Inter-server communication scheduling method and system based on CPU load balancing

Abstract

The application provides a CPU load balancing-based inter-server communication scheduling method and system, and relates to the technical field of CPU load balancing, wherein the method comprises: performing quality evaluation on network state data to generate strategy data, and meanwhile converting electromagnetic intensity signals and temperature distribution signals in a computer room into corresponding digital signals to form an environment data set; fusing the strategy data and the environment state data set through a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching instruction; screening a set of available CPU cores according to load data; and finally distributing data packets to target CPU cores for processing and transmission according to the frequency modulation scheme and the channel switching instruction, establishing multiple concurrent connection channels, and thereby realizing load balancing and high stability transmission between servers. The application improves the communication efficiency and transmission reliability of a server cluster.

Description

Technical Field

[0001] This application relates to the field of CPU load balancing technology, and in particular to a method and system for scheduling inter-server communication based on CPU load balancing. Background Technology

[0002] With the rapid development of cloud computing and real-time interactive applications, large-scale cloud service platforms need to handle high-concurrency, low-latency data transmission tasks, such as online video streaming and real-time multi-person collaborative editing. These applications require server clusters to have the ability to dynamically avoid network congestion while ensuring the transmission quality and stability of critical data streams, which places higher demands on the real-time performance, adaptability, and reliability of inter-server communication scheduling.

[0003] The existing solution adopts a dynamic scheduling mechanism based on network status prediction. Specifically, this solution monitors the computing load and network link status of server nodes in real time, and trains a prediction model based on historical data to predict the network congestion situation and server load trend in the future. Then, based on the prediction results, a multi-objective optimization algorithm is used to generate a scheduling strategy, which allocates data packets to server nodes with lower predicted load, and manages the data transmission order through a priority queue mechanism to achieve load balancing and communication quality assurance.

[0004] However, such solutions have limitations in adaptability to complex environments; moreover, their scheduling decisions rely on logical layer data of network and server status, ignoring the impact of actual data center environmental factors on communication quality; furthermore, since factors such as electromagnetic interference and temperature changes can directly lead to signal attenuation and transmission performance fluctuations, scheduling strategies based solely on logical state predictions may have decision-making biases in actual deployments, affecting transmission stability and resource scheduling accuracy. Summary of the Invention

[0005] This application provides a server communication scheduling method and system based on CPU load balancing to solve the problems of low communication efficiency and poor transmission reliability in the existing technology of server clusters.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a server communication scheduling method based on CPU load balancing, comprising: Collect data packets to be scheduled for transmission in the cloud service platform, CPU core load data, task quantity data, network status data, and electromagnetic intensity and temperature distribution signals in the computer room. The network status data is quality assessed, and policy data is generated based on the assessment results; The electromagnetic intensity signal and the temperature distribution signal are converted into corresponding analog electrical signals, and the analog electrical signals are sampled and quantized to generate an environmental state dataset. The strategy data and the environmental state dataset are fused and calculated using a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching command. Based on the load data and the number of tasks, determine the set of available CPU cores with a load level lower than a preset standard from the CPU cores of each server; Based on the frequency modulation scheme and the channel switching command, the data packet is allocated to the target CPU core in the set of available CPU cores for processing and transmission, so as to establish multiple concurrent communication connection channels. The concurrent communication connection channels are used to achieve load balancing among the target servers.

[0007] Optionally, the step of using a dynamic weight adjustment algorithm to fuse and calculate the policy data and the environmental state dataset to generate a frequency modulation scheme and channel switching instructions includes: Extract channel priority information from the strategy data; Based on the channel priority information, the initial priority weight of each network channel is determined; The environmental status dataset is analyzed, and the environmental interference coefficient corresponding to each network channel is calculated based on the analysis results. Based on the environmental interference coefficient, the initial priority weight of each network channel is dynamically adjusted using a dynamic weight adjustment algorithm. Based on the adjusted priority weights, frequency modulation schemes and channel switching commands are generated.

[0008] Optionally, generating the frequency modulation scheme and channel switching command based on the adjusted priority weights includes: The adjusted priority weights are normalized to a preset numerical range using a weight mapping function to generate a weight set. Based on the weight set, calculate the channel quality index of each network channel and generate a channel switching command; Based on the channel quality index, the corresponding carrier frequency, modulation depth, and symbol rate parameters are selected from a preset mapping table to form a modulation parameter set; Based on the set of modulation parameters, a frequency modulation scheme is generated through a parameter optimization algorithm.

[0009] Optionally, the step of calculating the channel quality index of each network channel and generating a channel switching command based on the weight set includes: Based on the weight values ​​of each network channel in the weight set, combined with network throughput data and signal attenuation parameters, the channel quality index of each network channel is calculated. Based on the channel quality index, each network channel is classified into quality levels, and a channel priority list is generated. Based on the channel priority list and combined with the preset handover triggering conditions, a channel handover sequence including the primary channel, backup channel and handover threshold is constructed. The channel switching sequence is subjected to a consistency check to generate a channel switching command.

[0010] Optionally, the step of performing a quality assessment on the network state data and generating policy data based on the assessment results includes: Extract the traffic rate sequence and packet loss rate sequence from the network status data; Based on the traffic rate sequence, calculate the throughput fluctuation value of each network channel; Based on the packet loss rate sequence, calculate the stability quantification value of each network channel; The throughput fluctuation value and the stability quantification value are weighted and fused to obtain the comprehensive quality score of each network channel; Based on the comprehensive quality score, each network channel is sorted in descending order to generate policy data containing channel priority information.

[0011] Optionally, the step of converting the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and performing sampling and quantization processing on the analog electrical signals to generate an environmental state dataset includes: The electromagnetic intensity signal is amplified to obtain a first analog electrical signal; Cold junction compensation is performed on the temperature distribution signal to obtain a second analog electrical signal; The first analog electrical signal and the second analog electrical signal are converted from analog to digital using a multi-channel synchronous sampling analog-to-digital converter to obtain a first digital signal and a second digital signal. The first digital signal and the second digital signal are respectively filtered. The processed first digital signal and the processed second digital signal are aligned and encapsulated according to the time sequence to generate an environmental state dataset.

[0012] Optionally, the step of allocating the data packet to a target CPU core in the set of available CPU cores for processing and transmission based on the frequency modulation scheme and the channel switching instruction, in order to establish multiple concurrent communication connection channels, includes: The transmission time window of the data packet is determined according to the modulation timing and parameter adjustment rules in the frequency modulation scheme; The optimal transmission channel is dynamically selected according to the channel switching sequence and usage priority in the channel switching instruction. Select the CPU core with the lowest load from the set of available CPU cores as the target CPU core; In the target CPU core, the data packet is subjected to protocol encapsulation and error verification processing; Based on the optimal transmission channel and the transmission time window, the processed data packet will simultaneously send multiple connection establishment requests to different target servers. After the target server responds to the connection establishment request and completes the handshake protocol, multiple concurrent communication connection channels are established between the target server and each target server.

[0013] Secondly, this application provides a server communication scheduling system based on CPU load balancing, comprising: The data acquisition module is used to collect data packets to be scheduled and transmitted in the cloud service platform, load data of CPU cores in each server, task quantity data, network status data, and electromagnetic intensity signals and temperature distribution signals in the computer room. The evaluation module is used to evaluate the quality of the network status data and generate policy data based on the evaluation results. The conversion module is used to convert the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and to perform sampling and quantization processing on the analog electrical signals to generate an environmental state dataset. The calculation module is used to perform fusion calculations on the policy data and the environmental state dataset using a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching command. The determination module is used to determine, based on the load data and the task quantity data, a set of available CPU cores with a load level lower than a preset standard from the CPU cores of each server; The allocation module is used to allocate the data packet to the target CPU core in the set of available CPU cores for processing and transmission based on the frequency modulation scheme and the channel switching instruction, so as to establish multiple concurrent communication connection channels, which are used to achieve load balancing among target servers.

[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor, used to execute the computer program, implements the steps of the CPU load balancing-based inter-server communication scheduling method as described in the first aspect above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the CPU load balancing-based inter-server communication scheduling method described in the first aspect above.

[0016] The technical solution provided in this application has the following beneficial effects: This application first achieves a comprehensive understanding of the server cluster's operating status and the data center environment, thus providing a multi-dimensional data foundation for subsequent scheduling decisions. Based on this, it accurately identifies the real-time communication quality of each network channel and formulates a preliminary scheduling strategy based on network status. Simultaneously, it quantifies environmental factors into computable data, providing data support for adaptive scheduling. This enables multi-dimensional collaborative decision-making based on network status and environmental factors, generating optimized scheduling instructions that adapt to complex environments. Furthermore, it effectively identifies available computing resources to avoid single-point overload, thereby ensuring overall processing efficiency. Finally, it achieves intelligent scheduling and transmission of data packets, improving the reliability and efficiency of inter-server communication.

[0017] Furthermore, this application also determines the initial weights by extracting the priority information of the network channel, dynamically adjusts the weights in combination with the environmental interference coefficient, and finally generates frequency modulation and channel switching commands.

[0018] Furthermore, by dynamically integrating network status and environmental data, this application can achieve real-time optimization of scheduling strategies, thereby improving the accuracy and environmental adaptability of communication scheduling, and ultimately enhancing the system's stable operation capability in complex environments.

[0019] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart illustrating a server communication scheduling method based on CPU load balancing, provided for an embodiment of this application; Figure 2 A schematic diagram illustrating a specific implementation of a server communication scheduling method based on CPU load balancing provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of an inter-server communication scheduling system based on CPU load balancing, provided in an embodiment of this application. Detailed Implementation

[0022] To address the problems existing in the prior art, this application proposes a server communication scheduling method based on CPU load balancing. This method first collects multi-source data, including server operating status, network quality, and electromagnetic environment and temperature distribution in the data center, and converts the environmental signals into a quantifiable digital state set. Then, a dynamic weighting algorithm is used to fuse this digital state set with network policy data to generate frequency modulation and channel switching commands. By considering both logical state and environmental factors in the scheduling decision, this method achieves real-time response to dynamic environmental changes, effectively reducing the impact of interference on transmission quality and ultimately improving the stability of server communication and the accuracy of resource scheduling in complex environments.

[0023] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The core of this application is to provide a server communication scheduling method based on CPU load balancing, and a flowchart of one specific implementation is shown below. Figure 1 As shown, the method includes: Step 101: Collect the data packets to be scheduled for transmission in the cloud service platform, the load data of the CPU cores in each server, the number of tasks, the network status data, and the electromagnetic intensity signal and temperature distribution signal in the computer room.

[0025] In step 101, the data packet to be scheduled for transmission represents a data unit in the cloud service platform that needs to be transmitted across servers. The data packet to be scheduled for transmission may contain user requests or inter-service communication content. The load data of the central processing unit (CPU) core represents the proportion of computing resources occupied by each CPU core in the current processing task, reflecting the real-time computing pressure.

[0026] Task quantity data represents the number of tasks waiting to be processed by CPU cores, reflecting the degree of task backlog; network status data includes the real-time traffic rate and packet loss rate of the network channel, characterizing the communication quality; the cloud service platform is the overall architecture that provides distributed computing services, each server is a specific computing node that constitutes the platform, and the data center is the infrastructure environment for deploying these servers and network devices.

[0027] The electromagnetic intensity signal represents the original analog signal of electromagnetic interference in the computer room, which affects the stability of signal transmission; the temperature distribution signal represents the original analog signal of temperature in the server's heat dissipation area, and high temperature may lead to a decrease in equipment performance.

[0028] For example, in a video stream processing scenario on a cloud service platform, the monitoring agent collects data showing that the load of CPU core A in the server cluster is 65% and the number of tasks is 8; the network probe detects that the traffic rate of channel 1 is 95Mbps and the packet loss rate is 0.3%; the electromagnetic sensor collects the electromagnetic intensity signal of area X of the cabinet as 0.2mV, and the temperature sensor collects the temperature signal of the area as 35mV; all data is packaged and sent to the dispatch center.

[0029] Step 102: Perform a quality assessment on the network status data and generate policy data based on the assessment results.

[0030] In step 102, the policy data representation includes intermediate decision data that prioritizes each network channel.

[0031] In this embodiment of the application, the traffic rate and packet loss rate in the network status data are calculated to obtain the signal quality evaluation value of each channel; then the channels are sorted according to the evaluation value to generate policy data containing priority information.

[0032] For example, based on the traffic rate of 95 Mbps and packet loss rate of 0.3% for channel 1 in step 101, a signal quality evaluation value is calculated, as exemplified by: Based on the score ranking, Channel 1 is marked as the highest priority, and the strategy data "Channel 1 Priority" is generated.

[0033] Step 103: Convert the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and perform sampling and quantization processing on the analog electrical signals to generate an environmental state dataset.

[0034] In step 103, the analog electrical signal represents the continuous voltage signal output by the sensor, the digital sampling represents the process of converting the analog signal into discrete digital values, the quantization encoding represents the numerical mapping and format standardization of the sampled values, and the environmental state dataset represents a data set containing digitally quantized values ​​of electromagnetic intensity and temperature.

[0035] In this embodiment, the electromagnetic intensity signal and temperature signal are conditioned by an amplifier and then synchronously sampled by an analog-to-digital converter to generate a digital sequence; the digital sequence is then filtered and encoded to generate a standardized environmental state dataset.

[0036] For example, the electromagnetic intensity signal of 0.2mV in step 101 is amplified 100 times to become 20mV, and the temperature signal of 35mV is compensated to become 37mV; then, each signal is converted from analog to digital by a 12-bit analog-to-digital converter. For example: Electromagnetic digital value = Temperature numerical value = Generate a dataset {electromagnetic intensity: 16, temperature: 30}.

[0037] Step 104: Using a dynamic weight adjustment algorithm, perform fusion calculations on the policy data and the environmental state dataset to generate a frequency modulation scheme and channel switching instructions.

[0038] In step 104, the frequency modulation scheme represents the configuration instructions that define the carrier frequency and modulation parameters for data transmission, and the channel switching instruction represents the control command that specifies the channel switching conditions and sequence.

[0039] In this embodiment, channel priority is extracted from policy data and environmental interference coefficient is calculated in combination with environmental data; then, environmental adaptive weights are generated through a weight adjustment algorithm; finally, frequency modulation parameters and channel switching sequences are generated based on the weights.

[0040] For example, in the strategy data, the priority of channel 1 is 0.6, and in the environmental data, the electromagnetic intensity of 16 corresponds to an interference coefficient of 0.016; the adjusted weight = 0.6 × (1 - 0.016) = 0.5904; then, based on the adjusted weight, the modulation parameters are generated: carrier frequency 2.4 GHz, modulation depth 75%, and channel switching instruction "main channel is 1, backup channel is 2".

[0041] Step 105: Based on the load data and the task quantity data, determine the set of available CPU cores with a load level lower than a preset standard from the CPU cores of each server.

[0042] In step 105, the preset standard refers to quantitative indicators such as CPU utilization threshold and processing queue length limit set according to real-time load data, which can be represented by a set of CPU cores that meet the load standard.

[0043] In this embodiment, the comprehensive load index of each CPU core is first calculated. This index can be obtained by combining real-time load and queue depth. Then, the comprehensive load index is compared with a preset threshold to filter out low-load CPU cores and form a set.

[0044] For example, CPU core A has a load of 65% and a queue depth of 8. The overall load index is 65% × 0.7 + 8 × 0.3 = 45.5 + 2.4 = 47.9. The preset threshold is 50, so core A can be classified into the available set.

[0045] Step 106: Based on the frequency modulation scheme and the channel switching instruction, the data packet is allocated to the target CPU core in the set of available CPU cores for processing and transmission, so as to establish multiple concurrent communication connection channels. The concurrent communication connection channels are used to realize load balancing among the target servers.

[0046] In step 106, the target CPU core refers to the specific processing core selected from the available set, and the concurrent communication connection channel refers to multiple data transmission paths established simultaneously.

[0047] In this embodiment, the data packet transmission frequency is adjusted according to the frequency modulation parameters, and the corresponding transmission channel is selected according to the channel switching instruction; at the same time, the CPU core with the lowest load is selected from the set of available CPU cores to process the data, and concurrent connections are established through a multi-threading mechanism.

[0048] For example, the data packet is assigned to CPU core A for processing and sent at a carrier frequency of 2.4 GHz with modulation parameters; it is transmitted to the target server B through channel 1, while three concurrent connection channels are established.

[0049] This method generates adaptive scheduling instructions by acquiring multi-dimensional data and sensing the environment, and dynamically fusing network strategies and environmental states. These instructions enable load balancing and high-stability transmission in inter-server communication, and effectively improve system response speed and resource utilization in complex environments.

[0050] To address the issue that existing scheduling methods do not adequately consider the impact of environmental factors on communication quality, some embodiments include step 104: utilizing a dynamic weight adjustment algorithm to fuse the policy data and the environmental state dataset to generate a frequency modulation scheme and channel switching instructions, such as... Figure 2 As shown, it includes: Step 201: Extract channel priority information from the strategy data.

[0051] In step 201, the channel priority information refers to the priority order list of each network channel generated after quality assessment of the network status data. This list reflects the channel ranking based on network performance.

[0052] In this embodiment of the application, the policy data file is parsed, and the score value and corresponding sorting result of each network channel recorded therein are read to obtain a channel order list arranged from high to low quality.

[0053] Step 202: Determine the initial priority weight of each network channel based on the channel priority information.

[0054] In step 202, the initial priority weight refers to the initial importance value assigned to the channel based on its position in the priority ranking, where a higher weight value indicates that the channel has a higher priority in scheduling.

[0055] In this embodiment, firstly, weights are assigned to each channel in a linearly decreasing manner according to the channel's ranking, with the first-ranked channel receiving the highest weight, and the weights of subsequent channels decreasing sequentially, forming an initial weight set.

[0056] Step 203: Analyze the environmental state dataset and calculate the environmental interference coefficient corresponding to each network channel based on the analysis results.

[0057] In step 203, the environmental interference coefficient is a parameter that quantifies the degree of influence of environmental factors on the transmission quality of the network channel. This coefficient combines the negative impacts of electromagnetic interference and temperature changes on signal transmission.

[0058] In this embodiment of the application, the digital signal values ​​in the environmental status dataset are read, and then the electromagnetic intensity value and temperature value are substituted into the preset calculation formula to obtain the interference coefficient corresponding to each cabinet area. Then, the interference coefficient is mapped to the network channel in its area.

[0059] Step 204: Based on the environmental interference coefficient, dynamically adjust the initial priority weight of each network channel using a dynamic weight adjustment algorithm.

[0060] In this embodiment, the initial priority weight of each channel and its environmental interference coefficient are input into the weight adjustment function. The function will reduce the weight value according to the magnitude of the interference coefficient and output the adjusted weight set.

[0061] Step 205: Generate frequency modulation scheme and channel switching command based on the adjusted priority weights.

[0062] In this embodiment, the channels are reordered according to the adjusted weight values. The channel with the highest weight is assigned the optimal modulation parameters and set as the main channel, while the remaining channels are selected as alternatives and their corresponding switching thresholds are set, thereby forming a complete scheduling instruction.

[0063] Here is a specific example: In the video stream processing scenario of the cloud service platform, based on the fact that channel 1 has the highest priority in the policy data, its initial priority weight is determined to be 0.6. Then, the first digital signal value after centralized processing of the environmental state data is 16 and the second digital signal value after processing is 30. Based on the analysis results, the environmental interference coefficient calculation formula is used. Calculate the coefficients, where Indicates the environmental interference coefficient. Indicates the value of the first digital signal. Represents the second digital signal value, where , and All values ​​are dimensionless; substituting them into the calculation yields... ; Then, based on the environmental interference coefficient, the weight correction formula in the dynamic weight adjustment algorithm is used. Perform weight adjustment, where Indicates the adjusted weights. Indicates the initial weights. Represents the environmental interference coefficient, where , All values ​​are dimensionless; substituting them into the calculation yields... ; Finally, a frequency modulation scheme is generated based on the adjusted weight of 0.2304, and the carrier frequency is set to 2.4 GHz and the modulation depth to 75%. At the same time, a channel switching command is generated, specifying the main channel as channel 1 and switching to the backup channel 2 when the quality index is lower than 0.35, thus completing the fusion calculation process.

[0064] In this embodiment of the application, by dynamically fusing network status and environmental data, the scheduling strategy can respond to environmental changes in real time, thereby improving the stability and transmission efficiency of communication links in complex environments, and thus achieving more accurate resource scheduling.

[0065] To further improve the accuracy and environmental adaptability of scheduling instruction generation, in some embodiments, step 205: generating frequency modulation schemes and channel switching instructions based on adjusted priority weights, includes: Step 301: Normalize the adjusted priority weights to a preset numerical range using a weight mapping function to generate a weight set.

[0066] In step 301, the weight mapping function is a mathematical function that linearly transforms weight values ​​of different ranges to a unified standard range. The preset numerical range is a predefined standardized range, and the weight set is a data set containing standardized weight values ​​of all network channels.

[0067] In this embodiment, inputting the adjusted weights of each network channel into a linear transformation function enables all weight values ​​to be mapped proportionally to the range of 0 to 1, while maintaining the original relative size relationship, thereby forming a standardized weight set.

[0068] Step 302: Based on the weight set, calculate the channel quality index of each network channel and generate a channel switching command.

[0069] In step 302, the channel quality index is a channel quality evaluation value calculated by combining the weight value and the real-time network performance.

[0070] In this embodiment, the weight value of each channel in the weight set is taken and multiplied by the real-time network quality coefficient of that channel to obtain the channel quality index. Then, the channel priority list is generated by sorting according to the index size. Finally, the instruction containing the primary channel, backup channel and handover threshold is formed by combining the preset handover conditions.

[0071] Step 303: Based on the channel quality index, select the corresponding carrier frequency, modulation depth, and symbol rate parameters from the preset mapping table to form a modulation parameter set.

[0072] In step 303, the mapping table is a pre-established data table that corresponds to the channel quality index and the modulation parameters, and the modulation parameter set is a set containing the modulation parameter configurations for each channel.

[0073] In this embodiment, firstly, the corresponding carrier frequency, modulation depth, and symbol rate parameters are found by querying a preset mapping table based on the channel quality index value of each channel, and then the modulation parameters of all channels are organized into a structured data table.

[0074] Step 304: Based on the set of modulation parameters, generate a frequency modulation scheme through a parameter optimization algorithm.

[0075] In the embodiments of this application, by analyzing the parameter configuration in the modulation parameter set and fine-tuning the parameters in combination with the real-time network load, a complete scheme containing detailed modulation parameters for each channel and adaptive adjustment rules can be generated.

[0076] Here is a specific example: In video stream processing scenarios on cloud service platforms, based on the adjusted weight of 0.2304, the weight of channel 1 is normalized to the interval between 0 and 1 using a weight mapping function. This weight mapping function can employ a linear scaling transformation formula. ,in Indicates the normalized weights. This represents the adjusted weights; both are dimensionless values. Substituting them into the calculation yields... The weight set channel 1 is generated with a weight of 0.606. Then, the channel quality index is calculated based on this set, specifically using the formula... ,in Indicates the channel quality index. This represents the real-time network quality coefficient, both of which are dimensionless values. Taking the real-time network quality coefficient for channel 1 as 0.85, the calculation is as follows: ; Next, a channel switching command is generated based on the index 0.515, and the main channel is set as channel 1 with a switching threshold of 0.35. Then, based on the channel quality index 0.515, a preset mapping table is queried to select the corresponding carrier frequency 2.4GHz, modulation depth 75%, and symbol rate 10Mbps to form a modulation parameter set. Finally, based on this table, a dynamic adjustment rule is added through a parameter optimization algorithm. When the network latency exceeds 50 milliseconds, the symbol rate is increased to 12Mbps to generate the final frequency modulation scheme and complete the command generation process.

[0077] In this embodiment, the generated modulation scheme and switching instructions can accurately reflect the actual network status through weight standardization, quality index calculation, and parameter optimization processes, thereby improving the accuracy of communication scheduling and environmental adaptability, and ultimately ensuring the stability of data transmission.

[0078] To further improve the accuracy of channel quality assessment and the reliability of handover control, in some embodiments, step 302: calculating the channel quality index of each network channel based on the weight set and generating a channel handover command includes: Step 401: Calculate the channel quality index of each network channel based on the weight values ​​of each network channel in the weight set, combined with network throughput data and signal attenuation parameters.

[0079] In step 401, the weight values ​​in the weight set are the final weight results after dynamic adjustment of the environmental interference coefficient, while the priority weights before adjustment are the initial weights without considering environmental impact.

[0080] Network throughput data is a real-time data transmission rate indicator collected from the network interface controller. Signal attenuation parameter is a quantitative value of signal loss calculated by analyzing changes in electromagnetic radiation intensity in the environmental state dataset. Channel quality index is a comprehensive evaluation index of channel quality calculated by combining weight values, throughput, and signal attenuation.

[0081] In this embodiment of the application, the weight value of each network channel in the weight set is taken, multiplied by the real-time throughput coefficient of the channel, and then divided by the signal attenuation coefficient to obtain the channel quality index value of each channel. The real-time throughput coefficient is calculated by the ratio of the current rate to the maximum theoretical rate, while the signal attenuation coefficient can be calculated by environmental monitoring data.

[0082] Step 402: Divide each network channel into quality levels according to the channel quality index and generate a channel priority list.

[0083] In step 402, the quality level division refers to dividing the channel into different quality levels according to the magnitude of the index, and the channel priority list is a list of channels arranged in order of quality from high to low.

[0084] In this embodiment of the application, the channel quality index of each network channel is compared with a preset quality level threshold to determine the quality level of each channel. Then, the channels are sorted according to the level and index size to generate a priority list containing channel number and quality level.

[0085] Step 403: Based on the channel priority list and the preset handover triggering conditions, construct a channel handover sequence that includes the primary channel, the backup channel, and the handover threshold.

[0086] In step 403, the handover triggering condition refers to the network state change condition that triggers channel handover, and the channel handover sequence is an ordered set of instructions that specifies the relationship between primary and backup channels and the handover timing.

[0087] In this embodiment of the application, the channel with the highest quality is selected from the channel priority list as the primary channel, and the channel with the second highest quality is selected as the backup channel; then, a switching threshold is set according to historical operation data, thereby constructing a switching sequence that includes primary and backup channel identifiers and switching condition parameters.

[0088] Step 404: Perform a consistency check on the channel switching sequence to generate a channel switching command.

[0089] In this embodiment of the application, it is first checked whether the primary channel and the backup channel in the channel switching sequence belong to the same link, and it is verified whether the switching conditions are mutually exclusive and complete. Conflict detection refers to checking whether the primary channel and the backup channel share a physical link, while logical verification confirms that the switching conditions meet the mutual exclusion requirement, that is, only one condition can be triggered at the same time. After passing the detection and verification, an instruction containing specific switching parameters and execution conditions is generated. For example, when it is detected that the primary channel 1 and the backup channel 3 share the same fiber optic link, the collision detection marks the combination as illegal. At the same time, the logic verification ensures that the two conditions of "signal quality below the threshold" and "delay exceeding the limit" will not take effect at the same time. The final generated instruction clearly stipulates that when the quality index of the primary channel 1 is lower than 0.35, the switch is made to the backup channel 2. If the backup channel 2 is unavailable, the redundancy mode is enabled and the switch to the backup channel 3 is prohibited.

[0090] Here is a specific example: In video stream processing scenarios on cloud service platforms, the weight value of channel one in the weight set is used. It is 0.606, combined with the real-time network quality coefficient. The value is 0.85 (this coefficient is calculated from the ratio of the real-time network throughput of 85Mbps to the maximum theoretical throughput of 100Mbps), using the channel quality index calculation formula. The channel quality index of channel one is calculated. Then, based on the index value, the quality levels of each network channel are divided. With an index greater than 0.45 set as a high-quality channel, a channel priority list is generated, in which channel one has the highest priority. Based on this list and preset switching trigger conditions, such as network latency exceeding 50 milliseconds or packet loss rate exceeding 1%, a channel switching sequence is constructed. This sequence designates channel one as the primary channel and channel two as the backup channel, with a switching threshold of [value missing]. The value is below 0.35; collision detection is performed on the sequence to confirm that there is no link overlap between the primary and backup channels, and logical verification ensures that the switching conditions are mutually exclusive and complete. The final generated channel switching instruction includes the following: the primary channel is channel one, and the backup channel is channel two. When the value is below 0.35 or the network latency exceeds 50 milliseconds, a switching operation is performed to complete the instruction generation process.

[0091] In this embodiment, through multi-parameter comprehensive calculation and hierarchical processing, combined with conflict detection and logical verification, the generated channel switching command can accurately reflect the network quality status, thereby ensuring the reliability and execution efficiency of the switching decision, and ultimately improving the overall stability of the communication system.

[0092] To further improve the accuracy of network state assessment and the effectiveness of policy generation, in some embodiments, step 102: performing a quality assessment on the network state data and generating policy data based on the assessment results includes: Step 501: Extract the traffic rate sequence and packet loss rate sequence from the network status data.

[0093] In step 501, the traffic rate sequence refers to the sequence of data transmission rate values ​​of each network channel at consecutive time points in chronological order, and the packet loss rate sequence refers to the sequence of data packet loss ratios of each network channel at corresponding time points in chronological order.

[0094] In this embodiment of the application, network status data packets are parsed to separate rate monitoring data and packet loss monitoring data for each network channel, and then sorted according to timestamps to form rate sequences and packet loss sequences.

[0095] Step 502: Calculate the throughput fluctuation value of each network channel based on the traffic rate sequence.

[0096] In step 502, the network channel refers to the communication path for data transmission between servers in the cloud service platform. These channels are built on the network infrastructure in the data center and connect the servers through routers, switches and network links to form the communication carrier for data transmission between servers. The throughput fluctuation value is an indicator that quantifies the stability of the network channel transmission rate and reflects the degree of drastic rate changes.

[0097] In this embodiment of the application, the traffic rate sequence of each network channel is first statistically analyzed, and then the average value of the rate difference between adjacent time points in the sequence is calculated based on the traffic rate sequence to obtain the throughput fluctuation value.

[0098] Step 503: Based on the packet loss rate sequence, calculate the stability quantification value of each network channel.

[0099] In step 503, the stability quantification value is a parameter for evaluating the reliability of network channel transmission, reflecting the severity of packet loss.

[0100] In this embodiment of the application, the stability quantification value is obtained by counting the proportion of times the packet loss rate in the sequence exceeds a preset threshold to the total number of times.

[0101] Step 504: Weight and fuse the throughput fluctuation value and the stability quantification value to obtain the comprehensive quality score of each network channel.

[0102] In step 504, the overall quality score is a quantitative value that reflects the overall quality of the network channel, which is the evaluation result.

[0103] In this embodiment of the application, weighting coefficients are assigned to the throughput fluctuation value and the stability quantification value respectively. Then, the results of multiplying the two index values ​​with their corresponding weighting coefficients are added together to obtain the comprehensive quality score of each channel.

[0104] Step 505: Based on the comprehensive quality score, sort each network channel in descending order to generate policy data containing channel priority information.

[0105] In this embodiment of the application, the overall quality scores of each network channel are compared, and the channels are sorted in descending order of scores to generate policy data containing channel number and corresponding priority.

[0106] Here is a specific example: In the video stream processing scenario of the cloud service platform, the traffic rate sequence [95, 93, 97, 94, 96] at the most recent five time points of channel one is extracted based on network status data, with units of Mbps, and the packet loss rate sequence [0.3%, 0.4%, 0.2%, 0.3%, 0.3%]. The throughput fluctuation value is calculated based on the traffic rate sequence using the formula... ,in This represents the throughput fluctuation value, in Mbps. Indicates the first The rate at each time point, in Mbps. This represents the sequence length, a dimensionless value, which is obtained by substituting it into the calculation. ; Then, a stability quantification value is calculated based on the packet loss rate sequence. For example, the formula is used. ,in The value represents the stability quantification and is a dimensionless numerical value. Indicates packet loss rate Exceeding the threshold The number of times, for example, T=0.35%, The value represents the sequence length and is dimensionless. Substituting this value into the calculation, we find that the number of times the threshold is exceeded is 1. Therefore... The throughput fluctuation value of 2.75 and the stability quantification value of 0.2 are weighted and fused using the formula. ,in This indicates the overall quality score. This represents the throughput fluctuation value, in Mbps. This indicates the weight of the fluctuation value, such as a value of 0.6. Represents the stability quantization value. This represents the stability weight, such as a value of 0.4, and , , and All values ​​are dimensionless; substituting them into the calculation yields... ; Based on the overall quality score of 91.5, the network channels are sorted in descending order. Since channel one has the highest score, policy data containing channel priority information is generated to designate channel one as the highest priority channel.

[0107] In this embodiment, the quality score generated by multi-dimensional index calculation and weighted fusion can comprehensively reflect the operating status of the network channel. The resulting priority ranking provides an accurate basis for subsequent scheduling decisions, thereby improving the rationality of network resource allocation.

[0108] To address the issues of accuracy and data standardization in environmental signal acquisition, in some embodiments, step 103 involves converting the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and then performing sampling and quantization processing on the analog electrical signals to generate an environmental state dataset, including: Step 601: Amplify the electromagnetic intensity signal to obtain a first analog electrical signal.

[0109] In step 601, the high-impedance differential amplifier is a circuit module used for the precise measurement of weak voltage signals. It is deployed in the signal acquisition unit in the computer room and its function is to amplify the original electromagnetic intensity signal acquired by the electromagnetic field probe. This device is a component of the computer room environmental monitoring infrastructure and is connected to the cloud service platform and servers through a data acquisition network.

[0110] The first analog electrical signal is a voltage signal with increased intensity.

[0111] In this embodiment, the raw weak voltage signal collected by the electromagnetic sensor is sent to a high-impedance differential amplifier, where it is amplified by a factor through the internal circuitry of the amplifier, thereby outputting an analog voltage signal with increased strength and easier processing.

[0112] Step 602: Perform cold junction compensation on the temperature distribution signal to obtain a second analog electrical signal.

[0113] In step 602, the thermocouple signal conditioning circuit is a dedicated circuit for processing thermocouple output signals. It is deployed in the distributed temperature acquisition nodes in the computer room. Its function is to compensate and standardize the raw signals output by the temperature sensor. This circuit is the underlying hardware unit of the computer room environment monitoring system and transmits data with the server and cloud service platform through the monitoring network.

[0114] In this embodiment, the original signal output by the temperature sensor is first corrected for ambient temperature error by a cold junction compensation circuit, and then converted into a voltage signal that is linearly related to temperature by a linearization circuit, resulting in a more accurate temperature simulation signal.

[0115] Step 603: Use a multi-channel synchronous sampling analog-to-digital converter to perform analog-to-digital conversion on the first analog electrical signal and the second analog electrical signal to obtain a first digital signal and a second digital signal.

[0116] In step 603, the multi-channel synchronous sampling analog-to-digital converter is an integrated circuit that can simultaneously acquire multiple analog signals and convert them into digital signals. It is deployed in the data acquisition equipment of the data center environment monitoring system. Its function is to synchronously digitize the analog signals from the amplifier and conditioning circuit. This device is a bridge connecting the data center environment and the data processing layer of the cloud service platform, converting the signals into digital data that can be analyzed by the platform.

[0117] In the embodiments of this application, the amplified electromagnetic signal and the second signal are simultaneously input into different channels of the analog-to-digital converter, and synchronously sampled and converted under unified clock control to generate a time-aligned digital signal sequence.

[0118] Step 604: Filter the first digital signal and the second digital signal respectively.

[0119] In this embodiment, the digital signal is first passed through a digital filter to remove high-frequency noise generated during sampling, and then through a notch filter to remove power supply frequency and its harmonic interference, resulting in a clean digital signal.

[0120] Step 605: Align and encapsulate the processed first digital signal and the processed second digital signal according to the time sequence to generate an environmental state dataset.

[0121] In this embodiment of the application, the same time stamp is added to the processed electromagnetic and second digital signals, and then the two types of data are combined into a structured dataset according to a predetermined format. The structured dataset contains three fields: timestamp, electromagnetic intensity value, and temperature value.

[0122] In this embodiment, the original signal is converted into a high-precision standardized digital dataset through signal amplification and conditioning, analog-to-digital conversion, and digital processing, providing an accurate and reliable data foundation for environmental perception, thereby ensuring the accuracy of subsequent scheduling decisions.

[0123] To further improve the efficiency and reliability of data transmission, in some embodiments, step 106: based on the frequency modulation scheme and the channel switching instruction, allocating the data packet to a target CPU core in the set of available CPU cores for processing and transmission, in order to establish multiple concurrent communication connection channels, includes: Step 701: Determine the transmission time window of the data packet according to the modulation timing and parameter adjustment rules in the frequency modulation scheme.

[0124] In step 701, the modulation timing and parameter adjustment rules define the time sequence of carrier frequency changes during data packet transmission and the dynamic adjustment rules of parameters such as modulation depth and symbol rate; the transmission time window is a specific time period during which data packets are allowed to be transmitted.

[0125] In this embodiment of the application, the modulation time period and parameter change rules specified in the frequency modulation scheme are analyzed, and the most suitable time interval for data packet transmission is calculated in combination with the current network load, thereby determining the specific transmission time window.

[0126] Step 702: Dynamically select the optimal transmission channel according to the channel switching sequence and usage priority in the channel switching instruction.

[0127] In step 702, the channel switching sequence is the order in which each channel is used, the usage priority is the level of importance of the channel, and the optimal transmission channel is the communication channel with the best quality in the current environment.

[0128] In this embodiment of the application, the quality status of each channel is monitored in real time according to the primary and backup channel order and priority identifier specified in the channel switching instruction, and the currently available channel with the highest quality is selected as the data transmission channel.

[0129] Step 703: Select the CPU core with the lowest load from the set of available CPU cores as the target CPU core.

[0130] In step 703, the target CPU core is the specific core that is ultimately selected to process the data packet.

[0131] In this embodiment of the application, the real-time load index of each core in the available CPU core set is queried, the load size of each core is compared, and the core with the lightest current load is selected as the data packet processing unit.

[0132] Step 704: In the target CPU core, the data packet is encapsulated and error checked.

[0133] In step 704, protocol encapsulation adds header information required for data transmission, and error checking uses algorithms to detect and correct data transmission errors.

[0134] In this embodiment, a processing program is run on a selected target CPU core, and a protocol header and checksum are added to the original data packet, thereby completing the data format standardization and error detection functions.

[0135] Step 705: Based on the optimal transmission channel and the transmission time window, simultaneously send multiple connection establishment requests to different target servers using the processed data packets.

[0136] In step 705, the connection establishment request is initialization information that requests the establishment of a communication connection.

[0137] In this embodiment of the application, within a determined transmission time window, data packets containing connection parameters are sent in parallel to multiple target servers through the selected optimal transmission channel to request the establishment of a communication connection.

[0138] Step 706: After the target server responds to the connection establishment request and completes the handshake protocol, multiple concurrent communication connection channels are established between the target server and each target server.

[0139] In step 706, the handshake protocol is the process by which two communicating parties exchange confirmation information when establishing a connection.

[0140] In this embodiment of the application, the system waits for the target server to confirm the connection request in order to verify the connection parameters through a three-way handshake process, and finally establishes multiple simultaneously active data transmission channels between the source server and each target server.

[0141] In the embodiments of this application, efficient scheduling and transmission of data packets are achieved through precise time control, channel selection and parallel processing, and a stable multi-path communication connection can be established, thereby improving the throughput and reliability of data transmission.

[0142] Figure 3A schematic diagram of a server communication scheduling system based on CPU load balancing, provided in an embodiment of this application, is shown. The system includes: The acquisition module 31 is used to acquire data packets to be scheduled and transmitted in the cloud service platform, load data of CPU cores in each server, task quantity data, network status data, and electromagnetic intensity signals and temperature distribution signals in the computer room.

[0143] Evaluation module 32 is used to evaluate the quality of the network status data and generate policy data based on the evaluation results.

[0144] The conversion module 33 is used to convert the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and to perform sampling and quantization processing on the analog electrical signals to generate an environmental state dataset.

[0145] The calculation module 34 is used to perform fusion calculations on the policy data and the environmental state dataset using a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching command.

[0146] The determination module 35 is used to determine, based on the load data and the task quantity data, a set of available CPU cores with a load level lower than a preset standard from the CPU cores of each server.

[0147] The allocation module 36 is used to allocate the data packet to the target CPU core in the set of available CPU cores for processing and transmission based on the frequency modulation scheme and the channel switching instruction, so as to establish multiple concurrent communication connection channels, which are used to achieve load balancing among the target servers.

[0148] The CPU load balancing-based inter-server communication scheduling system of this application embodiment is used to implement the aforementioned CPU load balancing-based inter-server communication scheduling method. Therefore, the specific implementation of the CPU load balancing-based inter-server communication scheduling system can be found in the embodiment section of the CPU load balancing-based inter-server communication scheduling method above. The specific implementation can be referred to the description of the corresponding embodiments, and will not be repeated here.

[0149] This application also provides an electronic device, including: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of any of the above-described CPU load balancing-based inter-server communication scheduling methods.

[0150] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described CPU load balancing-based inter-server communication scheduling methods.

[0151] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0152] This application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the steps in any of the above embodiments of the CPU load balancing-based inter-server communication scheduling method.

[0153] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0154] The foregoing has provided a detailed description of a server communication scheduling method and system based on CPU load balancing provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A server communication scheduling method based on CPU load balancing, characterized in that, include: Collect data packets to be scheduled for transmission in the cloud service platform, CPU core load data, task quantity data, network status data, and electromagnetic intensity and temperature distribution signals in the computer room. The network status data is quality assessed, and policy data is generated based on the assessment results; The electromagnetic intensity signal and the temperature distribution signal are converted into corresponding analog electrical signals, and the analog electrical signals are sampled and quantized to generate an environmental state dataset. The strategy data and the environmental state dataset are fused and calculated using a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching command. Based on the load data and the number of tasks, determine the set of available CPU cores with a load level lower than a preset standard from the CPU cores of each server; Based on the frequency modulation scheme and the channel switching command, the data packet is allocated to the target CPU core in the set of available CPU cores for processing and transmission, so as to establish multiple concurrent communication connection channels. The concurrent communication connection channels are used to achieve load balancing among the target servers.

2. The method according to claim 1, characterized in that, The step of using a dynamic weight adjustment algorithm to fuse and calculate the policy data and the environmental state dataset to generate a frequency modulation scheme and channel switching instructions includes: Extract channel priority information from the strategy data; Based on the channel priority information, the initial priority weight of each network channel is determined; The environmental status dataset is analyzed, and the environmental interference coefficient corresponding to each network channel is calculated based on the analysis results. Based on the environmental interference coefficient, the initial priority weight of each network channel is dynamically adjusted using a dynamic weight adjustment algorithm. Based on the adjusted priority weights, frequency modulation schemes and channel switching commands are generated.

3. The method according to claim 2, characterized in that, The generation of frequency modulation schemes and channel switching instructions based on the adjusted priority weights includes: The adjusted priority weights are normalized to a preset numerical range using a weight mapping function to generate a weight set. Based on the weight set, calculate the channel quality index of each network channel and generate a channel switching command; Based on the channel quality index, the corresponding carrier frequency, modulation depth, and symbol rate parameters are selected from a preset mapping table to form a modulation parameter set; Based on the set of modulation parameters, a frequency modulation scheme is generated through a parameter optimization algorithm.

4. The method according to claim 3, characterized in that, The step of calculating the channel quality index for each network channel and generating a channel switching instruction based on the weight set includes: Based on the weight values ​​of each network channel in the weight set, combined with network throughput data and signal attenuation parameters, the channel quality index of each network channel is calculated. Based on the channel quality index, each network channel is classified into quality levels, and a channel priority list is generated. Based on the channel priority list and combined with the preset handover triggering conditions, a channel handover sequence including the primary channel, backup channel and handover threshold is constructed. The channel switching sequence is subjected to a consistency check to generate a channel switching command.

5. The method according to claim 1, characterized in that, The step of performing a quality assessment on the network state data and generating policy data based on the assessment results includes: Extract the traffic rate sequence and packet loss rate sequence from the network status data; Based on the traffic rate sequence, calculate the throughput fluctuation value of each network channel; Based on the packet loss rate sequence, calculate the stability quantification value of each network channel; The throughput fluctuation value and the stability quantification value are weighted and fused to obtain the comprehensive quality score of each network channel; Based on the comprehensive quality score, each network channel is sorted in descending order to generate policy data containing channel priority information.

6. The method according to claim 1, characterized in that, The step of converting the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and then sampling and quantizing the analog electrical signals to generate an environmental state dataset includes: The electromagnetic intensity signal is amplified to obtain a first analog electrical signal; Cold junction compensation is performed on the temperature distribution signal to obtain a second analog electrical signal; The first analog electrical signal and the second analog electrical signal are converted from analog to digital using a multi-channel synchronous sampling analog-to-digital converter to obtain a first digital signal and a second digital signal. The first digital signal and the second digital signal are respectively filtered. The processed first digital signal and the processed second digital signal are aligned and encapsulated according to the time sequence to generate an environmental state dataset.

7. The method according to claim 1, characterized in that, Based on the frequency modulation scheme and the channel switching command, the data packets are allocated to target CPU cores in the set of available CPU cores for processing and transmission, thereby establishing multiple concurrent communication connection channels, including: The transmission time window of the data packet is determined according to the modulation timing and parameter adjustment rules in the frequency modulation scheme; The optimal transmission channel is dynamically selected according to the channel switching sequence and usage priority in the channel switching instruction. Select the CPU core with the lowest load from the set of available CPU cores as the target CPU core; In the target CPU core, the data packet is subjected to protocol encapsulation and error verification processing; Based on the optimal transmission channel and the transmission time window, the processed data packet will simultaneously send multiple connection establishment requests to different target servers. After the target server responds to the connection establishment request and completes the handshake protocol, multiple concurrent communication connection channels are established between the target server and each target server.

8. A server communication scheduling system based on CPU load balancing, characterized in that, include: The data acquisition module is used to collect data packets to be scheduled and transmitted in the cloud service platform, load data of CPU cores in each server, task quantity data, network status data, and electromagnetic intensity signals and temperature distribution signals in the computer room. The evaluation module is used to evaluate the quality of the network status data and generate policy data based on the evaluation results. The conversion module is used to convert the electromagnetic intensity signal and the temperature distribution signal into corresponding analog electrical signals, and to perform sampling and quantization processing on the analog electrical signals to generate an environmental state dataset. The calculation module is used to perform fusion calculations on the policy data and the environmental state dataset using a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching command. The determination module is used to determine, based on the load data and the task quantity data, a set of available CPU cores with a load level lower than a preset standard from the CPU cores of each server; The allocation module is used to allocate the data packet to the target CPU core in the set of available CPU cores for processing and transmission based on the frequency modulation scheme and the channel switching instruction, so as to establish multiple concurrent communication connection channels, which are used to achieve load balancing among target servers.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the CPU load balancing-based inter-server communication scheduling method as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the inter-server communication scheduling method based on CPU load balancing as described in any one of claims 1 to 7.

Images