service to the base station and user equipment for processing.

[0083] Assume that Type 1 has N latency-sensitive computing applications, Type 2 has M latency-sensitive other applications, and Type 3 has X latency-insensitive applications. Assume that n∈N, and service n occupies bandwidth b. n ,Assume m∈M, the bandwidth occupied by service m is b m ,Assume x∈X, the bandwidth occupied by service x is b x Therefore, there is

[0084] Based on the low-latency queue rules, it is assumed that both type 1 and type 2 low-latency services have a priority queue with a minimum guaranteed bandwidth. When congested, the data volume of this queue cannot exceed the allowed bandwidth, otherwise it will be discarded. Assuming that when the services in this scenario increase, congestion occurs. At this time, the priority of X non-latency sensitive application services is the lowest and there is no minimum guaranteed bandwidth. Therefore, the occupied bandwidth at this moment is 0. Assuming that in this congested environment, the priority of latency-sensitive computing power application services is greater than the priority of latency-sensitive other types of application services, and the minimum guaranteed bandwidth of each service is set as follows: b n，1 ,n∈N,b m,1 ,m∈M, and when congested, there is

[0085] 2. Problem establishment and solution:

[0086] Since the priority of N delay-sensitive computing power application services is higher than that of M delay-sensitive other types of application services, the delay of any service n (n∈N) of the former should be higher than the delay of any service m (m∈M). The uplink transmission delay of tasks for delay-sensitive computing power services includes: the processing delay of local offload tasks, the transmission delay of uplink offload tasks, and the calculation delay. For delay-sensitive other types of services, there is only uplink transmission delay in the uplink. Therefore, there is:

[0087] T n ≤T m ,n∈N,m∈M

[0088] Where, T n is the delay for service n (n∈N), T m is the delay for service m (m∈M).

[0089] The latency of service n can be written as:

[0090] T n =max{T n,loc ,T n,off}

[0091] Where, T n,loc is the processing delay of the local offload task for service n (n∈N), T n,off is the transmission delay of the uplink offload task for service n (n∈N).

[0092] Assume that the task lengths of business n and business m are L respectively. n (bits),L m (bits), assuming the processing power of the user's local CPU is f ue (cycles / s), let C n represents the number of CPU cycles / bit required by user n to process each bit task. Let the computing power of the base station be q BS (cycles / s), let the user's uninstall factor for service n be β n ∈[0,1]. Therefore, the delay for service n can be expanded as:

[0093]

[0094]

[0095] The latency of service m can be written as:

[0096]

[0097] In summary, we can get:

[0098] T m ≤T m ,n∈N,m∈M

[0099] max{T n,loc ,T n,off}≤T m ,n∈N,m∈M

[0100]

[0101] Therefore, the above function can be changed to:

[0102] F(β n ,b n,1 ,b m,1 )≤0

[0103]

[0104] In addition, the constraints include:

[0105] 0≤β n ≤1

[0106]

[0107] Assuming that the transmission power and channel conditions of each service are consistent, the SINR n =SINR m At the same time, according to the agreement of the LLQ algorithm, the minimum guaranteed bandwidth of the same type of services is the same, so there is:

[0108]

[0109] At this time, the function F can be transformed into:

[0110]

[0111] At this point, the function F can be converted into the following form:

[0112]

[0113]

[0114] For the above function, function F and variable β n There is a linear relationship between them and the variable b n,1 There is an inverse function relationship between them, and both variables have upper and lower limits, so the above function F always has a solution.

[0115] Therefore, in summary, in actual application, the minimum guaranteed bandwidth can be obtained according to the actual user offloading parameters, or vice versa, the user's offloading factor can be obtained. Therefore, as long as the above formula is satisfied, the priority of service n can be higher than that of service m, and the task processing delay of service n can be guaranteed to be less than or equal to the task processing delay of service m, thereby improving the delay processing capability of computing power services.

[0116] Similarly, the above function F can be further expanded and optimized, such as minimizing the transmission and unloading delay of service n, etc. This proposal will not repeat it. This type of algorithm is also applicable to other types of task queues to ensure the delay requirements of any queue.

[0117] It should be understood that building a preset optimization model based on the computing power application business delay optimization algorithm can be to start the corresponding process of the computing power application business delay optimization algorithm to build a neural network model, conduct network model training in a timely manner, apply the training results to access users, and track user demand feedback in real time, adjust model parameters, and further optimize the model.

[0118] This embodiment first constructs a computing power application service delay optimization algorithm based on the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service, and then constructs a preset optimization model based on the computing power application service delay optimization algorithm, thereby being able to pre-build an accurate and reliable preset optimization model, thereby facilitating subsequent service delay optimization.

[0119] Reference Figure 5 , Figure 5 This is a flow chart of a third embodiment of a method for optimizing service delay of the present invention. Based on the above embodiments, a third embodiment of a method for optimizing service delay of the present invention is proposed.

[0120] In the third embodiment, the step S10 includes:

[0121] Step S101: When network congestion is detected, the computing processing capability of the base station and the processing capability of the user equipment are obtained.

[0122] It should be understood that in this embodiment, by obtaining the offloading factors of various application services using the computing processing capabilities of the base station and the processing capabilities of the user equipment, the accuracy of the offloading factors can be improved, thereby improving the effect of service delay optimization.

[0123] Step S102: Obtaining offloading factors of various application services according to the computing processing capability of the base station and the processing capability of the user equipment.

[0124] It can be understood that obtaining the unloading factor of various application services based on the computing power processing capability of the base station and the processing capability of the user equipment can be to compare the computing power processing capability of the base station with the processing capability of the user equipment, and obtain the unloading factor of various application services based on the comparison result.

[0125] This embodiment obtains the offloading factors of various application services by using the computing processing capability of the base station and the processing capability of the user equipment, thereby improving the accuracy of the offloading factors and further improving the effect of service delay optimization.

[0126] In addition, refer to Figure 6 The embodiment of the present invention further provides a service delay optimization device, the service delay optimization device comprising:

[0127] The acquisition module 10 is used to acquire the unloading factors of various application services transmitted uplink to the base station when network congestion is detected, wherein the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services, and the unloading factors are used to represent parameters for unloading services to the base station and the user equipment for processing;

[0128] A calculation module 20, configured to calculate the minimum guaranteed bandwidth of each type of application service through a preset optimization model according to the offloading factor, wherein the preset optimization model is configured to calculate the minimum guaranteed bandwidth of each type of application service under the condition that the priority of the delay-sensitive computing power application service is higher than that of the delay-sensitive non-computing power application service, and the task processing delay of the delay-sensitive computing power application service is not greater than the task processing delay of the delay-sensitive non-computing power application service;

[0129] The optimization module 30 is used to optimize the service delay of each type of application service according to the minimum guaranteed bandwidth.

[0130] In this embodiment, it is disclosed that when network congestion is detected, an offloading factor of various application services that are uplinked to a base station is obtained, and the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services. The offloading factor is used to represent parameters for offloading services to base stations and user equipment for processing. The minimum guaranteed bandwidth of various application services is calculated according to the offloading factor through a preset optimization model. The preset optimization model is used to calculate the minimum guaranteed bandwidth of various application services according to the offloading factor under the condition that the priority of delay-sensitive computing power application services is higher than that of delay-sensitive non-computing power application services, and the task processing delay of delay-sensitive computing power application services is not greater than the task processing delay of delay-sensitive non-computing power application services. The service delay of various application services is optimized according to the minimum guaranteed bandwidth. Since this embodiment takes into account the delay requirements of delay-sensitive computing power application services and delay-sensitive non-computing power application services when optimizing service delay, it is possible to meet the delay requirements of delay-sensitive computing power application services and optimize the delay of computing services.

[0131] In one embodiment, the service delay optimization device further includes:

[0132] A training module is used to calculate the task uplink transmission delay of delay-sensitive computing power application services and the task uplink transmission delay of delay-sensitive non-computing power application services; to construct a computing power application service delay optimization algorithm based on the task uplink transmission delay of the delay-sensitive computing power application services and the task uplink transmission delay of the delay-sensitive non-computing power application services; and to construct a preset optimization model based on the computing power application service delay optimization algorithm, wherein the preset optimization model is a neural network model.

[0133] In one embodiment, the training module is also used to calculate the processing delay of the local unloading task of the delay-sensitive computing power application service according to the unloading factor of the delay-sensitive computing power application service, the task length, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task; calculate the transmission delay of the uplink unloading task of the delay-sensitive computing power application service according to the unloading factor of the delay-sensitive computing power application service, the task length, the signal to interference plus signal-to-noise ratio, the minimum guaranteed bandwidth, the computing power processing capability of the base station, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task; compare the processing delay of the local unloading task with the transmission delay of the uplink unloading task, and determine the task uplink transmission delay of the delay-sensitive computing power application service according to the comparison result.

[0134] In one embodiment, the training module is also used to obtain the task length, signal to interference plus signal to noise ratio and minimum guaranteed bandwidth of the delay-sensitive non-computing power application service; and calculate the task uplink transmission delay of the delay-sensitive non-computing power application service based on the task length, signal to interference plus signal to noise ratio and minimum guaranteed bandwidth of the delay-sensitive non-computing power application service.

[0135] In one embodiment, the training module is also used to construct a computing power application service delay optimization algorithm based on the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service, assuming that the transmission power and channel conditions of various application services are consistent and the minimum guaranteed bandwidth of similar services is the same.

[0136] In one embodiment, the acquisition module 10 is also used to obtain the computing power processing capability of the base station and the processing capability of the user equipment when network congestion is detected; and obtain the unloading factors of various application services based on the computing power processing capability of the base station and the processing capability of the user equipment.

[0137] In one embodiment, the service delay optimization device further includes:

[0138] The detection module is used to collect user data of various application services transmitted uplink to the base station in real time and pre-process the user data; determine the service demand according to the pre-processed user data, and detect whether the network is congested according to the service demand and the current bandwidth used.

[0139] Other embodiments or specific implementations of the service delay optimization device of the present invention can refer to the above-mentioned method embodiments and will not be described in detail here.

[0140] In addition, an embodiment of the present invention further proposes a storage medium, on which a service delay optimization program is stored. When the service delay optimization program is executed by a processor, the service delay optimization method as described above is implemented.

[0141] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0142] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0143] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory image (ROM) / random access memory (RAM), a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0144] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A service delay optimization method, characterized in that: The service delay optimization method comprises: When network congestion is detected, obtaining unloading factors of various application services transmitted uplink to the base station, the various application services including delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services, the unloading factors are used to represent parameters for unloading services to the base station and user equipment for processing; The minimum guaranteed bandwidth of each type of application service is calculated according to the offloading factor through a preset optimization model, and the preset optimization model is used to calculate the minimum guaranteed bandwidth of each type of application service according to the offloading factor under the condition that the priority of the delay-sensitive computing power application service is higher than the delay-sensitive non-computing power application service, and the task processing delay of the delay-sensitive computing power application service is not greater than the task processing delay of the delay-sensitive non-computing power application service; Optimizing the service delay of each type of application service according to the minimum guaranteed bandwidth; Wherein, when network congestion is detected, obtaining the unloading factors of various application services transmitted uplink to the base station includes: When network congestion is detected, the computing processing capacity of the base station and the processing capacity of the user equipment are obtained; The computing processing capability of the base station is compared with the processing capability of the user equipment, and the offloading factor of each application service is obtained based on the comparison result.

2. The service delay optimization method according to claim 1, characterized in that: Before calculating the minimum guaranteed bandwidth of each type of application service through a preset optimization model according to the offloading factor, the method further includes: Calculate the uplink transmission delay of tasks for latency-sensitive computing application services and latency-sensitive non-computing application services; Constructing a computing power application service delay optimization algorithm according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service; A preset optimization model is constructed based on the computing power application service delay optimization algorithm, and the preset optimization model is a neural network model.

3. The service delay optimization method according to claim 2, characterized in that: The calculation of the task uplink transmission delay of the latency-sensitive computing power application service includes: Calculate the processing delay of the local offload task of the delay-sensitive computing power application service according to the offload factor of the delay-sensitive computing power application service, the task length, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task; Calculate the transmission delay of the uplink offload task of the delay-sensitive computing power application service according to the offload factor of the delay-sensitive computing power application service, the task length, the signal to interference plus signal to noise ratio, the minimum guaranteed bandwidth, the computing power processing capability of the base station, the processing capability of the user equipment, and the number of CPU revolutions required to process each bit of the task; The processing delay of the local offload task is compared with the transmission delay of the uplink offload task, and the task uplink transmission delay of the delay-sensitive computing power application service is determined according to the comparison result.

4. The service delay optimization method according to claim 2, characterized in that: The task uplink transmission delay of the computing delay-sensitive non-computing power application service includes: Obtain the task length, signal-to-interference plus signal-to-noise ratio, and minimum guaranteed bandwidth of latency-sensitive non-computing application services; The task uplink transmission delay of the delay-sensitive non-computing power application service is calculated according to the task length, signal-to-interference-plus-signal-to-noise ratio and minimum guaranteed bandwidth of the delay-sensitive non-computing power application service.

5. The service delay optimization method according to claim 2, characterized in that: The constructing of a computing power application service delay optimization algorithm according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service comprises: On the premise that the transmission power and channel conditions of various types of application services are consistent and the minimum guaranteed bandwidth of similar services is the same, a computing power application service delay optimization algorithm is constructed according to the task uplink transmission delay of the delay-sensitive computing power application service and the task uplink transmission delay of the delay-sensitive non-computing power application service.

6. The service delay optimization method according to any one of claims 1 to 5, characterized in that: Before obtaining the offloading factors of various application services transmitted uplink to the base station when network congestion is detected, the method further includes: Collect user data of various application services transmitted uplink to the base station in real time and pre-process the user data; The service demand is determined according to the preprocessed user data, and whether the network is congested is detected according to the service demand and the currently used bandwidth.

7. A service delay optimization device, characterized in that: The service delay optimization device comprises: An acquisition module, configured to acquire, when network congestion is detected, an unloading factor of various application services transmitted uplink to a base station, wherein the various application services include delay-sensitive computing power application services, delay-sensitive non-computing power application services, and non-delay-sensitive application services, and the unloading factor is used to represent a parameter for unloading the service to the base station and the user equipment for processing; A calculation module, used to calculate the minimum guaranteed bandwidth of each type of application service through a preset optimization model according to the offloading factor, wherein the preset optimization model is used to calculate the minimum guaranteed bandwidth of each type of application service under the condition that the priority of the delay-sensitive computing power application service is higher than that of the delay-sensitive non-computing power application service, and the task processing delay of the delay-sensitive computing power application service is not greater than the task processing delay of the delay-sensitive non-computing power application service; An optimization module, configured to optimize the service delay of each type of application service according to the minimum guaranteed bandwidth; Among them, the acquisition module is also used to obtain the computing power processing capability of the base station and the processing capability of the user equipment when network congestion is detected; compare the computing power processing capability of the base station with the processing capability of the user equipment, and obtain the unloading factor of each type of application service based on the comparison result.

8. A service delay optimization device, characterized in that: The service delay optimization device includes: a memory, a processor, and a service delay optimization program stored in the memory and executable on the processor. When the service delay optimization program is executed by the processor, the service delay optimization method according to any one of claims 1 to 6 is implemented.

9. A storage medium, characterized in that: The storage medium stores a service delay optimization program, and when the service delay optimization program is executed by the processor, the service delay optimization method according to any one of claims 1 to 6 is implemented.

Images