License scheduling method and device and readable storage medium

Abstract

The invention provides a License scheduling method and device and a readable storage medium. The method comprises the following steps: predicting a load prediction value of a base station in a next period; in response to the condition that the load prediction value is greater than or equal to a preset scheduling threshold value, calling a first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and storing the allocation record in the block chain; and when it is detected that the actual load value of the base station meets a preset rollback condition, calling a second interface of the smart contract to rollback the allocated License to the License pool. According to the method, the device and the medium, the problems that the resource utilization rate is low, the flexibility is poor, the cost of an operator is greatly increased and the network experience of a user is influenced due to the fact that an existing License scheduling method mainly adopts a fixed allocation mode and does not fully consider the difference between busy hours and idle hours and the actual demand of co-construction and sharing are solved.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and in particular, to a method and device for scheduling Licenses and a readable storage medium. Background Art

[0002] A License can be understood as a carrier frequency software. When communication services continue to expand and value-added features need to be used or the resources of base station equipment need to be expanded, an operator can purchase Licenses for the corresponding functions or resources of the equipment to meet the needs of the services. That is to say, when an operator purchases hardware equipment such as base stations from an equipment manufacturer, it must also purchase the corresponding Licenses in order to use the hardware equipment to normally carry out network services.

[0003] Currently, different operators can carry out co-construction and sharing of 4G and 5G networks, which is an innovative model with unique characteristics in the field of communication network construction. To fully create a new situation of high-quality development in the 5G era, each operator needs to further expand the advantages of co-construction and sharing, strengthen collaborative cooperation, and pool efforts.

[0004] However, the Licenses currently purchased by operators are all fixedly allocated to each base station cell for use. If there are n base station cells, at least n Licenses need to be purchased. This not only greatly increases the costs of operators, but also does not fully consider the differences between busy and idle times and the actual needs of co-construction and sharing, resulting in most Licenses not being effectively and reasonably utilized. During idle times, it will cause network equipment to be idle and resources to be wasted. During busy times, it will cause problems such as excessive load and insufficient resources. The overall flexibility is poor and the efficiency is low. Moreover, if the business requirements cannot be met in a timely manner due to a lack of Licenses, it will also affect the network experience of users. Summary of the Invention

[0005] The technical problem to be solved by the present invention is the above-mentioned deficiencies of the prior art. A method and device for scheduling Licenses and a readable storage medium are provided to solve the problem that the existing License scheduling method mainly adopts a fixed allocation method, without fully considering the differences between busy and idle times and the actual needs of co-construction and sharing, resulting in low resource utilization rate, poor flexibility, greatly increasing the costs of operators, and affecting the network experience of users.

[0006] In a first aspect, the present invention provides a method for scheduling a License, and the method includes:

[0007] Predicting the predicted load value of a base station in the next cycle;

[0008] In response to the load prediction value being greater than or equal to a preset scheduling threshold, call the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and save the record of this allocation to the blockchain; wherein, the License pool includes multiple sharable Licenses;

[0009] When it is detected that the actual load value of the base station meets the preset fallback condition, call the second interface of the smart contract to fallback the allocated License to the License pool, and save the record of this fallback to the blockchain.

[0010] Further, before predicting the load prediction value of the base station in the next cycle, the method further includes:

[0011] Construct a License pool including multiple sharable Licenses;

[0012] Register multiple sharable Licenses on the blockchain;

[0013] Deploy the smart contract to the blockchain.

[0014] Further, predicting the load prediction value of the base station in the next cycle specifically includes:

[0015] Collect the metric data of the base station in the current cycle and the previous N cycles; wherein, N is an integer greater than or equal to 1;

[0016] Obtain the normalized network load values of the current cycle and the previous N cycles according to the metric data;

[0017] Input the normalized network load values of the current cycle and the previous N cycles into a pre-constructed prediction model to obtain the load prediction value of the base station in the next cycle.

[0018] Further, the load prediction value is a single load prediction value or a comprehensive load prediction value. If the load prediction value is a single load prediction value, the metric data includes the Physical Resource Block (PRB) utilization rate. The obtaining the normalized network load values of the current cycle and the previous N cycles according to the metric data specifically includes:

[0019] Perform normalization processing on the PRB utilization rates of the current cycle and the previous N cycles respectively to obtain the normalized network load values of the current cycle and the previous N cycles;

[0020] If the load prediction value is a comprehensive load prediction value, the metric data includes PRB utilization rate, average reference signal receiving power (RSRP), latency, and packet loss rate. Obtaining the normalized network load values for the current cycle and the previous N cycles based on the metric data specifically includes:

[0021] Normalize the PRB utilization rate, average RSRP, latency, and packet loss rate for the current cycle and the previous N cycles respectively, and perform weighted summation on the normalized PRB utilization rate, average RSRP, latency, and packet loss rate according to preset weights to obtain the normalized network load values for the current cycle and the previous N cycles.

[0022] Further, after calling the first interface of the smart contract to allocate an idle License from the pre-constructed License pool to the base station, the method further includes:

[0023] Modify the current status of the allocated License from idle to allocated, and record the scheduling timestamp and geographical location of the allocated License;

[0024] After calling the second interface of the smart contract to roll back the allocated License to the License pool, the method further includes:

[0025] Modify the current status of the allocated License from allocated to idle, and clear the scheduling timestamp and geographical location.

[0026] Further, the rollback condition is one of the following:

[0027] The actual load value of the base station is less than the scheduling threshold and lasts for a preset time or a preset number of cycles;

[0028] The actual load value of the base station is less than a preset rollback threshold, where the rollback threshold is less than the scheduling threshold;

[0029] The actual load value of the base station is less than the rollback threshold and lasts for a preset time or a preset number of cycles.

[0030] Further, the method further includes:

[0031] Update the scheduling threshold according to the difference between the actual load value and the load prediction value; and,

[0032] Adjust the weight parameters of the prediction model according to the difference between the actual load value and the load prediction value.

[0033] Further, updating the scheduling threshold according to the difference between the actual load value and the load prediction value is specifically calculated according to the following formula:

[0034]

[0035] where represents the scheduling threshold, represents the updated scheduling threshold, k is a set learning rate, and ΔL is the difference between the actual load value and the load prediction value.

[0036] In a second aspect, the present invention provides a scheduling device for a license, and the device includes:

[0037] A load prediction module, configured to predict the load prediction value of the base station in the next period;

[0038] A License allocation module, connected to the load prediction module, and configured to, in response to the load prediction value being greater than or equal to a preset scheduling threshold, call a first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and save the record of this allocation to the blockchain; wherein, the License pool includes a plurality of sharable Licenses;

[0039] A License fallback module, connected to the License allocation module, and configured to, when detecting that the actual load value of the base station meets a preset fallback condition, call a second interface of the smart contract to fallback the allocated License to the License pool, and save the record of this fallback to the blockchain.

[0040] In a third aspect, the present invention provides a scheduling device for a license, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to implement the License scheduling method described in the first aspect above.

[0041] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the License scheduling method described in the first aspect above.

[0042] The License scheduling method, device, and readable storage medium provided by the present invention first predict the load prediction value of the base station in the next cycle; then, in response to the load prediction value being greater than or equal to a preset scheduling threshold, call the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and save the record of this allocation to the blockchain; wherein, the License pool includes multiple sharable Licenses; and when it is detected that the actual load value of the base station meets the preset fallback condition, call the second interface of the smart contract to fallback the allocated License to the License pool, and save the record of this fallback to the blockchain. The present invention monitors and judges the load prediction value and the actual load value of the base station in real time. When the load prediction value reaches the scheduling threshold, a License is allocated from the License pool; and when the actual load value meets the fallback condition, the License is fallback. This method fully considers the differences between busy and idle base stations, as well as the actual needs of co-construction and sharing, making up for the deficiencies of the existing fixed allocation method. It not only gives full play to and expands the advantages of co-construction and sharing, but also greatly improves resource utilization and scheduling flexibility, significantly reduces the cost of operators purchasing Licenses, and effectively improves the network experience of users. It solves the problem that the existing License scheduling methods mainly adopt a fixed allocation method, without fully considering the differences between busy and idle times and the actual needs of co-construction and sharing, resulting in low resource utilization and poor flexibility, which not only greatly increases the costs of operators, but also affects the network experience of users. Description of the Drawings

[0043] Figure 1 It is a flowchart of a License scheduling method according to Embodiment 1 of the present invention;

[0044] Figure 2 It is a schematic structural diagram of a License scheduling device according to Embodiment 2 of the present invention;

[0045] Figure 3 It is a schematic structural diagram of a License scheduling device according to Embodiment 3 of the present invention. Detailed Embodiments

[0046] To enable those skilled in the art to better understand the technical solutions of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0047] It can be understood that the specific embodiments and drawings described herein are only used to explain the present invention, rather than limiting the present invention.

[0048] It can be understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0049] It can be understood that, for the convenience of description, only the parts related to the present invention are shown in the drawings of the present invention, while the parts unrelated to the present invention are not shown in the drawings.

[0050] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one entity structure, or may be composed of multiple entity structures, or multiple units and modules may also be integrated into one entity structure.

[0051] It can be understood that the terms "first", "second", etc. in the embodiments of the present invention are used to distinguish different objects, or to distinguish different processes for the same object, rather than to describe a specific order of the objects.

[0052] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in a different order from that marked in the drawings.

[0053] It can be understood that in the flowcharts and block diagrams of the present invention, the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the embodiments of the present invention are shown. Among them, each block in the flowchart or block diagram may represent a unit, module, program segment, or code, which contains executable instructions for implementing the specified function. Moreover, each block or combination of blocks in the block diagram and flowchart may be implemented by a hardware-based system for implementing the specified function, or by a combination of hardware and computer instructions.

[0054] It can be understood that the units and modules involved in the embodiments of the present invention may be implemented in software or in hardware. For example, the units and modules may be located in the processor.

[0055] Embodiment 1:

[0056] This embodiment provides a method for scheduling Licenses. As Figure 1 shown, the method includes:

[0057] Step S101: Predict the load prediction value of the base station in the next cycle.

[0058] In this embodiment, the base station specifically refers to any base station that can share all Licenses in the License pool, and the License pool contains multiple sharable Licenses. The cycle can be divided according to a preset time period. For example, one cycle is T seconds (e.g., T = 90 seconds), or one cycle is T minutes (e.g., T = 1 minute), etc. The load prediction value is used to evaluate the load condition of the base station in the next cycle, providing a key basis for subsequent License scheduling strategies.

[0059] Step S102: In response to the load prediction value being greater than or equal to a preset scheduling threshold, call the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and save the record of this allocation to the blockchain; wherein, the License pool includes multiple sharable Licenses.

[0060] In this embodiment, the scheduling threshold can be preset according to the actual situation; the smart contract is used to automatically complete the allocation and fallback of Licenses according to preset rules without manual intervention, thereby reducing the risks brought by human errors or operation delays. Among them, the first interface is used to allocate an idle License from a pre-constructed License pool to the base station and save the record of this allocation to the blockchain. For example, the definition of the first interface can be as follows:

[0061] allocateLicense(operatorID, basestationID, predictedLoad, licenseType, timestamp)

[0062] Among them, operatorID represents the operator identifier (optional), basestationID represents the base station identifier (mandatory), predictedLoad represents the load prediction value (optional), licenseType represents the License type (optional), and timestamp represents the timestamp (mandatory at the time of allocation).

[0063] In this embodiment, saving the record of this allocation to the blockchain specifically includes: writing the base station identifier, License identifier, allocation time, and operation type (allocation) involved in this allocation into the blockchain. According to actual requirements, information such as the operator identifier and load prediction value can also be selectively written.

[0064] Optionally, before predicting the load prediction value of the base station in the next cycle, the method further includes:

[0065] Construct a License pool including multiple sharable Licenses;

[0066] Register multiple sharable Licenses on the blockchain;

[0067] Deploy the smart contract to the blockchain.

[0068] In this embodiment, all shareable Licenses are pre-registered on the blockchain. Assuming the initial quantity is N (for example, N = 100), each License object includes at least a License identifier and a current status, and may also include an assigned operator, a scheduling timestamp, a geographical location field, etc.

[0069] In this embodiment, the main function of deploying the smart contract is to provide an automated, transparent, and secure execution platform for the allocation and rollback of Licenses. The smart contract deployed on the blockchain records all operation information, and once these records are written, they cannot be tampered with, providing an open and transparent audit chain for the entire scheduling process and enhancing the trust of all parties in the system.

[0070] Optionally, the predicted load value of the base station in the next period specifically includes:

[0071] Collect the metric data of the base station in the current period and the previous N periods; where N is an integer greater than or equal to 1;

[0072] Obtain the normalized network load values of the current period and the previous N periods according to the metric data;

[0073] Input the normalized network load values of the current period and the previous N periods into a pre-constructed prediction model to obtain the predicted load value of the base station in the next period.

[0074] In this embodiment, N can be set according to the actual situation. For example, it can be 1, 2, 3, or 4, etc. When it is 3, the metric data of the base station in the current period, the previous period, and the period before the previous period will be collected.

[0075] In this embodiment, the normalized network load value, that is, the actual load value corresponding to the current period and the previous N periods, is a value between 0 and 1. By inputting the normalized network load values of the current period and the previous N periods into a pre-constructed prediction model, the predicted load value of the base station in the next period can be obtained.

[0076] In this embodiment, the prediction model is preferably a time series model, such as a weighted moving average model, an autoregressive integrated moving average model, etc.

[0077] Optionally, the predicted load value is a single predicted load value or a comprehensive predicted load value. If the predicted load value is a single predicted load value, the metric data includes the utilization rate of PRB (Physical Resource Block). The obtaining of the normalized network load values of the current period and the previous N periods according to the metric data specifically includes:

[0078] Normalize the PRB utilization rates of the current cycle and the previous N cycles respectively to obtain the network load values after normalization for the current cycle and the previous N cycles.

[0079] If the load prediction value is a comprehensive load prediction value, the metric data includes PRB utilization rate, average RSRP (Reference Signal Receiving Power), delay, and packet loss rate. Obtaining the network load values after normalization for the current cycle and the previous N cycles according to the metric data specifically includes:

[0080] Normalize the PRB utilization rate, average RSRP, delay, and packet loss rate of the current cycle and the previous N cycles respectively, and perform weighted summation on the normalized PRB utilization rate, average RSRP, delay, and packet loss rate according to preset weights to obtain the network load values after normalization for the current cycle and the previous N cycles.

[0081] It should be noted that when performing load prediction, for all base stations, single load prediction can be used only, or comprehensive load prediction can be used only, or for some base stations, single load prediction can be used, while for other base stations, comprehensive load prediction can be used. The present invention does not limit this.

[0082] In this embodiment, if the load prediction value is a single load prediction value, mainly collect the PRB utilization rates of the current cycle and the previous N cycles of the base station, perform normalization processing, and then use the PRB utilization rates after normalization for the current cycle and the previous N cycles as the network load values after normalization for the current cycle and the previous N cycles of the base station.

[0083] In this embodiment, if the load prediction value is a comprehensive load prediction value, mainly collect the PRB utilization rate, average RSRP, delay, and packet loss rate of the current cycle and the previous N cycles of the base station, perform normalization processing, and then perform weighted summation on the normalized PRB utilization rate, average RSRP, delay, and packet loss rate according to preset weights (that is, multiply each normalized metric by its corresponding weight respectively, and finally add these products) to obtain the network load values after normalization for the current cycle and the previous N cycles.

[0084] Step S103: When it is detected that the actual load value of the base station satisfies the preset fallback condition, call the second interface of the intelligent contract to fallback the allocated License to the License pool, and save the record of this fallback to the blockchain.

[0085] In this embodiment, when it is detected that the actual load value of the base station meets the preset fallback condition, which means that the network load has decreased or the problem has been improved, at this time, the license can be returned to the license pool so that other base stations can use it. Among them, the second interface is used to return the allocated license to the license pool and save the record of this fallback to the blockchain. For example, the definition of the second interface can be as follows:

[0086] releaseLicense(licenseID, timestamp)

[0087] Among them, licenseID represents the license identifier, and timestamp represents the timestamp (when falling back).

[0088] In this embodiment, saving the record of this fallback to the blockchain specifically includes: writing the license identifier, fallback time, and operation type (fallback) involved in this allocation into the blockchain. According to actual needs, information such as the actual load value and base station identifier can also be selectively written.

[0089] Optionally, after the first interface that calls the smart contract allocates an idle license from the pre-constructed license pool to the base station, the method further includes:

[0090] Changing the current status of the allocated license from idle to allocated, and recording the scheduling timestamp and geographical location of the allocated license;

[0091] After the second interface that calls the smart contract returns the allocated license to the license pool, the method further includes:

[0092] Changing the current status of the allocated license from allocated to idle, and clearing the scheduling timestamp and geographical location.

[0093] In this embodiment, the current states of all Licenses include idle and allocated. When each License object has a scheduling timestamp and a geographical location, during License allocation, an idle License is allocated from the License pool to the base station through the first interface of the smart contract, and the current state of the License is modified to allocated. In addition, the corresponding scheduling timestamp and geographical location of the License are recorded (if it is a co - construction and sharing scenario, the allocated operator can also be recorded). Meanwhile, the record of this allocation is saved to the blockchain. Similarly, during License fallback, the allocated License is returned to the License pool through the second interface of the smart contract, the current state of the License is modified to idle, and the scheduling timestamp and geographical location are cleared (if the allocated operator is recorded, it is also cleared). Meanwhile, the record of this allocation is saved to the blockchain, thus realizing the transparency, automated recording, and non - tamperable evidence storage of the entire process of License resource allocation and fallback.

[0094] Optionally, the fallback condition is one of the following:

[0095] The actual load value of the base station is less than the scheduling threshold and lasts for a preset time or a preset number of cycles;

[0096] The actual load value of the base station is less than a preset fallback threshold, where the fallback threshold is less than the scheduling threshold;

[0097] The actual load value of the base station is less than the fallback threshold and lasts for a preset time or a preset number of cycles.

[0098] In this embodiment, to avoid false triggering caused by instantaneous fluctuations and to avoid frequent fallback and scheduling, fallback can be performed when the actual load value of the base station is less than the scheduling threshold or the fallback threshold and lasts for a preset time or a preset number of cycles, or fallback can be performed when the actual load value of the base station is less than the preset fallback threshold. Among them, the preset time of duration is preferably an integer multiple of the cycle.

[0099] Optionally, the method further includes:

[0100] Updating the scheduling threshold according to the difference between the actual load value and the load prediction value; and,

[0101] Adjusting the weight parameters of the prediction model according to the difference between the actual load value and the load prediction value.

[0102] In this embodiment, to avoid over-allocation or under-allocation of License resources and to improve the utilization rate of License resources, the scheduling threshold can be updated according to the difference between the actual load value and the corresponding load prediction value in the same period to gradually adapt to the real network situation. At the same time, the difference between the actual load value and the load prediction value provides direct feedback for evaluating the accuracy of the prediction model. By analyzing this difference, it can be understood in which aspects the model has deviations, and then the weight parameters of the prediction model can be adjusted, which helps to improve the accuracy and adaptability of the prediction model, enabling the model to better capture the law of load changes and providing more reliable results for subsequent load prediction.

[0103] Optionally, updating the scheduling threshold according to the difference between the actual load value and the load prediction value is specifically calculated according to the following formula:

[0104]

[0105] Wherein, represents the scheduling threshold, represents the updated scheduling threshold, k is a set learning rate, and ΔL is the difference between the actual load value and the load prediction value.

[0106] In this embodiment, k represents the learning rate, which can be set according to actual needs. For example, k = 0.1. Assume that the current actual load value L of the base station actual = 0.83, and the load prediction value L pred = 0.85, and the current scheduling threshold Then: ΔL = 0.83 - 0.85 = -0.02. The updated scheduling threshold

[0107] It should be noted that the License scheduling method provided by the present invention can give full play to and expand the advantages of co-construction and sharing through the implementation of joint, efficient, accurate, intelligent, and dynamic License scheduling, and can significantly reduce the cost of operators purchasing Licenses. At the same time, operators can flexibly schedule Licenses according to actual business needs to obtain accurate network functions, and then provide users with a better network experience. Therefore, the method of the present invention has the potential for large-scale popularization and application.

[0108] In a specific embodiment, the License scheduling method may include the following steps:

[0109] Step 1. Construct a License pool and a security ledger system

[0110] Step 1.1. Initialize the License pool

[0111] (1) First, perform License registration. All shareable Licenses are pre-registered on the blockchain. Assume the initial quantity is N (for example, N = 100). Each License object contains:

[0112] A unique identifier ID (such as L001 to L100)

[0113] The current status (the initial status is "free")

[0114] The allocated operator (can be empty and is written when to be allocated)

[0115] The scheduling timestamp and geographical location fields (information is written when recording the allocation later and cleared after rollback)

[0116] It should be noted that the status of the License mainly includes two types: free and allocated. In actual applications, other statuses can also be included.

[0117] It should be noted that this embodiment is also applicable to the situation where multiple base stations of a single operator are shared. In this case, the allocated operator can be omitted. In the scenario of multiple operators sharing, since the base station identifier can directly or indirectly determine the operator to which it belongs, the allocated operator can also be replaced with the allocated base station identifier ID.

[0118] (2) Deploy the smart contract: Write the smart contract to provide the following core interfaces:

[0119] allocateLicense(operatorID, basestationID, predictedLoad, licenseType, timestamp)

[0120] Among them, operatorID represents the operator identifier, basestationID represents the base station identifier, predictedLoad represents the predicted load (i.e., the load prediction value), licenseType represents the License type, and timestamp represents the timestamp (at the time of allocation).

[0121] Function: Automatically select a License in the "free" state from the License pool according to the operator, base station identifier, predicted load, and the called License type (such as expansion) that needs to allocate the License, update its status to "allocated", and write all information to the blockchain.

[0122] releaseLicense(licenseID, timestamp)

[0123] Among them, licenseID represents the License identifier, and timestamp represents the timestamp (when rolling back).

[0124] Function: Update the specified License status to "free" and record the rollback time.

[0125] Automatically verify the logic embedded in the contract to ensure that there is an immutable record after each status change.

[0126] The main role of deploying the smart contract is to provide an automated, transparent, and secure execution platform for License allocation and rollback. The smart contract can automatically complete License allocation and rollback according to pre-set rules, without manual intervention, reducing the risks brought by human errors or operation delays. The smart contract deployed on the blockchain records all operation information, and once these records are written, they cannot be tampered with, providing a public and transparent audit chain for the entire scheduling process and enhancing the trust of all parties in the system. Since the smart contract is executed on a decentralized blockchain, it reduces the single-point failures and security vulnerabilities that may exist in a centralized system, ensuring that the License scheduling process is not subject to malicious attacks or tampering.

[0127] Step 1.2, Secure ledger record

[0128] Each License scheduling (allocation or rollback) calls the smart contract interface and stores the transaction record on the chain.

[0129] Example of allocation record:

[0130] {

[0131] "licenseID":"L042",

[0132] "operatorID":"OP_A",

[0133] "basestationID":{"02411"},

[0134] "predictedLoad":0.85,

[0135] "action":"allocate",

[0136] "timestamp":"2025-03-06T14:05:00Z"

[0137] }

[0138] Among them, licenseID is the license identifier, operatorID is the operator identifier, basestationID is the base station identifier, predictedLoad is the load prediction value, action is the operation type (allocation), and timestamp is the allocation time.

[0139] This approach ensures that subsequent resource allocation and scheduling decisions can be traced and audited.

[0140] It should be noted that license allocation, license rollback, transaction records, etc. are all automatically completed by smart contracts.

[0141] (1) License allocation: For example, when the load is high, a license application is required. Then an available license is automatically found from the virtual license pool and bound to the operator at the specified location. The license status is changed to "allocated" and the allocation time, type, predicted load, etc. are recorded and written into the blockchain to form an audit record that cannot be tampered with.

[0142] (2) License rollback: For example, if a license rollback is required due to load relief or indicator recovery, the license status will be automatically changed to "free" and the binding with the base station or operator will be released. The rollback time, trigger reason and other records will also be written into the blockchain to form an audit record that cannot be tampered with.

[0143] (3) Transaction records: For example, every allocation, rollback, and status change will be automatically written into the blockchain ledger by the smart contract.

[0144] Step 2: Build intelligent prediction module and adaptive parameter adjustment

[0145] Step 2.1: Collect data first:

[0146] The following indicators of base stations are collected from the base station network management, but are not limited to these indicators:

[0147] RSRP average value (signal strength, unit: dBm)

[0148] PRB utilization (refers to the PRB resource utilization of the base station, which reflects the load of the base station)

[0149] Delay (unit: ms)

[0150] Packet loss rate (unit: %)

[0151] Clean and normalize the data collected within each acquisition period of T seconds (e.g., T = 60 seconds) so that all metrics fall within a unified range (e.g., 0 to 1) for subsequent calculations. The processed data forms a unified time series data record table.

[0152] It should be noted that the base stations include all base stations that can share the License.

[0153] Step 2.2: Build a prediction model:

[0154] Select a weighted moving average model, and its calculation formula is:

[0155] L pred (t + 1)= w1 * L(t)+ w2 * L(t - 1)+ w3 * L(t - 2)

[0156] Where L(t) is the network load value after normalization in the current period (t), with a value range of 0 to 1, and L pred (t + 1) is the predicted load value for the next period (t + 1), and L(t - 1) and L(t - 2) respectively represent the network loads after normalization in the previous period (t - 1) and the period before the previous period (t - 2). w1, w2, and w3 are the corresponding weights, usually satisfying w1 + w2 + w3 = 1. For example, the network loads in each period can be as shown in Table 1:

[0157] Table 1: Example of network loads in each period

[0158] Cycle number (t) Time (assuming sampling once per minute) Actual load L t-2 14:00-14:01 0.78 t-1 14:01-14:02 0.81 t 14:02-14:03 0.85 t+1 14:03-14:04 Prediction phase

[0159] This load can be a single load. For example, if it is a capacity load, it is predicted based on the PRB utilization rate.

[0160] It can also be a comprehensive load. For the comprehensive load, the normalized average RSRP, PRB utilization rate, delay, packet loss rate, etc. are combined according to certain weights. For example: 0.4 × average RSRP + 0.2 × PRB utilization rate + 0.3 × reciprocal of packet loss rate + 0.1 × delay), where the reciprocal of the packet loss rate refers to the reciprocal of the packet loss rate.

[0161] Step 2.3: Make a prediction

[0162] Specifically, the model can be implemented using Python, and the example code is as follows:

[0163] import numpy as np

[0164] def predict_load(load_series):

[0165] "″"

[0166] Input: load_series is a list of load data for the last three acquisition cycles (normalized values).

[0167] Output: Predicted load value for the next cycle

[0168] "″"

[0169] # Define weights (i.e., weights for each acquisition cycle) to ensure the sum is 1

[0170] weights = np.array([0.5, 0.3, 0.2])

[0171] # When the data is less than 3 cycles, use the simple average

[0172] if len(load_series) < 3:

[0173] return np.mean(load_series)

[0174] # Take the load data for the last three cycles

[0175] recent_loads = np.array(load_series[-3:])

[0176] # Calculate the predicted value

[0177] prediction = np.dot(weights, recent_loads)

[0178] return prediction

[0179] # Example call

[0180] recent_loads_example = [0.75, 0.80, 0.78] # Assume normalized load data within three cycles

[0181] predicted_load = predict_load(recent_loads_example)

[0182] print("Predicted load value:", predicted_load)

[0183] Here, the function in the code is called every T seconds, and the latest load data is passed in to output the predicted load value for the next cycle. That is, the prediction model runs every T seconds to output the predicted load value L within the next minute. pred 。

[0184] Step 2.4: Invoke the corresponding License according to the load condition

[0185] Set the initial scheduling threshold T load (such as 0.8).

[0186] If a single load is used, such as a capacity load, and L pred ≥T load , it is considered that the base station may have insufficient capacity in the future, and then the expansion License is invoked. If a comprehensive load is adopted, the comprehensive scheduling License is invoked, and this License optimizes capabilities such as latency through a joint scheduling strategy. If, L pred <T load , then keep the current state.

[0187] It should be noted that the License types include expansion License and comprehensive scheduling License. In actual applications, the corresponding License types may not be distinguished, that is, all Licenses in the License pool are suitable for both single-load prediction and comprehensive-load prediction scenarios. When performing load prediction, for all base stations, single-load prediction can be used only, comprehensive-load prediction can be used only, or, single-load prediction can be used for some base stations, while comprehensive-load prediction can be used for other base stations. The present invention does not limit this.

[0188] Step 2.5: Perform adaptive parameter adjustment:

[0189] Record the actual load value L actual and the difference from the predicted load value L pred : ΔL = L actual - L pred ;

[0190] Set the learning rate k (for example, k = 0.1) and update the scheduling threshold:

[0191]

[0192] wherein, represents the old scheduling threshold, represents the updated scheduling threshold.

[0193] This adjustment ensures that the prediction model and the scheduling threshold gradually adapt to the real network situation, avoiding over-scheduling or under-scheduling.

[0194] At the same time, according to the error performance, the model weights can be adjusted. For example, through historical data fitting, w1, w2, and w3 are dynamically adjusted to minimize the prediction error.

[0195] For example, the current actual load value L of a certain base stationactual = 0.83, the load prediction value L pred = 0.85, then: ΔL = 0.83 - 0.85 = -0.02

[0196] The updated scheduling threshold = 0.8 + 0.1×(-0.02) = 0.8 - 0.002 = 0.798

[0197] The updated scheduling threshold makes the next scheduling decision closer to the actual demand.

[0198] Step 2.6, Integrate the model into the scheduling system

[0199] Take the prediction module as the core decision-making component of the License scheduling system: Call predict_load to calculate the load prediction value in each scheduling cycle. If the load prediction value exceeds the preset scheduling threshold T load , then trigger the License scheduling action (call the smart contract interface to allocate License). After scheduling, monitor the actual load in real time, calculate the error and update the model parameters to provide feedback for the next cycle decision.

[0200] Step 3, Distributed scheduling and real-time monitoring

[0201] Deploy lightweight edge nodes in each region, use local data to quickly respond to scheduling requirements, and keep data synchronized with the central server to achieve the combination of local decision-making and global optimization.

[0202] After scheduling, the system collects the actual load data every cycle T' (for example, 60 seconds) to monitor the License scheduling effect;

[0203] If the monitoring result shows that the network load drops or the problem improves, automatically call releaseLicense to roll back the License to the virtual pool. For example, if the actual load data is less than the scheduling threshold T load , and lasts for a period of M minutes, then roll back the License to the License pool, thus avoiding frequent scheduling and improving stability. Another example is that a rollback threshold can be set, which is less than the scheduling threshold. When the actual load data is less than the rollback threshold (or the actual load data is less than the rollback threshold and lasts for a period of time), roll back the License to the License pool, thus also being able to avoid frequent scheduling.

[0204] All operations are recorded on the blockchain in real time to achieve full traceability.

[0205] It should be noted that for License scheduling, the functions of the lightweight edge nodes include:

[0206] (1) Collect network key indicators of local base stations in real time (such as PRB utilization rate, RSRP, latency, packet loss, etc.), without relying on the central node, reducing collection latency;

[0207] (2) Integrate the prediction algorithm to predict the future load L pred (t + 1) at the local end;

[0208] (3) As lightweight blockchain nodes, edge nodes can directly call the intelligent contract interface on the chain for License scheduling, fallback, etc. The lightweight edge nodes are the key execution nodes for License scheduling, responsible for the whole process of "collection, judgment, request, feedback", and cooperate with the central node through the contract on the chain to achieve local fast scheduling and global trusted synchronization.

[0209] It should be noted that edge nodes can act as lightweight blockchain nodes, but not necessarily full nodes. Because edge nodes have limited capabilities, and full nodes of traditional blockchains need to synchronize the complete ledger and verify all historical transactions, with large resource overheads, which are not suitable for direct deployment on edge devices. The storage of the complete ledger and consensus verification can be undertaken by the central node or the core operator node.

[0210] Step 4, Manual intervention

[0211] Build a real-time monitoring large screen based on WEB to display the network status of each region (i.e., the indicators of each base station), License allocation situation, prediction trend, and scheduling records; provide a manual intervention interface for management personnel to adjust the scheduling strategy immediately when emergencies occur; at the same time, set up an alarm mechanism to notify relevant personnel in time for manual processing or review.

[0212] It should be noted that in this embodiment, by combining blockchain technology with intelligent prediction algorithms, transparent, automated management and dynamic scheduling of License resources are achieved. By constructing a secure virtual License pool and intelligent contracts, an immutable record of each resource allocation and fallback is ensured; at the same time, multi-dimensional network data and a weighted moving average model are used for real-time load prediction, and scheduling parameters are adaptively adjusted based on error feedback, so as to be able to respond to network changes proactively, improve resource utilization and service quality. This scheduling method that comprehensively applies cutting-edge technologies is significantly superior to traditional fixed-threshold schemes in a dynamic network environment, can greatly improve the overall efficiency of License resources while ensuring network stability, and effectively promotes the intelligent development process of the communication industry.

[0213] In another specific embodiment, the scheduling method of this License may include the following steps:

[0214] S1, Periodic data collection and prediction

[0215] For example, every T = 60 seconds, collect the metrics of the current base station and calculate the load L(t). Use the prediction model to calculate the load L in the next cycle. pred .

[0216] S2. Determine the scheduling requirement

[0217] If L pred ≥ T load , then call the smart contract interface allocateLicense to allocate a License to this base station.

[0218] The smart contract records the allocation information and changes the License status from "free" to "allocated". After scheduling, the system starts to record the allocation time and the expected load.

[0219] S3. Adaptive threshold update and monitoring feedback

[0220] Every T' (for example, T' = 60 seconds), the system collects the actual load L actual , calculates the error ΔL and updates the scheduling threshold T load . If it is detected that the network load drops or the load improves significantly after scheduling, automatically call releaseLicense to return the unnecessary License to the virtual pool.

[0221] S4. All scheduling actions are recorded through the blockchain smart contract to form an immutable operation log for subsequent auditing and optimization reference.

[0222] It should be noted that the License scheduling method provided by the embodiments of the present invention has the following characteristics:

[0223] a) Distributed smart contract mechanism based on blockchain: By constructing a virtual License pool, it realizes the transparency, automated recording and immutable evidence storage of the entire process of License resource allocation and return, ensuring multi-party trust and security.

[0224] b) Multi-dimensional data fusion and preprocessing: Comprehensively collect network performance metrics such as PRB utilization rate, latency, and packet loss rate. After data cleaning and normalization processing, a unified multi-dimensional data view is formed to provide comprehensive data support for load prediction.

[0225] c) Dynamic load prediction and adaptive scheduling strategy: Adopt a weighted moving average model to predict the network load in real time, and dynamically adjust the scheduling threshold and model weights through an error feedback mechanism to achieve forward-looking and accurate License scheduling and effectively cope with sudden load fluctuations.

[0226] d) Edge - center collaborative scheduling architecture mode: Deploy lightweight edge scheduling nodes in each region to achieve local fast response and real - time scheduling. At the same time, cooperate with the central server (i.e., the central node) to ensure the flexibility and efficiency of overall scheduling.

[0227] The License scheduling method provided by the embodiment of the present invention first predicts the load prediction value of the base station in the next cycle; then, in response to the load prediction value being greater than or equal to a preset scheduling threshold, calls the first interface of the smart contract to allocate an idle License from a pre - constructed License pool to the base station, and saves the record of this allocation to the blockchain; wherein, the License pool includes multiple sharable Licenses; and when it is detected that the actual load value of the base station meets the preset fallback condition, calls the second interface of the smart contract to fallback the allocated License to the License pool, and saves the record of this fallback to the blockchain. The present invention monitors and judges the load prediction value and the actual load value of the base station in real time. When the load prediction value reaches the scheduling threshold, it allocates a License from the License pool; and when the actual load value meets the fallback condition, it fallback the License. This method fully considers the differences between the busy and idle states of the base station, as well as the actual needs of co - construction and sharing, making up for the deficiencies of the existing fixed allocation method. It not only gives full play to and expands the advantages of co - construction and sharing, but also greatly improves the resource utilization rate and scheduling flexibility, significantly reduces the cost of operators purchasing Licenses, and effectively improves the network experience of users. It solves the problem that the existing License scheduling methods mainly adopt a fixed allocation method, without fully considering the differences between busy and idle times and the actual needs of co - construction and sharing, resulting in low resource utilization rate and poor flexibility, which not only greatly increases the cost of operators but also affects the network experience of users.

[0228] Embodiment 2:

[0229] As Figure 2 shown, this embodiment provides a License scheduling device for executing the above - mentioned License scheduling method, including:

[0230] A load prediction module 11, configured to predict the load prediction value of the base station in the next cycle;

[0231] A License allocation module 12, connected to the load prediction module 11, configured to, in response to the load prediction value being greater than or equal to a preset scheduling threshold, call the first interface of the smart contract to allocate an idle License from a pre - constructed License pool to the base station, and save the record of this allocation to the blockchain; wherein, the License pool includes multiple sharable Licenses;

[0232] The License fallback module 13, connected to the License allocation module 12, is used to call the second interface of the smart contract to fallback the allocated License to the License pool when it is detected that the actual load value of the base station meets the preset fallback condition, and save the record of this fallback to the blockchain.

[0233] Optionally, the device further includes:

[0234] The License pool construction module is used to construct a License pool including multiple sharable Licenses;

[0235] The License registration module is used to register multiple sharable Licenses on the blockchain;

[0236] The smart contract deployment module is used to deploy the smart contract to the blockchain.

[0237] Optionally, the load prediction module 11 includes:

[0238] The metric data acquisition unit is used to acquire the metric data of the base station in the current cycle and the previous N cycles; where N is an integer greater than or equal to 1;

[0239] The normalized load value acquisition unit is used to obtain the normalized network load values of the current cycle and the previous N cycles according to the metric data;

[0240] The model prediction unit is used to input the normalized network load values of the current cycle and the previous N cycles into a pre-constructed prediction model to obtain the load prediction value of the base station in the next cycle.

[0241] Optionally, the load prediction value is a single load prediction value or a comprehensive load prediction value. If the load prediction value is a single load prediction value, the metric data includes the Physical Resource Block (PRB) utilization rate, and the normalized load value acquisition unit is specifically used for:

[0242] Normalize the PRB utilization rates of the current cycle and the previous N cycles respectively to obtain the normalized network load values of the current cycle and the previous N cycles;

[0243] If the load prediction value is a comprehensive load prediction value, the metric data includes the PRB utilization rate, the average Reference Signal Receiving Power (RSRP), the latency, and the packet loss rate, and the normalized load value acquisition unit is specifically used for:

[0244] Normalize the PRB utilization rate, average RSRP, delay, and packet loss rate for the current cycle and the previous N cycles respectively, and perform weighted summation on the normalized PRB utilization rate, average RSRP, delay, and packet loss rate according to preset weights to obtain the normalized network load value for the current cycle and the previous N cycles.

[0245] Optionally, the License allocation module 12 is further configured to:

[0246] Modify the current status of the allocated License from idle to allocated, and record the scheduling timestamp and geographical location of the allocated License;

[0247] The License fallback module 13 is further configured to:

[0248] Modify the current status of the allocated License from allocated to idle, and clear the scheduling timestamp and geographical location.

[0249] Optionally, the fallback condition is one of the following:

[0250] The actual load value of the base station is less than the scheduling threshold and lasts for a preset time or a preset number of cycles;

[0251] The actual load value of the base station is less than a preset fallback threshold, where the fallback threshold is less than the scheduling threshold;

[0252] The actual load value of the base station is less than the fallback threshold and lasts for a preset time or a preset number of cycles.

[0253] Optionally, the device further includes:

[0254] A scheduling threshold update module, configured to update the scheduling threshold according to the difference between the actual load value and the load prediction value; and,

[0255] A model parameter adjustment module, configured to adjust the weight parameters of the prediction model according to the difference between the actual load value and the load prediction value.

[0256] Optionally, the scheduling threshold update module specifically updates the scheduling threshold according to the following formula:

[0257]

[0258] Where, represents the scheduling threshold, represents the updated scheduling threshold, k is a set learning rate, and ΔL is the difference between the actual load value and the load prediction value.

[0259] Example 3:

[0260] Reference Figure 3 , this embodiment provides a License scheduling device, including a memory 21 and a processor 22. A computer program is stored in the memory 21, and the processor 22 is configured to run the computer program to execute the License scheduling method in Embodiment 1.

[0261] Among them, the memory 21 is connected to the processor 22. The memory 21 can adopt flash memory, read-only memory or other memories, and the processor 22 can adopt a central processing unit or a single-chip microcomputer.

[0262] Example 4:

[0263] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the License scheduling method in Embodiment 1 above.

[0264] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules, or other data). The computer-readable storage medium includes, but is not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), flash memory or other memory technologies, CD-ROM (Compact Disc Read-Only Memory), digital versatile disc (DVD) or other optical disc storage, magnetic cassette, tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0265] In summary, the License scheduling method, device, and readable storage medium provided by the embodiments of the present invention first predict the load prediction value of the base station in the next cycle; then, in response to the load prediction value being greater than or equal to a preset scheduling threshold, call the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and save the record of this allocation to the blockchain; where the License pool includes multiple sharable Licenses; and when it is detected that the actual load value of the base station meets the preset fallback condition, call the second interface of the smart contract to fallback the allocated License to the License pool, and save the record of this fallback to the blockchain. The present invention monitors and judges the load prediction value and the actual load value of the base station in real time. When the load prediction value reaches the scheduling threshold, it allocates a License from the License pool; and when the actual load value meets the fallback condition, it fallback the License. This method fully considers the differences between busy and idle base stations, as well as the actual needs of co-construction and sharing, making up for the deficiencies of the existing fixed allocation method. It not only gives full play to and expands the advantages of co-construction and sharing, but also greatly improves resource utilization and scheduling flexibility, significantly reduces the cost for operators to purchase Licenses, and effectively improves the network experience of users. It solves the problem that the existing License scheduling methods mainly adopt a fixed allocation method, without fully considering the differences between busy and idle times and the actual needs of co-construction and sharing, resulting in low resource utilization and poor flexibility, which greatly increases the costs for operators and affects the network experience of users.

[0266] It can be understood that the above embodiments are merely exemplary embodiments adopted to illustrate the principles of the present invention, and the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also regarded as the protection scope of the present invention.

Claims

1. A scheduling method for a license, characterized in that, The method includes: Predicting the load prediction value of the base station in the next cycle; In response to the load prediction value being greater than or equal to a preset scheduling threshold, calling the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and saving the current allocation record to the blockchain; wherein, the License pool includes multiple sharable Licenses; When it is detected that the actual load value of the base station meets the preset fallback condition, calling the second interface of the smart contract to fallback the allocated License to the License pool, and saving the current fallback record to the blockchain.

2. The method according to claim 1, wherein Before predicting the load prediction value of the base station in the next cycle, the method further includes: Constructing a License pool including multiple sharable Licenses; Registering multiple sharable Licenses on the blockchain; Deploying the smart contract to the blockchain.

3. The method according to claim 1, characterized in that, Predicting the load prediction value of the base station in the next cycle specifically includes: Collecting the metric data of the base station in the current cycle and the previous N cycles; where N is an integer greater than or equal to 1; Obtaining the normalized network load values of the current cycle and the previous N cycles according to the metric data; Inputting the normalized network load values of the current cycle and the previous N cycles into a pre-constructed prediction model to obtain the load prediction value of the base station in the next cycle.

4. The method according to claim 3, characterized in that, The load prediction value is a single load prediction value or a comprehensive load prediction value. If the load prediction value is a single load prediction value, the metric data includes the Physical Resource Block (PRB) utilization rate. Obtaining the normalized network load values of the current cycle and the previous N cycles according to the metric data specifically includes: Normalizing the PRB utilization rates of the current cycle and the previous N cycles respectively to obtain the normalized network load values of the current cycle and the previous N cycles; If the load prediction value is a comprehensive load prediction value, the metric data includes the PRB utilization rate, the average Reference Signal Receiving Power (RSRP), the delay, and the packet loss rate. Obtaining the normalized network load values of the current cycle and the previous N cycles according to the metric data specifically includes: Normalizing the PRB utilization rate, the average RSRP, the delay, and the packet loss rate of the current cycle and the previous N cycles respectively, and performing weighted summation on the normalized PRB utilization rate, the average RSRP, the delay, and the packet loss rate according to a preset weight to obtain the normalized network load values of the current cycle and the previous N cycles.

5. The method according to claim 1, wherein After calling the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, the method further includes: Changing the current status of the allocated License from idle to allocated, and recording the scheduling timestamp and geographical location of the allocated License; After calling the second interface of the smart contract to fallback the allocated License to the License pool, the method further includes: Modify the current status of the allocated License from allocated to idle, and clear the scheduling timestamp and geographical location.

6. The method according to claim 1, wherein The fallback condition is one of the following: The actual load value of the base station is less than the scheduling threshold and persists for a preset time or a preset number of cycles; The actual load value of the base station is less than a preset fallback threshold, where the fallback threshold is less than the scheduling threshold; The actual load value of the base station is less than the fallback threshold and persists for a preset time or a preset number of cycles.

7. The method according to claim 3, wherein The method further includes: Updating the scheduling threshold according to the difference between the actual load value and the load prediction value; and Adjusting the weight parameters of the prediction model according to the difference between the actual load value and the load prediction value.

8. The method according to claim 7, wherein The updating of the scheduling threshold according to the difference between the actual load value and the load prediction value is specifically calculated according to the following formula: Among them, represents the scheduling threshold, represents the updated scheduling threshold, k is the set learning rate, and ΔL is the difference between the actual load value and the load prediction value.

9. A scheduling device for a license, characterized in that, The device includes: A load prediction module for predicting the load prediction value of the base station in the next cycle; A License allocation module connected to the load prediction module, configured to, in response to the load prediction value being greater than or equal to a preset scheduling threshold, call the first interface of the smart contract to allocate an idle License from a pre-constructed License pool to the base station, and save the record of this allocation to the blockchain; wherein the License pool includes a plurality of shareable Licenses; A License fallback module connected to the License allocation module, configured to, when it is detected that the actual load value of the base station satisfies a preset fallback condition, call the second interface of the smart contract to fallback the allocated License to the License pool, and save the record of this fallback to the blockchain.

10. A scheduling device for a license, characterized in that It includes a memory and a processor, and a computer program is stored in the memory. The processor is configured to run the computer program to implement the scheduling method of the License according to any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, it implements the scheduling method of the License according to any one of claims 1-8.

Images