Resource unit allocation method, device, equipment and storage medium

Abstract

The present invention discloses a resource unit allocation method, device, equipment and storage medium, the method comprising: collecting multiple status information of each client to obtain a data set; training a network model based on the data set to obtain an allocation model; wherein the allocation model comprises a first target network model and a second target network model; allocating resource units based on the first target network model and the second target network model and target status information to obtain an allocation result; allocating the resource units based on the allocation result. The technical solution of the present invention adopts a dual network to allocate resource units; wherein the first target network model is used to obtain data and performance-related characteristics of the client, and the second target network model is used to evaluate the final performance in combination with the current channel environment, so as to obtain a more reasonable resource unit allocation method.

Description

Technical Field

[0001] The present invention belongs to the technical field, and in particular relates to a resource unit allocation method, device, equipment and storage medium. Background Art

[0002] The 802.11ax protocol began to introduce Orthogonal Frequency Division Multiple Access (OFDMA) technology, in which the smallest unit of time and frequency is the Resource Unit (RU), that is, channels of various bandwidths are further divided into multiple RUs. In previous protocols, a designated channel can only be allocated to one user. In intensive use scenarios, other users can only wait for the previous user to complete transmission before competing for resources.

[0003] After the introduction of OFMDA, the channel is divided into multiple RUs and allocated to different users for simultaneous transmission to improve the communication experience in multi-user scenarios; with the development of the protocol, higher communication frequency bands and larger channel bandwidths are allowed to be used, so the RU allocation algorithm needs to consider more and more factors.

[0004] The commonly used allocation scheme at present is to maximize the evaluation index based on the number of packets sent by each client, channel state information (CSI) or remaining time. This scheme formulates an evaluation function according to the usage scenario, then traverses all possible allocation schemes and calculates the index gain, and finally selects the allocation scheme that can maximize the gain. However, the evaluation function used in this scheme generally only considers a specific indicator, such as throughput or remaining time of data packets, and fails to take into account other communication quality indicators such as quality of service (QoS); in addition, with the introduction of large bandwidths such as 160MHz or even 320MHz, the number of RUs that can be allocated has increased significantly, so the potential allocation schemes have increased significantly, and the complexity of the full traversal algorithm has also greatly increased, which is not convenient for implementation in real-time systems with limited computing resources. Summary of the invention

[0005] The present invention aims to solve one of the technical problems in the related art at least to a certain extent. To this end, one object of the present invention is to provide a resource unit allocation method, device, equipment and storage medium.

[0006] In order to solve the above technical problems, the embodiments of the present invention provide the following technical solutions:

[0007] A resource unit allocation method, comprising:

[0008] Collect multiple status information of each client to obtain each client in the data set;

[0009] Training the network model based on the data set to obtain an allocation model; wherein the allocation model includes a first target network model and a second target network model;

[0010] Allocate resource units based on the first target network model, the second target network model and the target state information to obtain an allocation result;

[0011] The resource units are allocated based on the allocation result.

[0012] Optionally, the step of obtaining a data set based on multiple pieces of status information of each client includes:

[0013] Acquire first status information, second status information, and third status information of each of the clients;

[0014] A data set is obtained based on the first state information, the second state information and the third state information.

[0015] Optionally, obtaining a data set based on the first state information, the second state information, and the third state information includes:

[0016] Acquire each scene corresponding to the first state information, the second state information, and the third state information;

[0017] Label each of the scenes to obtain assigned labels;

[0018] Based on the assigned labels, the data set is obtained.

[0019] Optionally, labeling each of the scenes to obtain assigned labels includes:

[0020] Based on the test, obtain the target allocation for each scenario; or

[0021] The target allocation method for each scenario is obtained by performing calculation based on the allocation gain index and the allocation gain of each allocation method.

[0022] Optionally, obtaining the allocation gain indicator of each client at each moment includes:

[0023] Obtaining a first parameter of each of the clients at each moment; the first parameter is positively correlated with the allocation gain index; obtaining a second parameter of each of the clients at each moment; the second parameter is negatively correlated with the allocation gain index;

[0024] The allocation gain index of each of the clients at each moment is obtained by performing calculation based on the first parameter and the second parameter.

[0025] Optionally, the training of the network model based on the data set to obtain the allocation model includes:

[0026] Preprocessing the data set to obtain a target data set;

[0027] Inputting the first state information and the second state information into a first network model, training the first network model to obtain a first target network model; and outputting an initial result based on the first target network model;

[0028] Inputting the initial result and the third state information into a second network model, training the second network model, and obtaining a second target network model;

[0029] The allocation model is obtained based on the first target network model and the second target network model.

[0030] Optionally, allocating resource units based on the first target network model, the second target network model, and the target state information to obtain an allocation result includes:

[0031] Acquire target state information; wherein the target state information includes first target state information, second target state information and third target state information;

[0032] Processing the first target state information and the second target state information based on the first target network model, and outputting a first result;

[0033] The first result and the third target state information are processed based on the second target network model, and the allocation result is output.

[0034] An embodiment of the present invention further provides a resource unit allocation device, comprising:

[0035] The acquisition module is used to collect multiple status information of each client and obtain each client of the data set;

[0036] A training module, used to train the network model based on the data set to obtain an allocation model; wherein the allocation model includes a first target network model and a second target network model;

[0037] A processing module, configured to allocate resource units based on the first target network model, the second target network model and the target state information to obtain an allocation result;

[0038] An allocation module is used to allocate the resource unit based on the allocation result.

[0039] An embodiment of the present invention further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the method described above when executing the computer program.

[0040] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method described above.

[0041] The embodiments of the present invention have the following technical effects:

[0042] The above-mentioned technical solution of the present invention, 1) adopts a dual network to allocate resource units; wherein, the first target network model is used to obtain the data and performance-related characteristics of the client, and the second target network model is used to evaluate the final performance in combination with the current channel environment. Combining the first target network model with the second target network model can more easily guide the relevant characteristics of the network to learn in stages, thereby obtaining a more reasonable resource unit allocation method.

[0043] 2) In addition to collecting data in actual scenarios, the data set can also be enhanced through relevant wireless simulation platforms to avoid the difficulties faced by large-scale collection work and reduce the complexity of collecting data.

[0044] 3) The reference index of allocation gain is added, which comprehensively considers parameters such as latency, QoS, and service fairness, takes into account both client service quality and wireless channel conditions, and can adjust the relevant weights according to the different focuses of actual scenarios, which can better meet the operation requirements of real-time systems.

[0045] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a schematic diagram of a flow chart of a resource unit allocation method provided by an embodiment of the present invention;

[0047] Figure 2 This is an example of the process of the resource unit allocation method provided by an embodiment of the present invention;

[0048] Figure 3 It is a structural diagram of a resource unit allocation device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0050] In order to facilitate the understanding of the embodiments by those skilled in the art, some terms are explained:

[0051] (1) BSR: Buffer Status Report, which refers to the buffer status report sent by each client in the embodiment of the present invention.

[0052] (2) BQR: Basic Qualification Review, available bandwidth query; in the embodiment of the present invention, it means the available bandwidth query of each client.

[0053] (3) RSSI: Received Signal Strength Indication, received signal strength indication.

[0054] (4) CQI: Channel Quality Indicator.

[0055] (5)NS-3: A brand new network simulator.

[0056] (6) SDR: Software Defined Radio.

[0057] At present, the solution of solving the target option by fully traversing the potential allocation schemes only considers a specific evaluation indicator, and does not involve the indicators that subsequent protocols focus on improving, such as QoS or delay parameters. Moreover, with the introduction of ultra-large bandwidth, the number of potential allocation schemes has increased significantly, and the complexity of the algorithm has also increased, which cannot meet the operation requirements of real-time systems well.

[0058] In order to solve the above-mentioned technical problems, the embodiments of the present invention provide the following technical solutions:

[0059] An embodiment of the present invention provides a resource unit allocation system, including:

[0060] Data collectors, model trainers, wireless communication simulation platforms, and distribution equipment;

[0061] Among them, the data collector, model trainer, wireless communication simulation platform and distribution equipment realize data interaction based on the network;

[0062] Specifically, the data collector collects multiple status information of multiple clients in multiple scenarios to obtain a data set;

[0063] In addition, in order to simulate more topologies or usage scenarios, more topologies or usage scenarios can be constructed based on the wireless communication simulation platform, and multiple state information of each constructed scenario can be collected based on the data collector, thereby obtaining more effective data sets;

[0064] Then, the data collector inputs the data set into the model trainer, and the model trainer trains the first network model based on the first state information and the second state information in the data set, and outputs an initial result; then the model trainer trains the second network model based on the initial result and the third state information;

[0065] Repeat the above steps multiple times to obtain the first target network model and the second target network model respectively;

[0066] After obtaining the first target network model and the second target network model, the first target network model and the second target network model can be actually deployed based on the allocation device, and then the target state information is processed based on the deployed first target network model and the second network target model, and the allocation result is output, so as to finally realize the allocation of resource units based on the allocation result.

[0067] like Figure 1 As shown, an embodiment of the present invention provides a resource unit allocation method, which is applied to the above system, including:

[0068] Step S1: collect multiple status information of each client to obtain a data set for each client;

[0069] Specifically, the step of obtaining a data set based on multiple status information of each client includes:

[0070] Acquire first status information, second status information, and third status information of each of the clients;

[0071] A data set is obtained based on the first state information, the second state information and the third state information.

[0072] For example, in an embodiment of the present invention, the first state information may be data state information; the second state information may be performance state information; the third state information may be channel state information;

[0073] Among them, 1) data status information includes but is not limited to the amount of cached data of each client (such as BSR, QoS, remaining life time information, etc.);

[0074] 2) Performance status information includes but is not limited to the sending rate of each client, available subchannels (such as BQR) and signal strength (such as RSSI of the last data packet);

[0075] 3) Channel state information includes but is not limited to the channel quality of each client (such as the CQI or CSI of each client), available channel resources or interference, etc.;

[0076] In an actual application scenario, a data set is obtained based on data status information, performance status information, and channel status information of each client in multiple scenarios.

[0077] It should be noted that the data set includes but is not limited to data status information, performance status information and channel status information. In actual application scenarios, other status information can also be collected based on actual needs. The embodiments of the present invention do not make specific limitations on this.

[0078] In an optional embodiment of the present invention, more topologies or usage scenarios may be simulated through a wireless communication simulation platform to reduce the difficulty of obtaining a data set or to enhance the data set;

[0079] For example: through NS-3 wireless communication simulation platform or other wireless SDR, etc.

[0080] In addition to collecting data in actual scenes, the embodiments of the present invention can also enhance the data set through a related wireless simulation platform, thereby avoiding the difficulties faced by a large amount of collection work and reducing the complexity of collecting data.

[0081] In an optional embodiment of the present invention, multiple evaluation indicators are combined in the process of collecting data sets, which can meet the operation requirements of real-time systems compared with existing technical solutions;

[0082] Specifically, obtaining a data set based on the first state information, the second state information, and the third state information includes:

[0083] Acquire each scene corresponding to the first state information, the second state information, and the third state information;

[0084] Label each of the scenes to obtain assigned labels;

[0085] Based on the assigned labels, the data set is obtained.

[0086] In an embodiment of the present invention, in order to realize the subsequent training of the first network model and the second network model by supervised learning, a distribution label is set for the focus in the actual application scenario; specifically, each scenario in the data set is provided with a corresponding distribution label, and the distribution label is used as a supervision label for the subsequent training of the first network model and the second network model by supervised learning;

[0087] Among them, the assigned labels need to be given in advance when obtaining the data set.

[0088] In actual application scenarios, there may be a variety of marking methods. The embodiments of the present invention take two of the marking methods as examples for explanation:

[0089] Specifically, labeling each of the scenes to obtain assigned labels includes:

[0090] Based on the test, obtain the target allocation for each scenario; or

[0091] The target allocation method for each scenario is obtained by performing calculation based on the allocation gain index and the allocation gain of each allocation method.

[0092] In actual application scenarios, while collecting multiple status information of multiple clients in multiple scenarios, the target allocation method of resource units in each scenario is tested based on the test method, and the test results are recorded:

[0093] For example, record the allocation method that achieves the highest overall throughput performance or the highest overall allocation fairness as an allocation label;

[0094] Alternatively, some algorithms with higher allocation efficiency are used to calculate the target allocation method for each scene, and the calculation results are used as the allocation label for each scene.

[0095] It should be noted that, in the embodiments of the present invention, the testing methods include but are not limited to the above two testing methods.

[0096] An optional embodiment of the present invention, since the above-mentioned test method may have the problem of complicated operation when obtaining allocation labels in actual scenarios or existing algorithms; or the target allocation method obtained by the existing algorithm is only a local target and has little actual reference value; it is also possible that the application scenario itself has a more autonomous focus, for example: paying more attention to optimizing QoS and real-time traffic performance, etc.

[0097] In order to optimize or improve the test method, an embodiment of the present invention provides another way to obtain the allocation label, that is, based on the allocation gain index and the allocation gain of each allocation method, the target allocation method of each scenario is obtained by calculation.

[0098] Further, obtaining the allocation gain index of each client at each moment includes:

[0099] Acquire a first parameter of each client at each moment; the first parameter is positively correlated with the allocation gain indicator; wherein the first parameter includes service quality, channel quality and transmission rate;

[0100] Obtaining a second parameter of each client at each moment; the second parameter is negatively correlated with the allocation gain indicator; wherein the second parameter includes a time window, a historical average rate, and a delay parameter;

[0101] The allocation gain index of each of the clients at each moment is obtained by performing calculation based on the first parameter and the second parameter.

[0102] For example, the allocation gain index of each client at each moment can be calculated based on the following calculation formula:

[0103]

[0104] Among them, S i(n) represents the allocation gain index of the i-th client at the n-th moment; δ i Indicates QoS service quality; t c is the time window, R i is the historical average rate of the i-th client before time n; T i is the delay parameter or the remaining life time; C i j is the channel quality at the jth RU; r i is the transmission rate; i, c and j are positive integers, and n is greater than 0.

[0105] Based on the above formula, it can be seen that the embodiment of the present invention takes into account the need to balance fairness and a certain service quality in resource unit allocation, avoiding problems such as some clients being unable to obtain services or QoS traffic being delayed; in addition, the allocation gain is positively correlated with the QoS level, and negatively correlated with the service duration and the remaining life time or delay parameters.

[0106] For example, the allocation gain index of each client at each moment can also be calculated based on the following calculation formula:

[0107]

[0108] Among them, α, β, γ, and λ are weights respectively.

[0109] In actual application scenarios, any one of α, β, γ, and λ can be adjusted in real time based on actual needs to change the weights of each item corresponding to α, β, γ, and λ in the above formula.

[0110] For example, if a server focuses on fairness, it can increase the weight of β, increase the proportion of the corresponding item of β, and obtain the assigned label based on this.

[0111] The embodiments of the present invention add reference indicators for allocating gain, comprehensively consider parameters such as latency, QoS, and service fairness, take into account client service quality and wireless channel conditions, and can adjust relevant weights according to different focuses of actual scenarios, thereby better meeting the operating requirements of real-time systems.

[0112] Specifically, the allocation gains obtained by different allocation methods can be completely traversed on the data set to obtain the target allocation method of resource units in each scenario; in addition, the embodiments of the present invention can achieve a balance between fairness and a certain service quality in the resource unit allocation process based on the adjustment of α, β, γ, and λ, thereby better meeting the operation requirements of the real-time system.

[0113] Step S2: training the network model based on the data set to obtain an allocation model; wherein the allocation model includes a first target network model and a second target network model;

[0114] Specifically, the training of the network model based on the data set to obtain the allocation model includes:

[0115] Preprocessing the data set to obtain a target data set;

[0116] Inputting the first state information and the second state information into a first network model, training the first network model to obtain a first target network model; and outputting an initial result based on the first target network model;

[0117] Inputting the initial result and the third state information into a second network model, training the second network model, and obtaining a second target network model;

[0118] The allocation model is obtained based on the first target network model and the second target network model.

[0119] In an embodiment of the present invention, the first network model and the second network model are trained and optimized by supervised learning based on the data set to obtain a final allocation model.

[0120] In order to make the obtained allocation model more accurate, before training the first network model and the second network model, the data set is first normalized;

[0121] For example, the data set may be normalized using an existing standard normalization process or other methods.

[0122] Specifically, the data state information and the performance state information are input into the first network model, and after being processed by the first network model, the gain characteristics of the data and performance are output; wherein the first network model may be a feature extraction network model, which performs feature fusion and feature extraction on the data state information and the performance state information, and then outputs the gain characteristics of the data and performance;

[0123] The data output by the first network model, the gain characteristics of the performance, and the channel state information are then input into the second network model.

[0124] The channel state information includes channel quality, available channel resources, interference and other related parameters;

[0125] Furthermore, since the channel state information is related to the frequency, and the convolutional neural network can effectively extract data features with specific position information (sorted by fixed frequency), and the computational complexity of the convolutional neural network is much lower than that of the fully connected network, the embodiment of the present invention adopts the convolutional neural network as the second network model;

[0126] In addition, during the training of the first network model and the second network model, the first input channel is the gain characteristics of the data and performance output by the first network model, the second input channel is the channel state information sorted by a fixed frequency, and the output channel is the training allocation result.

[0127] The above training process is repeated, and the first network model and the second network model are trained multiple times based on a large number of data sets, thereby obtaining the first target network model and the second target network model, and finally obtaining the allocation model.

[0128] Among them, during the training process, based on the supervised learning method, the accuracy of the training allocation results output by the second network model is judged by assigning labels, and an accuracy threshold can be set. When the accuracy of the training allocation results reaches the accuracy threshold, the first network model and the second network model corresponding to the current training allocation results can be determined as the first target network model and the second target network model.

[0129] In addition, a training number threshold may be set. When the training number threshold is reached, the training of the first network model and the second network model is stopped, and the first target network model and the second target network model are obtained.

[0130] It should be noted that the first network model in the embodiment of the present invention is illustrated by a convolutional neural network, but is not limited to a convolutional neural network.

[0131] Step S3: allocating resource units based on the first target network model, the second target network model and the target state information to obtain an allocation result;

[0132] Specifically, allocating resource units based on the first target network model, the second target network model, and the target state information to obtain an allocation result includes:

[0133] Acquire target state information; wherein the target state information includes first target state information, second target state information and third target state information;

[0134] Processing the first target state information and the second target state information based on the first target network model, and outputting a first result;

[0135] The first result and the third target state information are processed based on the second target network model, and the allocation result is output.

[0136] Wherein, when each round of OFDMA resource unit allocation is turned on, first target state information, second target state information and third target state information are collected based on each client running in the current scenario;

[0137] Furthermore, the first target state information corresponds to data state information of each client in the current scenario; the second target state information corresponds to performance state information of each client in the current scenario; and the third target state information corresponds to channel state information of each client in the current scenario.

[0138] In the actual application scenario, when each round of OFDMA resource unit allocation is turned on, the data status information, performance status information and channel status information of each client in the current scenario are collected, and the relevant data of the data status information and the performance status information are input into the first target network model. The first target network model performs feature fusion and extraction on the acquired data status information and performance status information, and then outputs feature data related to data and performance;

[0139] Then, the characteristic data related to the data and performance and the channel state information data are input into the second target network model. The second target network model evaluates the final performance based on the acquired data, the characteristic data related to the performance and the channel state information, and then outputs the allocation result of the resource unit.

[0140] The obtained first target network model and the second target network model are actually deployed to facilitate the subsequent calling of the allocation model.

[0141] Step S4: Allocate the resource units based on the allocation result.

[0142] In an actual application scenario, after the allocation result is output based on the second target network model, the allocation result is executed to allocate resource units.

[0143] In an embodiment of the present invention, a dual network is used to allocate resource units; wherein the first target network model is used to obtain data and performance-related characteristics of the client, and the second target network model is used to evaluate the final performance in combination with the current channel environment. Combining the first target network model with the second target network model makes it easier to guide the relevant characteristics of the network to learn in stages, thereby obtaining a more reasonable allocation method for resource units.

[0144] The above embodiments of the present invention can be implemented based on the following implementation methods:

[0145] Step S201: collecting multiple status information of each client to obtain a data set;

[0146] Step S202: obtaining a distribution label: for example, obtaining a distribution label based on a test method or a distribution gain index;

[0147] Step S203: construct two feature extraction networks to extract data performance features and channel features respectively, and perform training optimization;

[0148] Step S204: Deploy the network: deploy the two networks trained and optimized in step S203.

[0149] Step S205: In each round of allocation process, the status information of the relevant clients is collected, and the allocation result is obtained after passing through two networks.

[0150] The embodiment of the present invention further provides a resource unit allocation device 300, comprising:

[0151] The acquisition module 301 is used to collect multiple status information of each client to obtain a data set;

[0152] A training module 302 is used to train the network model based on the data set to obtain an allocation model; wherein the allocation model includes a first target network model and a second target network model;

[0153] The processing module 303 is used to allocate resource units based on the first target network model, the second target network model and the target state information to obtain an allocation result;

[0154] The allocation module 304 is configured to allocate the resource units based on the allocation result.

[0155] Optionally, the step of obtaining a data set based on multiple pieces of status information of each client includes:

[0156] Acquire first status information, second status information, and third status information of each of the clients;

[0157] A data set is obtained based on the first state information, the second state information and the third state information.

[0158] Optionally, obtaining a data set based on the first state information, the second state information, and the third state information includes:

[0159] Acquire each scene corresponding to the first state information, the second state information, and the third state information;

[0160] Label each of the scenes to obtain assigned labels;

[0161] Based on the assigned labels, the data set is obtained.

[0162] Optionally, labeling each of the scenes to obtain assigned labels includes:

[0163] Based on the test, obtain the target allocation for each scenario; or

[0164] The target allocation method for each scenario is obtained by performing calculation based on the allocation gain index and the allocation gain of each allocation method.

[0165] Optionally, obtaining the allocation gain indicator of each client at each moment includes:

[0166] Obtain a first parameter of each of the clients at each moment; the first parameter is positively correlated with the allocation gain index; obtain a second parameter of each of the clients at each moment; the second parameter is negatively correlated with the allocation gain index; calculate based on the first parameter and the second parameter to obtain the allocation gain index of each of the clients at each moment.

[0167] Optionally, the training of the network model based on the data set to obtain the allocation model includes:

[0168] Preprocessing the data set to obtain a target data set;

[0169] Inputting the first state information and the second state information into a first network model, training the first network model to obtain a first target network model; and outputting an initial result based on the first target network model;

[0170] Inputting the initial result and the third state information into a second network model, training the second network model, and obtaining a second target network model;

[0171] The allocation model is obtained based on the first target network model and the second target network model.

[0172] Optionally, allocating resource units based on the first target network model, the second target network model, and the target state information to obtain an allocation result includes:

[0173] Acquire target state information; wherein the target state information includes first target state information, second target state information and third target state information;

[0174] Processing the first target state information and the second target state information based on the first target network model, and outputting a first result;

[0175] The first result and the third target state information are processed based on the second target network model, and the allocation result is output.

[0176] An embodiment of the present invention further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the method described above when executing the computer program.

[0177] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method described above.

[0178] In addition, other structures and functions of the device in the embodiment of the present invention are known to those skilled in the art and will not be described in detail here to reduce redundancy.

[0179] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0180] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0181] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0182] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0183] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0184] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0185] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0186] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A resource unit allocation method, characterized in that: include: Collect multiple status information of each client to obtain a data set; Training the network model based on the data set to obtain an allocation model; wherein the allocation model includes a first target network model and a second target network model; Allocate resource units based on the first target network model, the second target network model and the target state information to obtain an allocation result; Allocating the resource units based on the allocation result; The step of obtaining a data set based on multiple status information of each client includes: Acquire first status information, second status information, and third status information of each of the clients; Based on the first state information, the second state information and the third state information, obtaining a data set; The obtaining of a data set based on the first state information, the second state information, and the third state information includes: Acquire each scene corresponding to the first state information, the second state information, and the third state information; Label each of the scenes to obtain assigned labels; Based on the assigned labels, the data set is obtained.

2. The method according to claim 1, characterized in that The step of labeling each of the scenes to obtain a distribution label includes: Based on the test, obtain the target allocation for each scenario; or The target allocation method for each scenario is obtained by performing calculation based on the allocation gain index and the allocation gain of each allocation method.

3. The method according to claim 2, characterized in that Obtaining the allocation gain index of each client at each moment, including: Obtaining a first parameter of each of the clients at each moment; the first parameter is positively correlated with the allocation gain indicator; Obtain a second parameter of each of the clients at each moment; the second parameter is negatively correlated with the allocation gain indicator; and calculate based on the first parameter and the second parameter to obtain the allocation gain indicator of each of the clients at each moment.

4. The method according to claim 1, characterized in that The training of the network model based on the data set to obtain the allocation model includes: Preprocessing the data set to obtain a target data set; Inputting the first state information and the second state information into a first network model, training the first network model to obtain a first target network model; and outputting an initial result based on the first target network model; Inputting the initial result and the third state information into a second network model, training the second network model, and obtaining a second target network model; The allocation model is obtained based on the first target network model and the second target network model.

5. The method according to claim 1, characterized in that Allocating resource units based on the first target network model, the second target network model, and the target state information to obtain an allocation result includes: Acquire target state information; wherein the target state information includes first target state information, second target state information and third target state information; Processing the first target state information and the second target state information based on the first target network model, and outputting a first result; The first result and the third target state information are processed based on the second target network model, and the allocation result is output.

6. A resource unit allocation device, characterized in that: include: The acquisition module is used to collect multiple status information of each client and obtain each client of the data set; A training module, used to train the network model based on the data set to obtain an allocation model; wherein the allocation model includes a first target network model and a second target network model; A processing module, configured to allocate resource units based on the first target network model, the second target network model and the target state information to obtain an allocation result; An allocation module, configured to allocate the resource units based on the allocation result; The step of obtaining a data set based on multiple status information of each client includes: Acquire first status information, second status information, and third status information of each of the clients; Based on the first state information, the second state information and the third state information, obtaining a data set; The obtaining of a data set based on the first state information, the second state information, and the third state information includes: Acquire each scene corresponding to the first state information, the second state information, and the third state information; Label each of the scenes to obtain assigned labels; Based on the assigned labels, the data set is obtained.

7. An electronic device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the method according to any one of claims 1 to 5 when executing the computer program.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 5.

Images