Low-delay data compression method and system for 5G communication network

Abstract

The invention discloses a low-delay data compression method and system for a 5G communication network, and the method comprises the steps: obtaining the characteristics of an input data flow, and determining an initial resource distribution proportion and an output load distribution parameter which are needed by dynamic load balancing through analyzing the size and arrival frequency of a data packet; according to the load distribution parameters, an edge node scheduling algorithm is adopted to adjust and calculate the resource distribution proportion, and if the data flow burst amount exceeds a preset flow threshold value, the edge node processing capacity is increased, and an optimized resource scheduling scheme is obtained; obtaining a current data stream processing request from an edge node through an optimized resource scheduling scheme, generating a compression task queue in combination with an efficient coding technology, and outputting a to-be-processed data block sequence; and for the to-be-processed data block sequence, executing high-compression-rate coding by adopting a rapid compression algorithm, and if the compression time exceeds 5G short-time slot limitation, switching to a low-calculation-complexity mode. According to the invention, the system resource utilization rate and the data transmission efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular discloses a low-latency data compression method and system for a 5G communication network. Background Art

[0002] The research on low-latency data compression for 5G communication networks is one of the core fields in the development of modern communication technologies, and its importance is self-evident. With the wide deployment of 5G networks in scenarios such as ultra-reliable low-latency communication, high-bandwidth applications, and large-scale Internet of Things, data transmission efficiency and real-time performance have become key indicators for measuring network performance. An efficient data compression method can not only significantly reduce bandwidth occupancy, but also directly affect the end-to-end latency, and thus determine the success or failure of 5G technology in high-demand scenarios such as industrial automation and intelligent transportation. Currently, although certain progress has been made in the application of data compression technology in 5G networks, many bottlenecks still need to be urgently broken through.

[0003] Existing solutions mostly rely on traditional centralized compression or single coding algorithms, which are effective in specific scenarios, but obvious limitations are exposed in the face of the diverse requirements of 5G. For example, centralized processing easily leads to an extended transmission path and increased latency; general compression algorithms are difficult to balance high compression ratio and low computational overhead, and it is difficult to adapt to the differentiated requirements brought by 5G network slicing. In addition, the disconnection between the protocol layer and the compression mechanism limits the room for latency optimization in the handshake stage. These defects are particularly prominent in ultra-low latency scenarios, directly restricting the full play of 5G potential.

[0004] The core challenge in the research field is how to achieve a balance between compression efficiency and low latency under the conditions of limited edge computing capabilities, strong data flow heterogeneity, and demanding real-time requirements. Specifically, the dynamic allocation of computing resources at edge nodes is insufficient, resulting in limited response capabilities for handling bursty data; the computational complexity of compression coding algorithms has poor adaptability to the short time slot structure of 5G, making it difficult to achieve efficient parallel processing; the lack of coordination between the transmission protocol and the compression mechanism further amplifies the latency loss. These technical factors are intertwined, forming a unique problem, that is, how to synchronously optimize the compression and transmission processes in resource-constrained edge environments.

[0005] Therefore, how to design a low-latency data compression method that combines dynamic scheduling of edge computing, efficient parallel processing of coding, and deep integration of the protocol layer to achieve dual optimization of compression efficiency and real-time performance in 5G networks has become the key issue of this research. Summary of the Invention

[0006] The present invention provides a low-latency data compression method and system for a 5G communication network, aiming to solve at least one of the defects existing in the above-mentioned prior art.

[0007] One aspect of the present invention relates to a low-latency data compression method for a 5G communication network, comprising the following steps:

[0008] Obtain the characteristics of the input data stream, determine the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and output the load distribution parameters;

[0009] According to the load distribution parameters, adopt an edge node scheduling algorithm to adjust the calculation resource allocation ratio. If the data stream burst volume exceeds the preset traffic threshold, increase the processing capacity of the edge node to obtain an optimized resource scheduling scheme;

[0010] Through the optimized resource scheduling scheme, obtain the current data stream processing request from the edge node, combine with an efficient coding technique to generate a compression task queue, and output a sequence of data blocks to be processed;

[0011] For the sequence of data blocks to be processed, perform high-compression ratio encoding using a fast compression algorithm. If the compression time exceeds the 5G short time slot limit, switch to a low computational complexity mode to obtain compressed data units;

[0012] Obtain the compressed data units, adjust the protocol handshake process in combination with the transmission protocol optimization rules, and output a transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets.

[0013] Further, the step of obtaining the characteristics of the input data stream, determining the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and outputting the load distribution parameters includes:

[0014] Capture the input data stream, extract the packet size and arrival frequency to obtain the data stream characteristics;

[0015] Use statistical methods to analyze the packet size and arrival frequency to determine the change trend in the data stream characteristics;

[0016] According to the change trend, calculate the resource allocation ratio and output the initial resource allocation result;

[0017] If the initial resource allocation result exceeds the preset quantity threshold, adjust the resource allocation ratio through a linear regression algorithm to obtain an optimized allocation scheme;

[0018] Generate load distribution parameters through the optimized allocation scheme and judge the load balancing state;

[0019] Obtain the load balancing state, combine with the data stream characteristics to determine the real-time adjustment requirement for dynamic load balancing;

[0020] For the real-time adjustment requirement, process the input data stream using a sliding window method and output the final load distribution parameters.

[0021] Further, in the step of processing the input data stream by using a sliding window method and outputting the final load distribution parameters for real-time adjustment requirements, the final load distribution parameters are as follows:

[0022]

[0023] Among them, γ(t) represents the load distribution parameter at time t, m represents the number of monitoring indicators, and Δ i (t) represents the change amount of the i-th indicator, and rate i (t) represents the change rate of the i-th indicator.

[0024] Further, according to the load distribution parameters, using an edge node scheduling algorithm to adjust the calculation resource allocation ratio, and if the data stream burst volume exceeds a preset traffic threshold, then increasing the processing capacity of the edge node, the steps for obtaining an optimized resource scheduling scheme include:

[0025] Using the load distribution parameters, adjusting the calculation resource allocation ratio by using an edge node scheduling algorithm to obtain a preliminary resource scheduling result.

[0026] If the data stream burst volume exceeds the second preset traffic threshold, then adjust the resource allocation ratio by using a linear regression algorithm, increase the calculation resources, and generate an adjusted resource scheduling scheme;

[0027] For the adjusted resource scheduling scheme, obtain the edge node status, judge whether the calculation resources meet the load distribution requirements, and obtain the resource allocation status;

[0028] Using the resource allocation status, analyze the change of the data stream burst volume by using a sliding window method to determine the real-time adjustment requirements of the optimization scheme;

[0029] According to the real-time adjustment requirements, adjust the parameters of the edge node scheduling algorithm to obtain the updated allocation ratio of the edge nodes;

[0030] For the updated allocation ratio of the edge nodes, obtain the change of the load distribution parameters, judge the dynamic adjustment range of the edge node processing capacity, and generate the final resource scheduling scheme.

[0031] Further, in the step of adjusting the parameters of the edge node scheduling algorithm according to the real-time adjustment requirements to obtain the updated allocation ratio of the edge nodes, the updated allocation ratio of the edge nodes is as follows:

[0032]

[0033] Among them, α represents the updated resource allocation ratio of the edge node, n represents the total number of edge nodes, and w i represents the weight coefficient of the i-th node, and d irepresents the demand of the i-th node, c i represents the computing power of the i-th node.

[0034] Furthermore, the steps of obtaining the current data stream processing request from the edge node through the optimized resource scheduling scheme, generating a compressed task queue by combining an efficient coding technique, and outputting a sequence of data blocks to be processed include:

[0035] Obtain a data stream processing request from the edge node through the optimized resource scheduling scheme, generate a compressed task queue by using an efficient coding technique, and determine a sequence of data blocks to be processed;

[0036] For the sequence of data blocks to be processed, extract data blocks from the task queue, use a preset traffic threshold to determine whether the data stream needs to be segmented, and obtain segmented data units;

[0037] According to the segmented data units, obtain the node status from the edge node, determine whether the resource scheduling meets the current processing request, and determine the need for adjusting the allocation ratio;

[0038] Based on the need for adjusting the allocation ratio, use a dynamic adjustment method to update the resource scheduling of the edge node to obtain an adjusted task allocation scheme;

[0039] For the adjusted task allocation scheme, obtain the real-time changes of the data stream from the edge node, use a sliding window method to analyze the execution status of the compression tasks, and determine the processing order of the task queue;

[0040] According to the processing order of the task queue, perform secondary compression on the data blocks by using an efficient coding technique to determine the final data processing result;

[0041] Based on the final data processing result, obtain feedback information from the edge node, determine the dynamic adjustment range of the resource scheduling, and generate an updated allocation ratio of the data blocks to be processed.

[0042] Furthermore, for the sequence of data blocks to be processed, perform high-compression-rate coding using a fast compression algorithm. If the compression time exceeds the 5G short time slot limit, switch to a low computational complexity mode. The steps of obtaining the compressed data units include:

[0043] For the sequence of data blocks, perform coding using a fast compression algorithm to obtain a preliminary compression result;

[0044] If the compression time exceeds the short time slot limit, adjust to the low complexity mode through mode switching to determine the compressed data units;

[0045] According to the compressed data units, extract the task priorities from the processing sequence and determine the execution order of the compression tasks;

[0046] For the execution order of the compression task, a time judgment method is adopted to analyze the resource allocation under short time slot constraints, and an adjusted task queue is obtained;

[0047] Through the adjusted task queue, real-time status data is obtained from the edge nodes to determine the processing progress of the data unit;

[0048] According to the processing progress of the data unit, a sliding window method is adopted to analyze the execution status of the compression task and judge the subsequent coding requirements;

[0049] For the subsequent coding requirements, a fast compression algorithm is adopted to optimize and adjust the data unit to determine the final processing result.

[0050] Furthermore, the steps of obtaining the compressed data unit, adjusting the protocol handshake process in combination with the transmission protocol optimization rules, and outputting the transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets include:

[0051] Obtain the compressed data, determine the initial configuration of the transmission protocol through data unit division, and output the protocol adjustment parameters;

[0052] Through the protocol adjustment parameters, update the handshake process using the optimization rules to obtain a streamlined flow sequence;

[0053] For the streamlined flow sequence, judge the triggering conditions of redundant acknowledgments. If the conditions are met, reduce the number of acknowledgment packets and output the adjusted packet sequence;

[0054] According to the adjusted packet sequence, obtain the generation status of the transmission frame from the data processing module and determine the frame output order;

[0055] Adopt a sliding window method to analyze the frame output order and judge the real-time load of the transmission frame to obtain the load balancing configuration;

[0056] Through the load balancing configuration, adjust the processing priority of the data unit and output the optimized transmission data frame;

[0057] For the optimized transmission data frame, adopt a fast compression algorithm to perform secondary adjustment on the data unit to determine the final transmission unit.

[0058] Another aspect of the present invention relates to a low-latency data compression system for a 5G communication network, which is applied to the above-mentioned low-latency data compression method for a 5G communication network and includes:

[0059] The first acquisition module is used to acquire the characteristics of the input data stream, determine the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and output the load distribution parameters;

[0060] A processing module, configured to adjust the calculation resource allocation ratio according to the load distribution parameter by using an edge node scheduling algorithm. When the data stream burst volume exceeds a preset traffic threshold, the processing capacity of the edge node is increased to obtain an optimized resource scheduling scheme;

[0061] A first output module, configured to obtain the current data stream processing request from the edge node through the optimized resource scheduling scheme, generate a compression task queue by combining an efficient coding technique, and output a sequence of data blocks to be processed;

[0062] A second acquisition module, configured to perform high-compression-rate coding on the sequence of data blocks to be processed by using a fast compression algorithm. When the compression time exceeds the 5G short time slot limit, it switches to a low computational complexity mode to obtain compressed data units;

[0063] A second output module, configured to obtain the compressed data units, adjust the protocol handshake process by combining transmission protocol optimization rules, and output a transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets.

[0064] Further, the first acquisition module includes:

[0065] An extraction unit, configured to capture the input data stream and extract the data packet size and arrival frequency to obtain the data stream characteristics;

[0066] A first determination unit, configured to analyze the data packet size and arrival frequency by using a statistical method to determine the change trend in the data stream characteristics;

[0067] A calculation unit, configured to calculate the resource allocation ratio according to the change trend and output an initial resource allocation result;

[0068] An acquisition unit, configured to, when the initial resource allocation result exceeds a preset quantity threshold, adjust the resource allocation ratio by using a linear regression algorithm to obtain an optimized allocation scheme;

[0069] A judgment unit, configured to generate a load distribution parameter through the optimized allocation scheme and judge the load balancing state;

[0070] A second determination unit, configured to obtain the load balancing state and determine the real-time adjustment requirement for dynamic load balancing in combination with the data stream characteristics;

[0071] An output unit, configured to process the input data stream by using a sliding window method for the real-time adjustment requirement and output the final load distribution parameter.

[0072] The beneficial effects achieved by the present invention are:

[0073] The present invention provides a low-latency data compression method and system for a 5G communication network, which obtains the characteristics of the input data stream, determines the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and outputs the load distribution parameters; according to the load distribution parameters, adopts an edge node scheduling algorithm to adjust the calculation resource allocation ratio, and if the data stream burst volume exceeds the preset traffic threshold, increases the processing capacity of the edge node to obtain an optimized resource scheduling scheme; through the optimized resource scheduling scheme, obtains the current data stream processing request from the edge node, combines with an efficient coding technology to generate a compression task queue, and outputs a sequence of data blocks to be processed; for the sequence of data blocks to be processed, adopts a fast compression algorithm to perform high-compression-rate encoding, and if the compression time exceeds the 5G short time slot limit, switches to a low computational complexity mode to obtain compressed data units; obtains the compressed data units, combines with the transmission protocol optimization rules to adjust the protocol handshake process, and outputs a transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets. The low-latency data compression method and system for a 5G communication network provided by the present invention, by analyzing the characteristics of the input data stream, determines the initial resource allocation ratio and outputs the load distribution parameters, according to the load distribution parameters, dynamically adjusts the calculation resource allocation by adopting an edge node scheduling algorithm, and increases the processing capacity of the edge node when the data stream bursts; obtains the data stream processing request from the edge node, generates a compression task queue and outputs a sequence of data blocks to be processed, for these data blocks, adopts a fast compression algorithm to perform high-compression-rate encoding, and switches to a low-complexity mode when the compression time exceeds the limit; combines with the transmission protocol optimization rules, adjusts the handshake process and reduces the redundant acknowledgment packets, and outputs an optimized transmission data frame, realizing efficient dynamic load balancing and data compression transmission in an edge computing environment, and improving the system resource utilization rate and data transmission efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 FIG. is a schematic flowchart of an embodiment of a low-latency data compression method for a 5G communication network according to the present invention. DETAILED DESCRIPTION

[0075] In order to better understand the above technical solutions, the following will describe the above technical solutions in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0076] As Figure 1 shown, the first embodiment of the present invention proposes a low-latency data compression method for a 5G communication network, including the following steps:

[0077] Step S100, obtain the characteristics of the input data stream, determine the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and output the load distribution parameters.

[0078] Data Flow is the process of data transmission, storage, and processing along a specific path in a computer system or information system, characterized by real-time, continuity, and dynamics.

[0079] The packet size refers to the amount of data in a packet, usually measured in bytes (Byte). The packet size determines the amount of information that can be carried in a single transmission.

[0080] The arrival frequency refers to the rate at which packets arrive at the receiving end, usually measured by the number of packets arriving per second (packets / second) or the data rate (bits / second).

[0081] Dynamic load balancing is a technology that dynamically adjusts data and task allocation according to the real-time state of the system, aiming to improve the performance and stability of the system. By monitoring the load situation of the server in real time and dynamically reallocating tasks based on this information, the load balance of the system is ensured.

[0082] The load distribution parameters of the data flow are key indicators and configuration rules used to describe the characteristics of task, request, or resource allocation in the data flow processing system, aiming to optimize system performance and achieve load balance. The load distribution parameters include traffic characteristic parameters, resource usage parameters, load balancing algorithm parameters, distribution model parameters, and dynamic adjustment parameters, etc. Among them, the traffic characteristic parameters include data throughput and the ratio of real-time to historical traffic. The resource usage parameters include CPU and memory usage rates and network bandwidth occupancy rates. The load balancing algorithm parameters include weights and priorities, as well as the number of connections and failure thresholds. The distribution model parameters include the task arrival rate distribution and service time distribution. The dynamic adjustment parameters include response time sensitivity and intelligent scheduling algorithm parameters.

[0083] Step S200: According to the load distribution parameters, use the edge node scheduling algorithm to adjust the calculation resource allocation ratio. If the data flow burst volume exceeds the preset traffic threshold, increase the processing capacity of the edge node to obtain an optimized resource scheduling plan.

[0084] The edge node scheduling algorithm plays a core role in resource optimization and task allocation in edge computing. By pre-simulating the combination of computing tasks and edge nodes, generating hash values, and establishing a solution library, the optimal strategy is directly called during operation.

[0085] The data flow burst volume refers to the phenomenon of a sudden increase in data traffic within a short period, commonly seen in Internet of Things, 5G communication, and real-time service scenarios. Its processing requires the combination of network architecture optimization and edge computing technology.

[0086] The edge node adopts a heterogeneous computing architecture (CPU + GPU / FPGA), supporting trillion-level parallel computing per second in industrial scenarios and meeting the needs of complex data processing.

[0087] The resource scheduling scheme can adopt the edge computing resource scheduling scheme, which is a dynamic resource scheduling strategy used to generate a hash value by simulating the combination of computing tasks and edge nodes in advance, establish a scheduling scheme library with the minimum delay, and directly call the optimal strategy during operation to reduce the real-time computing overhead.

[0088] Step S300: Through the optimized resource scheduling scheme, obtain the current data stream processing request from the edge node, combine it with the efficient coding technology to generate a compressed task queue, and output the sequence of data blocks to be processed.

[0089] The efficient coding technology compresses or transforms the original data into a more compact form through specific conversion rules or algorithms. The efficient coding technology can significantly reduce the size of the data, thereby saving storage space and transmission bandwidth. The efficient coding technology can adopt algorithms such as Huffman coding, LZW (string table compression algorithm), or arithmetic coding.

[0090] The compressed task queue is a technology for processing data compression asynchronous tasks, mainly used to process those tasks that take a long time, such as sending emails, generating reports, etc. By putting the tasks into the queue, the system can process these tasks asynchronously without blocking the main program.

[0091] The data block sequence refers to a sequence formed by arranging data blocks in a certain order in the fields of computer science and information technology. A data block is a group of records arranged continuously in order, and it is the data unit for transmission between the main memory and input devices, output devices, or external memories.

[0092] Step S400: For the sequence of data blocks to be processed, perform high-compression ratio coding using a fast compression algorithm. If the compression time exceeds the 5G short time slot limit, switch to the low computational complexity mode to obtain the compressed data unit.

[0093] The fast compression algorithm is an algorithm that can efficiently compress data in a short time. The fast compression algorithm has a wide range of applications in multiple fields, such as data transmission, storage, and processing.

[0094] The 5G short time slot limit refers to the technical constraint condition for minimizing the data transmission delay by flexibly adjusting the number of symbols and the subcarrier spacing (Numerology) in the time slot structure. In the 5G network, a time slot is the basic unit of time resource, usually composed of multiple orthogonal frequency division multiplexing (OFDM) symbols. 5G defines a short time slot structure called mini-slot, which is mainly used for ultra-high reliable and low latency application scenarios.

[0095] A compressed data unit refers to a structured data block formed by encoding and compressing the original data through a specific algorithm. Its core objective is to improve storage or transmission efficiency by reducing redundant information.

[0096] The Low Computational Complexity Mode refers to a technique or strategy in algorithm, system, or hardware design that significantly reduces the time, space, or energy consumption required for a computing task by optimizing the computing steps, resource usage, or processing flow. Its core objective is to efficiently complete the target while reducing the demand for computing resources, and it is applicable to resource-constrained scenarios (such as mobile devices, embedded systems, real-time processing, etc.).

[0097] Step S500: Obtain the compressed data unit, adjust the protocol handshake process in combination with the transmission protocol optimization rules, and output a transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets.

[0098] The transmission protocol optimization rules are systematic strategies designed for the characteristics of 5G networks, aiming to improve data transmission efficiency by reducing latency, enhancing reliability, optimizing resource utilization, etc. The transmission protocol optimization rules can adopt adaptive data compression to dynamically select compression algorithms according to the data type (text / video / sensor data).

[0099] The protocol handshake process is a process of a class of network protocols that allows the client and the server to confirm each other's identities.

[0100] The transmission protocol optimization rules adopt a protocol enhancement mechanism, such as the integration of the QUIC (Quick UDP Internet Connections) protocol. In a weak network environment, it switches to the QUIC protocol and replaces the TCP (Transmission Control Protocol) three-way handshake with a 1-RTT handshake (compatible with TLS1.3), reducing the latency by 50%.

[0101] The Redundant Acknowledgment Packet is a mechanism in the TCP (Transmission Control Protocol) used for quickly detecting and recovering data loss.

[0102] The transmission data frame is the basic unit for data exchange in a computer network. During the transmission process, the transmitted data is divided into individual data frames and then transmitted through the network.

[0103] Furthermore, for the low-latency data compression method for a 5G communication network provided in this embodiment, step S100 includes:

[0104] Step S110: By capturing the input data stream, extract the packet size and arrival frequency to obtain the data stream characteristics.

[0105] By capturing the input data stream and extracting the packet size and arrival frequency, it can be understood as obtaining basic information from network transmission. For example, in a video stream transmission scenario, assume the average packet size is 500KB and the arrival frequency is 10 packets per second. Such characteristics reflect the stability and intensity of the data stream.

[0106] Step S120: Use statistical methods to analyze the packet size and arrival frequency to determine the trend of change in the data stream characteristics.

[0107] When using statistical methods for analysis, it can be observed whether the packet size fluctuates over a period of time. For example, if the size changes from 500KB to 800KB and the arrival frequency drops from 10 per second to 8 per second during a certain period, this indicates that the traffic may have changed due to user behavior or network conditions.

[0108] Specifically, the average value and variance can be used to quantify these trends of change. For example, calculate the average packet size and frequency within 5 minutes to determine whether the traffic tends to be stable or surges. Such analysis helps to anticipate resource requirements in advance and avoid system overload caused by sudden traffic changes.

[0109] Step S130: According to the trend of change, calculate the resource allocation ratio and output the initial resource allocation result.

[0110] When calculating the resource allocation ratio according to the trend of change, in one possible implementation, the total throughput can be estimated based on the product of the packet size and frequency. Exemplarily, if the initial throughput is 500KB×10 = 5000KB / second and the system presets that each unit of resource processes 1000KB / second, then initially allocate 5 resource units.

[0111] Step S140: If the initial resource allocation result exceeds the preset quantity threshold, then adjust the resource allocation ratio through the linear regression algorithm to obtain an optimized allocation plan.

[0112] If the preset quantity threshold is 4 units, the excess part triggers the adjustment. Preferably, use the linear regression algorithm to predict the traffic trend in the next 10 minutes. For example, according to historical data, it is fitted that the packet size may increase to 600KB, the frequency rises to 12 per second, and the total throughput becomes 7200KB / second. Therefore, adjust to 7 resource units. Such an optimized plan can more accurately match the demand, avoid resource waste or shortage, and improve system efficiency.

[0113] Step S150: Through the optimized allocation plan, generate load distribution parameters and judge the load balancing state.

[0114] When generating the load distribution parameters through the optimized allocation scheme, it can be understood as mapping resource allocation to the server cluster.

[0115] For example, if 7 units are allocated to 3 servers, which are 3, 2, and 2 respectively, the load distribution parameters may be the utilization rate of each server. Ideally, it is close to balance, such as 70%, 60%, and 65%.

[0116] When judging the load balancing state, if the utilization rate of a certain server exceeds 90%, it indicates imbalance.

[0117] Step S160: Obtain the load balancing state, and combine with the data flow characteristics to determine the real-time adjustment requirements for dynamic load balancing.

[0118] Combined with the data flow characteristics, such as a sudden increase in frequency, real-time adjustment may be required. For real-time adjustment requirements, the sliding window method for processing data streams is an efficient means.

[0119] Specifically, with a 1-minute window, monitor the average throughput of the last 5 windows. If it rises from 5000 KB / second to 6000 KB / second, dynamically add 1 resource unit to the server with a high utilization rate. This way can quickly respond to traffic changes and ensure service stability.

[0120] Step S170: For real-time adjustment requirements, adopt the sliding window method to process the input data stream and output the final load distribution parameters.

[0121] It should be noted that the selection of the size and step length of the sliding window affects the adjustment accuracy. For example, too small a window may lead to frequent adjustments and increase system overhead; too large a window may result in slow response. In one embodiment, combining a 1-minute window during stable traffic and a 30-second window during mutations can balance efficiency and sensitivity.

[0122] The final load distribution parameters, such as the utilization rate of each server stabilizing at about 75% after adjustment, indicate good load balancing effect. This dynamic adjustment not only improves resource utilization but also reduces response latency and enhances the user experience. It can be understood that this method is applicable to traffic management in a single scenario, has strong scalability and low implementation cost, and is particularly suitable for services with high real-time requirements.

[0123] In this embodiment, the final load distribution parameters are:

[0124]

[0125] In formula (1), γ(t) represents the load distribution parameter at time t, m represents the number of monitoring indicators, Δ i (t) represents the change amount of the i-th indicator, rate i (t) represents the change rate of the i-th indicator.

[0126] Preferably, for the low-latency data compression method for a 5G communication network provided in this embodiment, step S200 includes:

[0127] Step S210: Adjust the calculation resource allocation ratio by using an edge node scheduling algorithm through load distribution parameters to obtain a preliminary resource scheduling result.

[0128] When adjusting the calculation resource allocation ratio through load distribution parameters, it can be understood as scheduling based on the load status of edge nodes. For example, in a video stream transmission scenario, assume that the load distribution parameter of edge node A shows that the current utilization rate is 80%, while that of node B is 50%.

[0129] By using an edge node scheduling algorithm, some calculation tasks can be transferred from A to B. The preliminary scheduling result may be that A drops to 65% and B rises to 60%. This method can quickly match resources with demands by analyzing load distribution parameters.

[0130] Step S220: If the data stream burst volume exceeds the second preset traffic threshold, then adjust the resource allocation ratio through a linear regression algorithm, increase the calculation resources, and generate an adjusted resource scheduling plan.

[0131] In a possible implementation manner, if the data stream burst volume exceeds the second preset threshold, for example, the throughput surges from 5000 KB / second to 7000 KB / second, the processing capabilities of edge nodes need to be re-evaluated. Exemplarily, the processing upper limit of node A is 6000 KB / second. At this time, resources need to be increased, such as adding an auxiliary node C with a processing capability of 2000 KB / second. The adjusted scheduling plan may be that A maintains 5000 KB / second and C shares 2000 KB / second. This dynamic allocation can quickly respond to burst traffic and ensure that the service is not interrupted.

[0132] Step S230: For the adjusted resource scheduling plan, obtain the edge node status, judge whether the calculation resources meet the load distribution requirements, and obtain the resource allocation status.

[0133] For the adjusted resource scheduling plan, it is particularly crucial to obtain the edge node status. Specifically, the real-time utilization rates of nodes A, B, and C can be monitored to judge whether the load requirements are met. For example, if the utilization rate of A is stable at 70%, that of B is 60%, and that of C is 50%, it indicates that the resource allocation status is good. If the utilization rate of C is only 20%, it means that there is resource redundancy and further optimization is needed.

[0134] Step S240: Through the resource allocation status, use a sliding window method to analyze the change in the data stream burst volume and determine the real-time adjustment requirements of the optimization plan.

[0135] Preferably, the sliding window method can be used to analyze the change in the burst volume of the data stream. In one embodiment, with a 30 - second window, observing the throughput of the past 5 windows, it is found that it gradually rises from 6000 KB / second to 7500 KB / second, indicating the need for real - time adjustment. When adjusting the parameters of the edge node scheduling algorithm, the resource allocation ratio can be changed from A:B:C = 5:3:2 to 4:3:3 to ensure that C undertakes more tasks. This dynamic parameter adjustment can better adapt to traffic fluctuations.

[0136] Step S250: According to the real - time adjustment requirement, adjust the parameters of the edge node scheduling algorithm to obtain the updated allocation ratio of the edge nodes.

[0137] It should be noted that the updated allocation ratio needs to be verified in combination with the change in the load distribution parameters. For example, after adjustment, the utilization rate of A drops to 60%, and that of C rises to 65%, indicating that the dynamic adjustment range of the node processing capacity is reasonable. If the utilization rate of C soars to 90%, then resources need to be added again. The final resource scheduling plan may be A:B:C = 4:3:4, and the total processing capacity is increased to 8000 KB / second to cope with future traffic growth.

[0138] It can be understood that this solution forms a complete closed - loop from core scheduling to dynamic adjustment. For example, the initial scheduling depends on load parameters, additional resources are allocated during bursts, status monitoring ensures balance, the sliding window optimizes real - time performance, and the final plan verifies the adjustment range. This multi - faceted supported logic not only ensures the efficient use of resources but also improves the system stability. Especially during peak traffic periods, it can significantly reduce latency and guarantee the user experience.

[0139] Step S260: For the updated allocation ratio of the edge nodes, obtain the change in the load distribution parameters, judge the dynamic adjustment range of the edge node processing capacity, and generate the final resource scheduling plan.

[0140] In one possible implementation, if the burst traffic subsides, for example, the throughput drops back to 5500 KB / second, the trend can be confirmed through the sliding window, and the resource allocation for C can be gradually reduced to 1 unit. This flexibility avoids resource waste while maintaining the efficiency of scheduling.

[0141] Specifically, the dynamic adjustment range of the edge node processing capacity can also predict the traffic peak through historical data. For example, by analyzing the burst pattern in the past week, 20% of redundant resources can be reserved in advance. This forward - looking scheduling can further enhance the robustness of the solution.

[0142] Furthermore, in step S250, the updated allocation ratio of the edge nodes is:

[0143]

[0144] In formula (2), α represents the resource allocation ratio after the update of the edge node, n represents the total number of edge nodes, w i represents the weight coefficient of the i-th node, d i represents the demand of the i-th node, c i represents the computing power of the i-th node.

[0145] Furthermore, the low-latency data compression method for a 5G communication network provided in this embodiment, step S300 includes:

[0146] Step S310, obtain a data stream processing request from an edge node through an optimized resource scheduling scheme, generate a compression task queue using an efficient coding technique, and determine a sequence of data blocks to be processed.

[0147] After obtaining a data stream processing request from an edge node through an optimized resource scheduling scheme, generate a compression task queue using an efficient coding technique. For example, when an edge node receives a data stream of 10 GB per second, the data volume can be compressed to about 6 GB through a lossless compression technique to form a task queue for subsequent processing.

[0148] Specifically, this efficient coding can select a dictionary-based compression method to reduce redundant information according to the repetitive characteristics of the data stream. It should be noted that the generation of the compression task queue needs to consider the real-time nature of the data stream to ensure that the tasks in the queue do not accumulate due to coding delays.

[0149] Step S320, for the sequence of data blocks to be processed, extract a data block from the task queue, and use a preset traffic threshold to determine whether the data stream needs to be segmented to obtain segmented data units.

[0150] For the sequence of data blocks to be processed, after extracting a data block, use a preset traffic threshold to determine whether segmentation is required.

[0151] In a possible implementation, if the size of a single data block exceeds 2 MB, it is segmented into multiple 512 KB data units.

[0152] Exemplarily, this segmentation method can be adjusted according to the type of data block. For example, video stream segmentation pays more attention to frame boundaries, while log data is segmented according to timestamps.

[0153] It can be understood that the segmented data units can more flexibly adapt to the processing capabilities of edge nodes.

[0154] Step S330, according to the segmented data units, obtain the node status from the edge node, determine whether the resource scheduling meets the current processing request, and determine the need for adjusting the allocation ratio.

[0155] When obtaining the node status from the edge node according to the sharded data units, it is possible to determine whether the resource scheduling meets the requirements by checking the CPU occupancy rate and memory usage.

[0156] Preferably, if the CPU occupancy rate of a certain node exceeds 80%, it indicates that the current resources are insufficient and the allocation ratio needs to be adjusted.

[0157] For example, the processing task ratio of this node is increased from 30% to 40% to cope with the surge in data flow. This dynamic judgment method can quickly respond to load changes.

[0158] Step S340: Adjust the demand through the allocation ratio, and use the dynamic adjustment method to update the resource scheduling of the edge node to obtain the adjusted task allocation plan.

[0159] When adjusting the demand through the allocation ratio and updating the resource scheduling using the dynamic adjustment method, a priority mechanism can be introduced. In one embodiment, high-priority tasks, such as real-time monitoring data, are allocated to edge nodes with stronger performance, while low-priority tasks are allocated to standby nodes.

[0160] For example, if the processing capacity of an edge node is 5GB of data per second, 4GB is preferentially allocated to high-priority tasks, and the remaining 1GB is used to process other tasks. This method can effectively improve the processing efficiency of critical tasks.

[0161] Step S340: For the adjusted task allocation plan, obtain the real-time changes in the data flow from the edge node, and use the sliding window method to analyze the execution status of the compression task to determine the processing order of the task queue.

[0162] When analyzing the execution status of the compression task for the adjusted task allocation plan, the sliding window method can be set to a 10-second window to observe the task completion rate. For example, if the task completion rate within the window is lower than 90%, the task queue order is adjusted to give priority to processing backlogged tasks. Specifically, this analysis can timely detect bottlenecks and avoid delays caused by an overly long task queue.

[0163] Step S350: According to the processing order of the task queue, use efficient coding technology to perform secondary compression on the data blocks to determine the final data processing result.

[0164] When performing secondary compression on the data blocks using efficient coding technology, an adaptive compression algorithm can be selected to adjust the compression ratio according to the data characteristics.

[0165] For example, for text data, the compression rate can reach 70%, while for image data, it is approximately 30%. In one embodiment, after secondary compression, 1GB of data can be reduced to 400MB. This method significantly reduces the transmission and storage pressure.

[0166] Step S360: Obtain feedback information from the edge nodes based on the final data processing result, determine the dynamic adjustment range of resource scheduling, and generate the updated allocation ratio of the data blocks to be processed.

[0167] When obtaining feedback information through the final data processing result, the response time and throughput of the edge nodes can be concerned about.

[0168] For example, if the response time of a certain node increases from 50 ms to 100 ms, it indicates that the resource scheduling needs to be further adjusted, and the allocation ratio may need to be increased by 10%.

[0169] It should be noted that this feedback mechanism can provide a basis for the next scheduling and form a closed-loop optimization.

[0170] Exemplarily, when determining the dynamic adjustment range of resource scheduling, it can be predicted in combination with the historical data stream peak value. For example, the peak value in the past hour was 15 GB per second, and the current is 12 GB per second, then 20% of additional resources are reserved, that is, the adjustment range is the processing capacity of 2.4 GB per second. This predictive adjustment can effectively cope with sudden traffic and ensure the stable operation of the system.

[0171] Preferably, for the low-latency data compression method for 5G communication networks provided in this embodiment, step S400 includes:

[0172] Step S410: Perform encoding on the data block sequence using a fast compression algorithm to obtain a preliminary compression result.

[0173] Perform encoding on the data block sequence using a fast compression algorithm to obtain a preliminary compression result.

[0174] Exemplarily, the fast compression algorithm can be based on the principle of dictionary encoding. By constructing a repetition pattern table of data blocks, the frequently occurring sequences are replaced with short identifiers.

[0175] For example, when processing the log data stream at the edge node, assuming that a certain data block contains a repeated "ERROR" field, the algorithm can map it to a 2-byte identifier to generate a preliminary compression result.

[0176] Step S420: If the compression time exceeds the short time slot limit, adjust to the low-complexity mode through mode switching to determine the compressed data unit.

[0177] If the compression time exceeds the short time slot limit, adjust to the low-complexity mode through mode switching. In a possible implementation, the short time slot limit is set to 50 milliseconds. If the compression takes 60 milliseconds, switch to the low-complexity mode, such as directly skipping the complex entropy encoding and only retaining the basic replacement rules, to determine the compressed data unit. It can be understood that this can ensure that tasks do not pile up in time-sensitive scenarios.

[0178] Step S430: Extract the task priorities from the processing sequence based on the compressed data units, and determine the execution order of the compression tasks.

[0179] Specifically, the task priorities can be defined according to the size or urgency of the data units. For example, the priority of 10KB real-time monitoring data is higher than that of 5KB backup data, and it enters the execution queue first.

[0180] Step S440: For the execution order of the compression tasks, use a time judgment method to analyze the resource allocation under short time slot constraints, and obtain an adjusted task queue.

[0181] In one embodiment, if the short time slot is 50 milliseconds and the current node processing capacity is 20 tasks per second, then through time judgment, the lagging tasks in the queue are reallocated to idle nodes to optimize the queue turnover.

[0182] Step S450: Through the adjusted task queue, obtain the real-time status data from the edge nodes to determine the processing progress of the data units.

[0183] For example, if node A reports that 80% of the data units have been processed and node B has only processed 50%, then subsequent allocations can be adjusted accordingly. This real-time status acquisition enables the system to dynamically respond to load changes.

[0184] Step S460: According to the processing progress of the data units, use a sliding window method to analyze the execution status of the compression tasks and judge the subsequent coding requirements.

[0185] Specifically, the sliding window can be set to 10 tasks wide, and analyze the average time consumption of the last 10 tasks. If it is found that the time consumption increases by 20%, it indicates that coding optimization is required.

[0186] Step S470: For the subsequent coding requirements, use a fast compression algorithm to optimize and adjust the data units to determine the final processing result.

[0187] In one embodiment, if the window analysis shows that the compression ratio of some data units is lower than 30%, then adjust the algorithm parameters and add a preprocessing step, such as splitting large data blocks into multiple small units before compression.

[0188] For example, a 20KB data unit is split into 4 5KB units, encoded separately and then merged, ultimately improving the compression efficiency. It should be noted that this optimization can significantly reduce the storage pressure on the edge nodes and at the same time speed up data transmission.

[0189] For example, in the scenario of processing video stream segments at the edge nodes, fast compression can encode the repeated frame features as indexes, and the low complexity mode skips the detail optimization to ensure that tasks are completed within a short time slot.

[0190] The priority judgment gives priority to the processing of key frames, while the sliding window analysis dynamically adjusts the coding depth. The final result takes into account both real-time performance and resource utilization. This multi-faceted collaborative approach can form a complete solution within a single business domain, ensuring core requirements while enriching applicability through expansion and adjustment.

[0191] Furthermore, the low-latency data compression method for a 5G communication network provided in this embodiment, step S500 includes:

[0192] Step S510, obtain compressed data, determine the initial configuration of the transmission protocol through data unit division, and output protocol adjustment parameters.

[0193] Regarding the acquisition of compressed data and the determination of the initial configuration of the transmission protocol through data unit division, it can be understood that this process aims to lay a foundation for subsequent transmission.

[0194] Exemplarily, when processing real-time data at an edge node, assuming the data stream is compressed data of 10MB per second, through data unit division, the data can be divided into blocks of 1MB in size, and each block corresponds to a transmission unit. This division method facilitates the protocol to initialize the configuration according to the network bandwidth.

[0195] For example, in a scenario with a bandwidth of 50Mbps, the transmission time for each block is approximately 0.16 seconds, and the protocol can set the initial timeout parameter to 0.2 seconds accordingly to ensure smooth transmission. This configuration can effectively reduce the transmission delay caused by overly large data blocks.

[0196] Step S520, through the protocol adjustment parameters, update the handshake process using an optimization rule to obtain a streamlined flow program sequence.

[0197] The implementation of streamlining the flow program sequence by optimizing the handshake process through protocol adjustment parameters is worthy of attention.

[0198] Specifically, the handshake process usually involves multiple confirmations, and the optimization rule can reduce unnecessary interactions. In one possible implementation, if the initial handshake requires 3 round trips and takes 60ms, after adjusting the parameters, it can be streamlined to 2 round trips, and the time consumption is reduced to 40ms. The streamlined flow program sequence is more compact and suitable for high-frequency short-time slot transmission scenarios. This optimization can significantly improve the protocol response speed.

[0199] Step S530, for the streamlined flow program sequence, judge the triggering condition of redundant confirmation. If the condition is met, reduce the number of confirmation packets and output the adjusted packet sequence.

[0200] Regarding the judgment of the triggering condition of redundant confirmation, preferably, it can be set that if the confirmation time interval of 3 consecutive data units is less than 5ms, it is considered that the network is stable and the number of confirmation packets is reduced.

[0201] For example, the confirmation interval is adjusted from per unit to per three units, and the packet sequence is reduced from 10 to 4. This adjustment can reduce the network load, especially in high-throughput scenarios.

[0202] Step S540: According to the adjusted packet sequence, obtain the generation status of the transmission frame from the data processing module and determine the frame output order.

[0203] When obtaining the transmission frame generation status according to the adjusted packet sequence and determining the frame output order, exemplarily, if the processing module feedbacks that the current frame generation rate is 50 frames per second, the output can be arranged in timestamp order to avoid out-of-order.

[0204] Step S550: Analyze the frame output order using the sliding window method, judge the real-time load of the transmission frame, and obtain the load balancing configuration.

[0205] Analyzing the real-time load using the sliding window method further optimizes this process. In one embodiment, the window size is set to 5 frames. If the load peak reaches 80% at a certain moment, the window can be dynamically reduced to 3 frames to balance the load. This configuration can ensure transmission stability and resource utilization.

[0206] Step S560: Through the load balancing configuration, adjust the processing priority of the data unit and output the optimized transmission data frame.

[0207] When adjusting the processing priority of the data unit and outputting the optimized transmission data frame, it can be understood that the priority adjustment is based on the real-time state. For example, if the processing progress of a certain data unit lags behind by 20%, its priority can be raised to the front of the queue to ensure the priority transmission of key data.

[0208] Step S570: For the optimized transmission data frame, perform a secondary adjustment on the data unit using the fast compression algorithm to determine the final transmission unit.

[0209] The secondary adjustment of the data unit by the fast compression algorithm is more refined. In one embodiment, if the compression ratio of a certain data unit only reaches 60%, the algorithm optimization parameters can be used to increase it to 80%, and the final transmission unit is more efficient. This method can enhance the overall consistency of data transmission. For example, in the edge computing scenario, the above methods can work together. The initial configuration ensures reasonable data chunking, streamlining handshakes improves efficiency, reducing redundant confirmations reduces overhead, load balancing optimizes resource allocation, priority adjustment safeguards key tasks, and finally, the secondary compression improves transmission quality. This multi-faceted support scheme not only ensures real-time performance but also enhances the system robustness and is suitable for high-dynamic network environments.

[0210] The present invention also provides a low-latency data compression system for a 5G communication network, which is applied to the above-mentioned low-latency data compression method for a 5G communication network. The system includes a first acquisition module, a processing module, a first output module, a second acquisition module, and a second output module. Among them, the first acquisition module is used to acquire the characteristics of the input data stream, determine the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and output the load distribution parameters; the processing module is used to adjust the calculation resource allocation ratio according to the load distribution parameters by using the edge node scheduling algorithm. If the data stream burst volume exceeds the preset traffic threshold, the processing ability of the edge node is increased to obtain an optimized resource scheduling scheme; the first output module is used to obtain the current data stream processing request from the edge node through the optimized resource scheduling scheme, generate a compression task queue in combination with the efficient coding technology, and output the sequence of data blocks to be processed; the second acquisition module is used to perform high-compression ratio encoding on the sequence of data blocks to be processed by using the fast compression algorithm. If the compression time exceeds the 5G short time slot limit, it switches to the low computational complexity mode to obtain the compressed data unit; the second output module is used to obtain the compressed data unit, adjust the protocol handshake process in combination with the transmission protocol optimization rules, and output the transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets.

[0211] Further, for the low-latency data compression system for a 5G communication network provided in this embodiment, the first acquisition module includes an extraction unit, a first determination unit, a calculation unit, an acquisition unit, a judgment unit, a second determination unit, and an output unit. Among them, the extraction unit is used to extract the packet size and arrival frequency by capturing the input data stream to obtain the data stream characteristics; the first determination unit is used to analyze the packet size and arrival frequency by using a statistical method to determine the change trend in the data stream characteristics; the calculation unit is used to calculate the resource allocation ratio according to the change trend and output the initial resource allocation result; the acquisition unit is used to adjust the resource allocation ratio by using the linear regression algorithm if the initial resource allocation result exceeds the preset quantity threshold to obtain an optimized allocation scheme; the judgment unit is used to generate the load distribution parameters through the optimized allocation scheme and judge the load balancing state; the second determination unit is used to obtain the load balancing state and determine the real-time adjustment requirement of dynamic load balancing in combination with the data stream characteristics; the output unit is used to process the input data stream by using the sliding window method for the real-time adjustment requirement and output the final load distribution parameters.

[0212] This embodiment provides a low-latency data compression method and system for a 5G communication network. Compared with the existing technology, it obtains the characteristics of the input data stream, determines the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and outputs the load distribution parameters; according to the load distribution parameters, it uses the edge node scheduling algorithm to adjust the calculation resource allocation ratio. If the data stream burst volume exceeds the preset traffic threshold, it increases the processing capacity of the edge node to obtain an optimized resource scheduling scheme; through the optimized resource scheduling scheme, it obtains the current data stream processing request from the edge node, combines with the efficient coding technology to generate a compression task queue, and outputs the sequence of data blocks to be processed; for the sequence of data blocks to be processed, it uses the fast compression algorithm to perform high-compression-rate encoding. If the compression time exceeds the 5G short time slot limit, it switches to the low computational complexity mode to obtain the compressed data unit; it obtains the compressed data unit, combines with the transmission protocol optimization rules to adjust the protocol handshake process, and outputs the transmission data frame with optimized protocol by reducing the number of redundant acknowledgment packets. The low-latency data compression method and system for a 5G communication network provided by this embodiment determine the initial resource allocation ratio and output the load distribution parameters by analyzing the characteristics of the input data stream. According to the load distribution parameters, it dynamically adjusts the calculation resource allocation using the edge node scheduling algorithm and increases the processing capacity of the edge node when the data stream bursts; it obtains the data stream processing request from the edge node, generates a compression task queue and outputs the sequence of data blocks to be processed. For these data blocks, it uses the fast compression algorithm to perform high-compression-rate encoding and switches to the low complexity mode when the compression time exceeds the limit; it combines with the transmission protocol optimization rules to adjust the handshake process and reduce the redundant acknowledgment packets, and outputs the optimized transmission data frame, realizing efficient dynamic load balancing and data compression transmission in the edge computing environment, and improving the system resource utilization rate and data transmission efficiency.

[0213] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A low-latency data compression method for 5G communication networks, characterized in that: The following steps are involved: Obtain the input data flow characteristics, determine the initial resource allocation ratio required for dynamic load balancing by analyzing the packet size and arrival frequency, and output the load distribution parameters; According to the load distribution parameters, the edge node scheduling algorithm is used to adjust the computing resource allocation ratio. If the data flow burst exceeds the preset flow threshold, the edge node processing capacity is increased to obtain an optimized resource scheduling solution; Through the optimized resource scheduling scheme, the current data stream processing request is obtained from the edge node, and the compression task queue is generated by combining the efficient coding technology to output the sequence of data blocks to be processed; For the sequence of data blocks to be processed, a fast compression algorithm is used to perform high compression rate encoding. If the compression time exceeds the 5G short time slot limit, it is switched to a low computational complexity mode to obtain a compressed data unit; The compressed data unit is obtained, and the protocol handshake process is adjusted in combination with the transmission protocol optimization rule, and the transmission data frame optimized by the protocol is output by reducing the number of redundant confirmation packets.

2. The low-latency data compression method for 5G communication network according to claim 1, characterized in that: The step of obtaining input data flow characteristics, determining the initial resource allocation ratio required for dynamic load balancing by analyzing the data packet size and arrival frequency, and outputting load distribution parameters comprises: By capturing the input data stream, extracting the packet size and arrival frequency, the data stream characteristics are obtained; Analyzing the size and arrival frequency of data packets using statistical methods to determine the changing trends in the characteristics of the data stream; According to the change trend, the resource allocation ratio is calculated and the initial resource allocation result is output; If the burst volume of the data flow exceeds the first preset flow threshold, the resource allocation ratio is adjusted by a linear regression algorithm to obtain an optimized allocation plan; Generate load distribution parameters through the optimized allocation scheme and determine the load balancing state; Obtain the load balancing state, and determine the real-time adjustment requirements of dynamic load balancing in combination with data flow characteristics; In response to the real-time adjustment requirements, a sliding window method is used to process the input data stream and output the final load distribution parameters.

3. The low-latency data compression method for 5G communication network according to claim 2, characterized in that: In the step of processing the input data stream using a sliding window method and outputting the final load distribution parameter according to the real-time adjustment requirement, the final load distribution parameter is: Among them, γ(t) represents the load distribution parameter at time t, m represents the number of monitoring indicators, and Δ i (t) represents the change of the i-th indicator, rate i (t) represents the rate of change of the i-th indicator.

4. The low-latency data compression method for 5G communication network according to claim 1, characterized in that: According to the load distribution parameters, the edge node scheduling algorithm is used to adjust the computing resource allocation ratio. If the data flow burst exceeds the preset flow threshold, the edge node processing capacity is increased. The steps of obtaining the optimized resource scheduling scheme include: Through the load distribution parameters, the edge node scheduling algorithm is used to adjust the computing resource allocation ratio and obtain the preliminary resource scheduling results. If the burst volume of the data flow exceeds the second preset flow threshold, the resource allocation ratio is adjusted through a linear regression algorithm, computing resources are increased, and an adjusted resource scheduling plan is generated; According to the adjusted resource scheduling plan, the edge node status is obtained to determine whether the computing resources meet the load distribution requirements and obtain the resource allocation status; Through the resource allocation status, the sliding window method is used to analyze the changes in the burst volume of data flow and determine the real-time adjustment requirements of the optimization plan; According to the real-time adjustment requirements, the edge node scheduling algorithm parameters are adjusted to obtain the updated edge node allocation ratio; According to the updated allocation ratio of edge nodes, the load distribution parameter changes are obtained, the dynamic adjustment range of the edge node processing capacity is determined, and the final resource scheduling plan is generated.

5. The low-latency data compression method for 5G communication network according to claim 4, characterized in that: In the step of adjusting the edge node scheduling algorithm parameters according to the real-time adjustment requirements to obtain the updated edge node allocation ratio, the updated edge node allocation ratio is: Among them, α represents the updated resource allocation ratio of edge nodes, n represents the total number of edge nodes, and w i represents the weight coefficient of the i-th node, d i represents the demand of the ith node, c i Represents the computing power of the i-th node.

6. The low-latency data compression method for 5G communication network according to claim 1, characterized in that: The steps of obtaining the current data stream processing request from the edge node through the optimized resource scheduling scheme, generating a compression task queue in combination with the efficient coding technology, and outputting a sequence of data blocks to be processed include: Through the optimized resource scheduling scheme, data stream processing requests are obtained from edge nodes, and the compression task queue is generated using efficient coding technology to determine the sequence of data blocks to be processed; For the sequence of data blocks to be processed, extract the data blocks from the task queue, use the preset flow threshold to determine whether the data stream needs to be processed in fragments, and obtain the fragmented data units; According to the sharded data units, the node status is obtained from the edge node to determine whether the resource scheduling meets the current processing request and determine the allocation ratio adjustment requirements; Adjust the demand through the allocation ratio, adopt a dynamic adjustment method to update the resource scheduling of the edge node, and obtain an adjusted task allocation plan; According to the adjusted task allocation scheme, the real-time changes of data streams are obtained from edge nodes, and the execution status of compression tasks is analyzed using the sliding window method to determine the processing order of the task queue; According to the processing order of the task queue, the data block is compressed twice using the efficient coding technology to determine the final data processing result; Through the final data processing results, feedback information is obtained from the edge nodes to determine the dynamic adjustment range of resource scheduling and generate the updated allocation ratio of the data blocks to be processed.

7. The low-latency data compression method for 5G communication network according to claim 1, characterized in that: The step of using a fast compression algorithm to perform high compression rate encoding for a sequence of data blocks to be processed, and switching to a low computational complexity mode if the compression time exceeds the 5G short time slot limit, to obtain a compressed data unit includes: For the data block sequence, a fast compression algorithm is used to perform encoding to obtain a preliminary compression result; If the compression time exceeds the short time slot limit, the mode is adjusted to a low complexity mode by mode switching to determine the compressed data unit; According to the compressed data unit, the task priority is extracted from the processing sequence to determine the execution order of the compression tasks; According to the execution order of the compressed tasks, the time judgment method is used to analyze the resource allocation under the short time slot constraint, and the adjusted task queue is obtained; Obtain real-time status data from edge nodes through the adjusted task queue to determine the processing progress of data units; According to the processing progress of the data unit, the sliding window method is used to analyze the execution status of the compression task and determine the subsequent encoding requirements; In response to subsequent encoding requirements, a fast compression algorithm is used to optimize and adjust the data units to determine the final processing results.

8. The low-latency data compression method for 5G communication network according to claim 1, characterized in that: The steps of obtaining the compressed data unit, adjusting the protocol handshake process in combination with the transmission protocol optimization rule, and outputting the protocol optimized transmission data frame by reducing the number of redundant confirmation packets include: Obtain compressed data, determine the initial configuration of the transmission protocol through data unit division, and output protocol adjustment parameters; Adjust parameters through the protocol, adopt optimization rules to update the handshake process, and obtain a streamlined process sequence; For the streamlined process sequence, determine the triggering conditions for redundant confirmation. If the conditions are met, reduce the number of confirmation packages and output the adjusted package sequence. According to the adjusted packet sequence, the generation state of the transmission frame is obtained from the data processing module to determine the frame output sequence; The frame output sequence is analyzed by using a sliding window method to determine the real-time load of the transmission frame and obtain a load balancing configuration; By means of the load balancing configuration, the processing priority of the data unit is adjusted, and an optimized transmission data frame is output; For the optimized transmission data frame, a fast compression algorithm is used to perform a second adjustment on the data unit to determine the final transmission unit.

9. A low-latency data compression system for 5G communication networks, applied to the low-latency data compression method for 5G communication networks as claimed in any one of claims 1 to 8, characterized in that: include: The first acquisition module is used to acquire the input data flow characteristics, determine the initial resource allocation ratio required for dynamic load balancing by analyzing the data packet size and arrival frequency, and output the load distribution parameters; A processing module, configured to adjust the computing resource allocation ratio by using an edge node scheduling algorithm according to the load distribution parameters, and if the data flow burst exceeds a preset flow threshold, increase the edge node processing capacity to obtain an optimized resource scheduling solution; The first output module is used to obtain the current data stream processing request from the edge node through the optimized resource scheduling scheme, generate a compression task queue in combination with the efficient coding technology, and output a sequence of data blocks to be processed; A second acquisition module is used to perform high compression rate encoding using a fast compression algorithm for the sequence of data blocks to be processed, and if the compression time exceeds the 5G short time slot limit, switch to a low computational complexity mode to obtain a compressed data unit; The second output module is used to obtain the compressed data unit, adjust the protocol handshake process in combination with the transmission protocol optimization rule, and output the protocol optimized transmission data frame by reducing the number of redundant confirmation packets.

10. The low-latency data compression system for 5G communication network according to claim 9, characterized in that: The first acquisition module includes: An extraction unit, used for capturing an input data stream, extracting a data packet size and an arrival frequency, and obtaining data stream characteristics; A first determining unit, configured to analyze the size and arrival frequency of data packets using a statistical method to determine a change trend in the data flow characteristics; A calculation unit, used to calculate the resource allocation ratio according to the change trend and output an initial resource allocation result; An acquisition unit, configured to adjust the resource allocation ratio by a linear regression algorithm to obtain an optimized allocation scheme if the burst volume of the data stream exceeds a first preset flow threshold; A judging unit, configured to generate load distribution parameters and judge a load balancing state through the optimized allocation scheme; A second determination unit, configured to obtain the load balancing state and determine the real-time adjustment requirements of the dynamic load balancing in combination with data flow characteristics; The output unit is used to process the input data stream using a sliding window method according to the real-time adjustment demand and output the final load distribution parameters.

Images