A data communication system and method based on future network

Abstract

The present invention discloses a data communication system and method based on future networks, which specifically relates to the field of intelligent traffic management under future network environments, and is used to solve the problems of dynamic task splitting and equipment resource optimization. First, the load regulation index and task load intensity index of the equipment are monitored in real time to accurately obtain the equipment status and task requirements. Based on the multi-objective optimization model of genetic algorithm, the equipment load and task computing requirements are dynamically balanced, and the optimal task splitting scheme is generated to reasonably split and distribute the tasks to the equipment with lower load. Combined with the network topology and path priority model, the task is split into multiple subtasks, and the optimal path allocation is used to improve the task processing efficiency and reduce network congestion. During the task execution process, the equipment load changes and task progress are continuously monitored, the task granularity is dynamically adjusted according to the resource status and task demand changes, and the task splitting or merging is automatically triggered to achieve efficient resource utilization and avoid equipment overload or resource waste.

Description

Technical Field

[0001] The present invention relates to the field of intelligent traffic management in a future network environment, and more specifically, to a data communication system and method based on a future network. Background Art

[0002] With the development of 6G network technology, intelligent traffic management systems have gradually become an important infrastructure for urban traffic scheduling and optimization. The high bandwidth, low latency and powerful computing power of the 6G network enable real-time transmission and processing of traffic data, supporting intelligent vehicle scheduling, traffic flow prediction and rapid response to emergencies. The system provides efficient path planning, traffic signal optimization and other services by analyzing and processing large-scale real-time traffic data to ensure smooth and safe traffic operation. However, in the face of complex traffic scenarios, different types of tasks need to be reasonably split and assigned to various network element devices for processing. The effective splitting of tasks has become a key factor affecting system efficiency.

[0003] In the prior art, task splitting is usually based on pre-set fixed rules, which lack dynamic integration with task requirements and network element device load status. In particular, the lack of flexibility in adjusting task granularity has become a core issue restricting system performance. When dealing with highly complex tasks, the traditional fixed splitting mechanism cannot make appropriate granularity adjustments to tasks based on actual conditions. Specifically, when the task computing requirements change or the network element device load is uneven, the task splitting scheme still uses static rules, resulting in overload of some network element devices, while other network element device resources are idle, which seriously affects the execution efficiency of the task and the overall response speed of the system.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a data communication system and method based on the future network, which significantly improves the flexibility of task processing and resource utilization efficiency through multi-level task splitting and execution optimization. First, by real-time monitoring of the load regulation index and task load intensity index of the device, the current state of the device and the task requirements are accurately obtained, providing accurate data support. The multi-objective optimization model based on the genetic algorithm can dynamically balance the device load and the task computing requirements, generate the optimal task splitting scheme, ensure that the task is reasonably split and matched to the device with lower load, and maximize resource utilization. Combined with the network topology and path priority model, the task is split into multiple subtasks in real time and allocated to the appropriate device through the optimal path, which significantly improves the overall efficiency of task processing and reduces the risk of network congestion. During the task execution process, the device load changes and task progress are continuously monitored, the task granularity is dynamically adjusted according to the changes in resource status and task requirements, and the further splitting or merging operations of the task are automatically triggered to achieve efficient utilization of resources and avoid device overload or resource waste. The overall solution ensures the stability and efficiency of task processing through real-time monitoring, intelligent scheduling, and dynamic task adjustment mechanism, while ensuring the optimal allocation and use of network resources to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] Monitor the usage of each network element in real time, analyze the load of network element and the complexity of tasks, and reflect the status of tasks and network element;

[0008] A multi-objective optimization model based on genetic algorithms balances task computing requirements and network resources to obtain the optimal task splitting solution;

[0009] According to the optimal splitting scheme, the task is split into multiple subtasks, and the subtasks are preferentially assigned to low-load network element devices based on the network topology and delay conditions;

[0010] Split or merge existing subtasks based on task requirements and network element equipment resource conditions.

[0011] In a preferred embodiment, the use of each network element device is monitored in real time, the network element device load and task complexity are analyzed, and the status of the task and the network element device is reflected, including the following contents:

[0012] Deploy a distributed sensor network and embed a dedicated status data acquisition module in each network element device to collect the parameters of the network element device in real time. The edge node transmits the parameters of the network element device to the central dispatching center, which calculates the load regulation index of each network element device based on the occupancy of different resources. According to the load intensity brought by the task during the processing, the pressure and computing intensity on the network element device during task execution are calculated to obtain the task load intensity index.

[0013] In a preferred embodiment, the load regulation index acquisition logic is:

[0014] First, a nonlinear logarithmic function is used in combination with an exponential decay function to handle the dynamic changes in data transmission traffic; the formula is as follows: ; : Input data traffic of network element equipment; : Output data traffic of network element equipment; : The amount of data to be processed that is not currently processed by the network element device; : Idle time of network element equipment refers to the period of time when the network element equipment is not processing data; : The reaction coefficient of the network element device when processing idle;

[0015] Task feedback delay is based on the ratio of the actual task response time to the ideal response time, and is used to measure the time efficiency of network element devices in processing tasks. The formula is as follows: ; : Feedback time of actual tasks; : The time to complete the task under ideal conditions; : The amount of data currently being processed by the task; : The amount of unprocessed data accumulated by network element devices; : A positive number to prevent the denominator from being zero.

[0016] Taking into account the data transmission intensity and task feedback delay, the load regulation index is calculated using nonlinear adjustment items. The formula is as follows: ; : The difference between input and output data traffic, indicating the asymmetry of data traffic of network element devices; : Reflects the total scale of data processed and accumulated by current network element devices.

[0017] In a preferred embodiment, the logic for obtaining the task load intensity index is:

[0018] Calculate the dependency function of the task and recursively calculate the task load intensity. The recursive function form is as follows: ; : The first Sub-data processing rate; : Data dependency coefficient of the task; : A very small positive number to prevent the denominator from being zero;

[0019] Based on the relationship between the data processing rate and the recursive function, the task load intensity index is obtained by processing the cumulative value of the recursive function. The formula is as follows: ; : Indicates the accumulation of recursive processing rate values ​​at each stage of task processing; : The maximum processing rate value that occurs during task processing, indicating the processing peak value of the task at a certain moment; : Indicates the number of recursive accumulation stages during task execution.

[0020] In a preferred embodiment, a multi-objective optimization model based on a genetic algorithm is used to balance task computing requirements and network resources to obtain an optimal task splitting solution, including the following:

[0021] In the genetic algorithm, an initial population containing multiple task splitting schemes is first generated. Each individual represents a possible task splitting scheme, including the number of subtasks into which the task is split and the size of each subtask. Assuming that the task needs to be split into M subtasks, each individual is represented by a vector ,in Indicates The size of each subtask, the dimension of each individual is associated with the available resource status of the network element device;

[0022] The generation of the initial population combines the load regulation index and task load intensity index of all current network element devices, and takes the load regulation index of the network element device as a constraint to ensure that the subtasks after task splitting will not exceed the resource carrying capacity of the network element device; the task load intensity index of the task is taken as the target, so that the split subtasks can balance the task processing efficiency and the network element device load;

[0023] The quality of each individual is determined by its fitness function. The fitness function combines the load regulation index and the task load intensity index to optimize through two core goals: one is to balance the computing requirements of the task and the resource utilization of the network element equipment, and the other is to maximize the overall task processing efficiency of the system. The fitness function is designed as a comprehensive function of the two main goals: ;in: :Network element equipment Load regulation index; : The average load regulation index of all network element devices in the current system; : After the task is split into subtasks, the sum of the task load intensity indexes of each subtask; : A very small positive number to prevent division by zero errors;

[0024] Selection, crossover and mutation operations are performed based on the fitness function to continuously generate new task splitting schemes and gradually approach the optimal solution; after multiple generations of iterations, the algorithm finally converges and obtains the optimal task splitting scheme.

[0025] In a preferred implementation, according to the optimal splitting scheme, the task is split into multiple subtasks, and the subtasks are preferentially allocated to low-load network element devices in combination with the network topology and delay conditions, including the following:

[0026] In order to ensure that the subtasks of a task can be preferentially assigned to the most appropriate network element device, the topology of the entire network is first obtained; the network topology is represented in the form of a weighted directed graph, where each node represents a network element device, and the weight of each edge reflects the delay or bandwidth limitation between network element devices;

[0027] In order to allocate task subtasks to the most appropriate network element devices, a path priority model is constructed, which comprehensively considers the link delay, bandwidth availability and load regulation index of the network element devices in the topology; the priority of each path Calculated by the following formula: ;in: : represents the delay of the corresponding path; : The average value of the load regulation index of all network element devices passing through the corresponding path; : Minimum link bandwidth on the corresponding path; : A positive number used to prevent division by zero errors.

[0028] In a preferred embodiment, based on the optimal task splitting scheme, the task is split into multiple subtasks, and these subtasks are allocated to different network element devices according to the path priority model. The specific task allocation process is implemented by the following steps:

[0029] According to the size of each subtask in the optimal splitting solution, the subtask with small load is preferentially allocated to the network element device with low load;

[0030] According to the path priority between each network element device and the source node, the path with the highest priority is selected, and the task subtasks are assigned to the network element devices on the corresponding path;

[0031] After the subtasks are assigned to the network element devices, the execution status of the subtasks of each network element device is monitored, and the execution progress of the subtasks is adjusted in real time according to the load regulation index. When the load index of a certain network element device rises rapidly, the reallocation mechanism is triggered to transfer some unfinished subtasks to other network element devices with lower loads.

[0032] In a preferred implementation, existing subtasks are split or merged according to task requirements and network element device resource conditions, including the following:

[0033] According to the changes in the network element load index and task load index, the task splitting or merging operation is automatically triggered. When the following two situations are detected, different operation strategies will be adopted:

[0034] When the task load intensity index is higher than the task load threshold and the load regulation index is lower than the load threshold, the task splitting operation will be triggered;

[0035] When the task load intensity index is lower than the task load threshold and the load regulation index is higher than the load threshold, the task splitting operation will be triggered.

[0036] A data communication system based on a future network, comprising: a monitoring module, an optimization module, an allocation module and an adjustment module;

[0037] Monitoring module: monitors the usage of each network element device in real time, analyzes the network element device load and task complexity, reflects the status of tasks and network element devices, and transmits the analyzed network element device load and task complexity to the optimization module;

[0038] Optimization module: A multi-objective optimization model based on genetic algorithms balances task computing requirements and network resources, obtains the optimal task splitting solution, and passes the optimal splitting solution to the allocation module;

[0039] Allocation module: splits the task into multiple subtasks according to the optimal splitting scheme, and preferentially allocates the subtasks to low-load network element devices based on the network topology and delay conditions. The matching results between the task and the network element device are sent to the adjustment module.

[0040] Adjustment module: Split or merge existing subtasks according to task requirements and network element equipment resources.

[0041] The technical effects and advantages of a data communication system and method based on a future network of the present invention are as follows:

[0042] The present invention significantly improves the flexibility of task processing and resource utilization efficiency through multi-level task splitting and execution optimization. First, by real-time monitoring of the load control index and task load intensity index of the device, the current state of the device and the task requirements are accurately obtained, providing accurate data support. The multi-objective optimization model based on the genetic algorithm can dynamically balance the device load and the task calculation requirements, generate the optimal task splitting scheme, ensure that the task is reasonably split and matched to the device with lower load, and maximize resource utilization. Combined with the network topology and path priority model, the task is split into multiple subtasks in real time and allocated to the appropriate device through the optimal path, which significantly improves the overall efficiency of task processing and reduces the risk of network congestion. During the task execution process, the device load change and task progress are continuously monitored, the task granularity is dynamically adjusted according to the changes in resource status and task requirements, and the further splitting or merging operation of the task is automatically triggered to achieve efficient utilization of resources and avoid device overload or resource waste. The overall solution ensures the stability and efficiency of task processing through real-time monitoring, intelligent scheduling, and dynamic task adjustment mechanism, while ensuring the optimal allocation and use of network resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The present invention is a flow chart of a data communication method based on a future network.

[0044] Figure 2 The present invention is a structural schematic diagram of a data communication system based on a future network. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Embodiment 1: Figure 1 The present invention provides a data communication method based on a future network, comprising:

[0047] Monitor the usage of each network element device in real time, analyze the load of network element devices and the complexity of tasks, and reflect the status of tasks and network element devices.

[0048] A multi-objective optimization model based on genetic algorithm is used to balance task computing requirements and network resources to obtain the optimal task splitting solution.

[0049] According to the optimal splitting plan, the task is split into multiple subtasks, and the subtasks are preferentially allocated to low-load network element devices based on the network topology and delay conditions.

[0050] Existing subtasks can be further split or merged based on task requirements and network element equipment resource conditions.

[0051] In the 6G intelligent traffic management system, real-time monitoring of each network element and dynamic management of task status are the core elements to ensure the efficient operation of the system. In complex traffic scenarios, the system needs to dynamically obtain the resource usage of each network element, and combine the computing requirements of the current task and the network data traffic to calculate the "network element load index" and "task complexity index" that reflect the network element and task status. These indicators provide key decision-making basis for subsequent task splitting and scheduling strategies.

[0052] Monitor the usage of each network element in real time, analyze the load of network element and the complexity of tasks, and reflect the status of tasks and network element, including the following:

[0053] Deploy a distributed sensor network and embed a dedicated status data collection module in each network element device to collect resource parameters such as CPU utilization, memory occupancy, network bandwidth utilization, and I / O throughput of network element devices in real time. Use lightweight edge computing nodes to pre-process the collected data locally to reduce network latency for transmission. Each edge node generates a preliminary network element device status report based on the resource occupancy of the network element device, and transmits the data to the central dispatching system after data compression.

[0054] The edge node transmits the parameters of the network element equipment to the central dispatching center, and the central dispatching center calculates the load control index of each network element equipment according to the occupancy of different resources.

[0055] According to the load intensity brought by the task during the processing, the pressure and computing intensity of the network element equipment during task execution are calculated to obtain the task load intensity index.

[0056] The logic for obtaining the load regulation index is:

[0057] First, the data transmission intensity (DTI) should not only reflect the transmission pressure of the network element equipment, but also consider the accumulated amount of data in its processing queue. Therefore, a nonlinear logarithmic function is used in combination with an exponential decay function to handle the dynamic changes of data transmission traffic.

[0058] The formula is as follows: ; : Input data traffic of network element equipment (the amount of data received per unit time). : Output data traffic of network element equipment (the amount of data transmitted per unit time). : The amount of data to be processed that is not currently processed by the network element device, reflecting the backlog of data pressure on the network element device. : The idle time of the network element device refers to the time period when the network element device does not process data, which is used to reflect the idleness of the network element device. : The reaction coefficient of the network element device when processing idle state, which indicates the speed at which the network element device responds to data in idle state. The value can be adjusted according to the network element device type and actual scenario.

[0059] A logarithmic function is introduced into the data transmission intensity to handle the nonlinear relationship between data flow and backlog data. At the same time, the idle time of the network element device is introduced through the exponential decay term to ensure that when the network element device is idle for a long time, the data transmission intensity value decays faster, reflecting the state of reduced pressure on the network element device.

[0060] Task feedback delay (TFL) is the ratio of the actual task response time to the ideal response time, which is used to measure the time efficiency of network element equipment in processing tasks. In order to further increase the complexity of the formula, the task feedback delay is combined with the dynamics of data accumulation in task processing, and nonlinear terms are added.

[0061] The formula is as follows: ; : The feedback time of the actual task, reflecting the time consumed by the network element equipment to complete the task. : The time to complete the task under ideal conditions, and the expected feedback time generated by the preset model. : The amount of data currently being processed by the task. : The amount of unprocessed data accumulated in network element devices. : A positive number to prevent the denominator from being zero.

[0062] This formula combines the actual task feedback time with the task backlog status of the network element device, and nonlinearly reflects the lag effect of task feedback. The more backlogged data, the greater the processing feedback delay of the network element device, which increases the complexity of load regulation.

[0063] On this basis, the load regulation index is calculated by comprehensively considering the data transmission intensity and task feedback delay and using nonlinear adjustment items. This ensures that when the task pressure of the network element equipment increases, the load regulation index will increase rapidly, and when it is idle, the load regulation index will decrease smoothly.

[0064] The formula is as follows: ; : The difference between input and output data traffic indicates the asymmetry of data traffic of network element devices. The larger the difference, the more likely there is a bottleneck in data processing efficiency. : Reflects the total scale of data processed and accumulated by the current network element equipment, and uses a logarithmic function to ensure that the load index has a smooth growth characteristic when the data scale increases.

[0065] The load regulation index is used to indicate the comprehensive load status of the network element device during data transmission and task processing. This index can reflect the load pressure and data flow processing efficiency of the network element device when processing the current task. The larger the load regulation index, the greater the transmission burden, the larger the task backlog or the obvious task feedback delay when the network element device processes data, indicating that the network element device resources are close to saturation or in a high-load state, and the task processing efficiency is low; on the contrary, the smaller the index, the higher the fluency of the network element device in processing data, the more timely the task feedback, the less the backlog of data, the network element device is in a relatively idle or normal load state, and has a greater resource carrying capacity. The change of this index directly reflects the dynamic adjustment of the load status of the network element device during operation, which facilitates the system to accurately evaluate and manage the load status of the network element device when scheduling tasks.

[0066] The logic for obtaining the task load intensity index is:

[0067] First, the data processing rate of a task refers to the amount of data that a network element device can process per unit time, and is the core parameter of the task load intensity index. The data processing rate of a task reflects the degree of dependence on data during the execution of the task, that is, whether the data input and processing can maintain sufficient fluency.

[0068] The data dependency of tasks in the processing process is introduced in the form of recursive functions. Task processing is usually continuous. When the data processing speed of a part of the task slows down, it will cause backlog and delay of subsequent processing. To this end, the dependency function of the task is calculated, and the task load intensity is recursively calculated.

[0069] The recursive function is as follows: ; : The first The data processing rate indicates the processing efficiency of the task at the current stage. : The data dependency coefficient of the task, which means that the stronger the dependency between the previous and next stages of the task during data processing, the larger the value. : A small positive number to prevent the denominator from being zero.

[0070] The recursive model shows that when the processing rate of a task decreases, data dependency will rapidly increase, increasing the load intensity of subsequent task execution. By recursively accumulating data dependency, the dynamic changes in the load during task execution can be effectively captured, reflecting the complexity of the task.

[0071] Based on the relationship between data processing rate and recursive function, a nonlinear complexity evaluation is formed by processing the cumulative value of the recursive function to obtain the task load intensity index (TLII).

[0072] The formula is as follows: ; : Indicates the accumulation of recursive values ​​of the processing rate at each stage of the task processing process, reflecting the changes in the overall complexity of the task. : The maximum processing rate value that occurs during task processing, indicating the processing peak value of the task at a certain moment. This item is used to correct the situation where the exponential growth is too fast, ensuring that the exponential will not be amplified due to a very fast processing speed. : Indicates the number of recursively accumulated stages during the task execution process, specifically referring to the different stages or time slices into which the task is divided from the start of execution to the current moment.

[0073] The task load intensity index is used to indicate the pressure and processing complexity of the task on system resources during execution. By analyzing the data processing rate of the task and its data dependency recursive model, it reflects the consumption of system resources by the task at different stages. The larger the index, the higher the data dependency and the lower the data processing rate of the task during processing, indicating that the task occupies a higher degree of system resources, the load is heavier, and there may be bottlenecks or delays in task execution; on the contrary, the smaller the index, the faster the task processing speed, the lower the data dependency, the system can execute the task more efficiently, the load is lighter, and the task has less demand for resources. The change of this index directly reflects the resource occupation of the task at different stages, which helps the system dynamically adjust the task splitting or merging strategy to ensure the optimal use of resources.

[0074] In the intelligent traffic management scenario of future networks, the task computing requirements and efficient scheduling of network resources determine the overall performance of the system. In order to ensure that the task splitting scheme can take into account the load conditions and task processing requirements of each network element device, a multi-objective optimization model based on genetic algorithm is used, combining the two core parameters of "load regulation index" and "task load intensity index". Under the premise of balancing network element device resources and task pressure, the optimal task splitting scheme is dynamically generated, and the size of the subtask is determined according to the load conditions of each network element device.

[0075] A multi-objective optimization model based on genetic algorithms balances task computing requirements and network resources to obtain the optimal task splitting solution, including the following:

[0076] In the genetic algorithm, it is first necessary to generate an initial population containing multiple task splitting schemes. Each individual represents a possible task splitting scheme, including the number of subtasks into which the task is split and the size of each subtask. Each individual in the population is represented by binary coding or other appropriate coding methods.

[0077] Individual representation: Assuming that the task needs to be split into M subtasks, each individual can be represented as a vector ,in Indicates The size of each subtask. Each individual dimension is associated with the available resource status of the network element device.

[0078] Generation rules: The generation of the initial population is based on the load regulation index and task load intensity index of all current network element devices in the system. The load regulation index of the network element device is used as a constraint to ensure that the subtasks after the task is split will not exceed the resource carrying capacity of the network element device; the task load intensity index of the task is used as a target to ensure that the split subtasks can balance the task processing efficiency and the network element device load.

[0079] Technical details: When the initial population is generated, multiple task splitting schemes are randomly generated, but it is necessary to ensure that these schemes can meet the current load conditions of the network element devices. For example, if the load regulation index of a network element device is high, the generated task splitting scheme should avoid too many subtasks being assigned to the network element device to prevent the network element device from being overloaded.

[0080] The quality of each individual (task splitting scheme) is determined by its fitness function. The fitness function combines the load regulation index and the task load intensity index and is optimized through two core goals: one is to balance the computing requirements of the task and the resource utilization of the network element equipment, and the other is to maximize the overall task processing efficiency of the system.

[0081] The fitness function is designed as a comprehensive function of two main objectives: ;in: :Network element equipment The load regulation index reflects the current resource usage status of the network element equipment. : The average load control index of all network element devices in the current system, indicating the overall load balance of network element devices. : After the task is split into subtasks, the sum of the task load intensity indexes of each subtask reflects the overall processing difficulty of the task. : A very small positive number to prevent division by zero errors.

[0082] The first part: By comparing the load regulation index of the network element device with the average load of the system, the load balance degree of the calculation task on different network element devices is calculated. If the load of a network element device is close to the average level of the system, its fitness is high, reflecting that its load state is relatively reasonable.

[0083] The second part: Use the task load intensity index as the optimization target to ensure that the computing requirements of the task are met. The lower the task load intensity index of the subtask, the higher the execution efficiency of the subtask after the task is split.

[0084] The core steps of the genetic algorithm are selection, crossover and mutation operations based on the fitness function to continuously generate new task splitting schemes and gradually approach the optimal solution.

[0085] Selection: Roulette selection or tournament selection is used to give priority to individuals with higher fitness for the next generation of reproduction. This ensures that solutions with good load balance and lower task load intensity index have a higher probability of being selected.

[0086] Crossover: Perform a crossover operation on the selected individuals to generate a new task splitting scheme. The crossover operation simulates genetic recombination in heredity, allowing some parts of the task split to be inherited from different parent schemes. The offspring individuals after crossover represent a new task splitting strategy, which can further optimize the system's task allocation.

[0087] Mutation: To maintain population diversity and prevent falling into a local optimal solution, some individuals are randomly selected for mutation operations, the size of some subtasks is randomly adjusted, or the number of subtasks into which a task is split is changed. This step ensures that the algorithm explores a wider solution space and enhances the flexibility of the task splitting scheme.

[0088] After multiple generations of iterations, the algorithm finally converged and obtained the optimal task splitting scheme. This scheme not only determined how many subtasks the task should be split into, but also determined the size of each subtask, and the task allocation scheme has been optimized according to the load of the network element equipment.

[0089] The subtasks under the optimal task splitting scheme are assigned to the network element device with the lowest load. The results are updated based on the real-time load regulation index of the network element device, and the network topology and transmission delay are taken into consideration to ensure efficient transmission and processing of task data.

[0090] Prioritize allocating subtasks to network element devices with lower load regulation indexes, and adjust the size of subtasks so that the load intensity of each network element device when processing subtasks is at a reasonable level to avoid overload or waste of resources.

[0091] According to the optimal splitting scheme, the task is split into multiple subtasks. Combined with the network topology and delay, the subtasks are preferentially assigned to low-load network element devices, including the following:

[0092] In order to ensure that the subtasks of a task can be assigned to the most appropriate network element device first, the topology of the entire network needs to be obtained first. Each network element device regularly sends network connectivity information between itself and other network element devices to the central dispatch center through the built-in network monitoring module, including bandwidth availability, delay, number of hops, link quality, etc.

[0093] The network topology is represented in the form of a weighted directed graph, where each node represents a network element and the weight of each edge reflects the delay or bandwidth limitation between network elements. The topology is updated regularly to ensure the dynamic nature of task allocation.

[0094] Each network element device obtains the real-time status of surrounding network element devices through the link state protocol, including information such as link bandwidth, congestion level and transmission delay.

[0095] The central dispatching system summarizes this status information to form a topology map that reflects the overall status of the network, which will serve as the basis for subtask allocation.

[0096] In order to allocate task subtasks to the most appropriate network element devices, a path priority model is constructed, which comprehensively considers the link delay, bandwidth availability and load regulation index of network element devices in the topology. The goal of this model is to give priority to paths with low load, low delay and high bandwidth to ensure the execution efficiency of subtasks and data transmission speed.

[0097] The priority of each path Calculated by the following formula: ;in: : Indicates the delay of the path, which indicates the delay of task data passing through the path. : The average value of the load regulation index of all network element devices passing through the corresponding path, reflecting the load status of the network element devices on the path. : The minimum link bandwidth on the path to ensure that the path bandwidth does not become a bottleneck. : A positive number used to prevent division by zero errors.

[0098] The path priority model calculates the priority based on the latency of each path and the network element load index, and gives priority to paths with low latency, light network element load and sufficient bandwidth for task allocation.

[0099] Based on the optimal task splitting solution, the task is split into multiple subtasks, and these subtasks are assigned to different network element devices according to the path priority model. The specific task allocation process is implemented through the following steps:

[0100] According to the size of each subtask in the optimal splitting solution, the subtasks with small loads are preferentially allocated to the network element devices with low loads to balance the network load.

[0101] The scheduling system selects the path with the highest priority according to the path priority between each network element device and the source node, and allocates the task subtasks to the network element devices on the path.

[0102] Once the allocation path of the subtask is determined, the dispatch center will start the data transmission process, the data packet will be transmitted to the corresponding network element device through the preferred path, and the subtask will start to be executed on the network element device.

[0103] After the subtasks are assigned to the network element devices, the subtask execution of each network element device is monitored, including the execution progress, data transmission success rate and load status of the network element device. The scheduling system will adjust the execution progress of the subtask in real time according to the load regulation index (MLRI) and reallocate unfinished subtasks when necessary.

[0104] When the load index of a network element device rises rapidly, the dispatch center will trigger the reallocation mechanism to transfer some unfinished subtasks to other network elements with lower loads to prevent the network element device from being overloaded.

[0105] If a path is congested or has a link failure, the path priority is recalculated based on real-time feedback information and the subtasks are reallocated.

[0106] During the task execution process, since the load status of network element devices and task requirements may change, the system needs to have a flexible adjustment mechanism to dynamically split or merge subtasks to ensure the best use of resources and efficient task execution. By real-time monitoring of the load of network element devices and the progress of task execution, the granularity of tasks can be dynamically adjusted according to demand to improve resource utilization efficiency and prevent network element device overload or resource waste.

[0107] Based on task requirements and network element equipment resource conditions, the existing subtasks are further split or merged, including the following:

[0108] Each network element device periodically calculates its current load regulation index (MLRI) and task load intensity index (TLII), and feeds them back to the central dispatching system at fixed time intervals (such as every 10 seconds).

[0109] According to the changes in the network element load index and task load index, the task is automatically split or merged to optimize resource utilization. Different operation strategies will be adopted when the following two situations are detected:

[0110] When the task load intensity index (TLII) increases, it means that the computing demand of the task is growing. If the load regulation index (MLRI) of the network element device decreases, it means that the network element device still has sufficient resources to handle more tasks. At this time, the existing subtasks are further split into smaller subtasks and distributed to more network element devices to speed up processing.

[0111] First, evaluate the size and execution progress of the current subtask, and conduct further granular analysis on each subtask based on the task load intensity index, splitting the subtask with large load into multiple smaller subtasks to avoid a single subtask occupying too many resources or taking too long to execute.

[0112] When the task load intensity index is higher than the task load threshold and the load regulation index is lower than the load threshold, the task splitting operation will be triggered.

[0113] Select subtasks with low load and merge them into a larger subtask, thereby reducing the overhead of task switching and alleviating the computing burden of network element devices. The merging operation is completed by analyzing the dependencies and data processing status of the current tasks to ensure that the merging of tasks does not affect the overall execution efficiency.

[0114] When the task load intensity index is lower than the task load threshold and the load regulation index is higher than the load threshold, the task splitting operation will be triggered.

[0115] A dual trigger mechanism is introduced to automatically determine the conditions for splitting and merging tasks by real-time monitoring of task requirements and network element equipment loads, ensuring that the granularity of tasks and the resource status of network element equipment are always dynamically balanced.

[0116] When the conditions for splitting or merging are met, the task adjustment operation is immediately performed to ensure that the newly generated subtasks can be quickly allocated to appropriate network element devices, or the subtasks are merged into suitable network element devices for continued execution.

[0117] First, the task will be finely divided according to the above conditions to generate more subtasks. The subtasks will be allocated to the network elements with lower loads according to the network topology and path priority model. The size of the subtask will be adjusted according to the resource status of the network element to ensure that each network element only bears the appropriate load.

[0118] When splitting tasks, you first need to decompose the data flow of the subtask, divide the data flow into multiple small flows according to the set ratio, and redistribute these data flows through the topology structure to ensure that data transmission between network element devices is minimized.

[0119] For tasks that trigger merging, the data dependencies between subtasks are first checked to ensure that the merged tasks will not cause data conflicts or incorrect task execution order. After that, the system will merge the data streams of these subtasks into a large stream, perform unified processing after merging, and redistribute them to network element devices that meet the processing conditions.

[0120] After the tasks are merged, the merged tasks will reduce the amount of data transmission through compression technology, and uniformly handle task status synchronization to ensure seamless connection when the network element equipment takes over the tasks.

[0121] Through the fine data flow decomposition and merging technology in the process of task splitting and merging, the data transmission bottleneck in task processing is avoided, and it is ensured that there is no delay or interruption in the execution process of the task after adjustment.

[0122] The present invention significantly improves the flexibility of task processing and resource utilization efficiency through multi-level task splitting and execution optimization. First, by real-time monitoring of the load control index and task load intensity index of the network element device, the current state and task requirements of the network element device are accurately obtained, providing accurate data support. The multi-objective optimization model based on the genetic algorithm can dynamically balance the load of the network element device and the task calculation requirements, generate the optimal task splitting scheme, ensure that the task is reasonably split and matched to the network element device with lower load, and maximize resource utilization. Combined with the network topology and path priority model, the task is split into multiple subtasks in real time and allocated to the appropriate network element device through the optimal path, which significantly improves the overall efficiency of task processing and reduces the risk of network congestion. During the task execution process, the load change of the network element device and the progress of the task are continuously monitored, the task granularity is dynamically adjusted according to the changes in the resource status and task requirements, and the further splitting or merging operation of the task is automatically triggered to achieve efficient utilization of resources and avoid overload or waste of resources of the network element device. The overall solution ensures the stability and efficiency of task processing through real-time monitoring, intelligent scheduling, and dynamic task adjustment mechanism, while ensuring the optimal allocation and use of network resources.

[0123] Embodiment 2: Figure 2 The present invention provides a data communication system based on a future network, comprising: a monitoring module, an optimization module, an allocation module and an adjustment module;

[0124] Monitoring module: monitors the usage of each network element device in real time, analyzes the network element device load and task complexity, reflects the status of tasks and network element devices, and transmits the analyzed network element device load and task complexity to the optimization module.

[0125] Optimization module: A multi-objective optimization model based on genetic algorithm balances the task computing requirements and network resources, obtains the optimal task splitting solution, and passes the optimal splitting solution to the allocation module.

[0126] Allocation module: According to the optimal splitting plan, the task is split into multiple subtasks. Combined with the network topology and delay conditions, the subtasks are preferentially allocated to low-load network element devices. The results of the matching task and network element device selection are sent to the adjustment module.

[0127] Adjustment module: further split or merge existing subtasks according to task requirements and network element equipment resources.

[0128] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0129] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0130] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or network element device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or network element device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method, article or network element device including the elements.

[0131] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A data communication method based on a future network, characterized in that: Includes steps: Monitor the usage of each network element in real time, analyze the load of network element and the complexity of tasks, and reflect the status of tasks and network element, including the following: Deploy a distributed sensor network and embed a dedicated status data collection module in each network element device to collect the parameters of the network element device in real time; The edge node transmits the parameters of the network element equipment to the central dispatching center. The central dispatching center calculates the load regulation index of each network element equipment according to the occupancy of different resources. According to the load intensity brought by the task during the processing, the pressure and computing intensity of the network element equipment during the task execution are calculated to obtain the task load intensity index. The logic for obtaining the load regulation index is: First, a nonlinear logarithmic function is used in combination with an exponential decay function to handle the dynamic changes in data transmission traffic; the formula is as follows: ; : Input data traffic of network element equipment; : Output data traffic of network element equipment; : The amount of data to be processed that is not currently processed by the network element device; : Idle time of network element equipment refers to the period of time when the network element equipment is not processing data; : The reaction coefficient of the network element device when processing idle; Task feedback delay is based on the ratio of the actual task response time to the ideal response time, and is used to measure the time efficiency of network element devices in processing tasks. The formula is as follows: ; : Feedback time of actual tasks; : The time to complete the task under ideal conditions; : The amount of data currently being processed by the task; : The amount of unprocessed data accumulated by network element devices; : positive number, to prevent the denominator from being zero; Taking into account the data transmission intensity and task feedback delay, the load regulation index is calculated using nonlinear adjustment items. The formula is as follows: ; : The difference between input and output data traffic, indicating the asymmetry of data traffic of network element devices; : Reflects the total scale of data processed and accumulated by current network element devices; The logic for obtaining the task load intensity index is: Calculate the dependency function of the task and recursively calculate the task load intensity. The recursive function form is as follows: ; : The first Sub-data processing rate; : Data dependency coefficient of the task; : A very small positive number to prevent the denominator from being zero; Based on the relationship between the data processing rate and the recursive function, the task load intensity index is obtained by processing the cumulative value of the recursive function. The formula is as follows: ; : Indicates the accumulation of recursive processing rate values ​​at each stage of task processing; : The maximum processing rate value that occurs during task processing, indicating the processing peak value of the task at a certain moment; : Indicates the number of recursive accumulation stages during task execution; A multi-objective optimization model based on genetic algorithms balances task computing requirements and network resources to obtain the optimal task splitting solution, including the following: In the genetic algorithm, an initial population containing multiple task splitting schemes is first generated. Each individual represents a possible task splitting scheme, including the number of subtasks into which the task is split and the size of each subtask. Assuming that the task needs to be split into M subtasks, each individual is represented by a vector ,in Indicates The size of each subtask, the dimension of each individual is associated with the available resource status of the network element device; The generation of the initial population combines the load regulation index and task load intensity index of all current network element devices, and takes the load regulation index of the network element device as a constraint to ensure that the subtasks after task splitting will not exceed the resource carrying capacity of the network element device; the task load intensity index of the task is taken as the target, so that the split subtasks can balance the task processing efficiency and the network element device load; The quality of each individual is determined by its fitness function. The fitness function combines the load regulation index and the task load intensity index to optimize through two core goals: one is to balance the computing requirements of the task and the resource utilization of the network element equipment, and the other is to maximize the overall task processing efficiency of the system. The fitness function is designed as a comprehensive function of the two main goals: ;in: :Network element equipment Load regulation index; : The average load regulation index of all network element devices in the current system; : After the task is split into subtasks, the sum of the task load intensity indexes of each subtask; : A very small positive number to prevent division by zero errors; Based on the fitness function, selection, crossover and mutation operations are performed to continuously generate new task splitting solutions and gradually approach the optimal solution. After multiple generations of iterations, the algorithm finally converges and obtains the optimal task splitting solution. According to the optimal splitting scheme, the task is split into multiple subtasks, and the subtasks are preferentially assigned to low-load network element devices based on the network topology and delay conditions; Split or merge existing subtasks based on task requirements and network element equipment resource conditions.

2. A data communication method based on future network according to claim 1, characterized in that: According to the optimal splitting scheme, the task is split into multiple subtasks. Combined with the network topology and delay, the subtasks are preferentially assigned to low-load network element devices, including the following: In order to ensure that the subtasks of a task can be preferentially assigned to the most appropriate network element device, the topology of the entire network is first obtained; the network topology is represented in the form of a weighted directed graph, where each node represents a network element device, and the weight of each edge reflects the delay or bandwidth limitation between network element devices; In order to allocate task subtasks to the most appropriate network element devices, a path priority model is constructed, which comprehensively considers the link delay, bandwidth availability and load regulation index of network element devices in the topology; the priority of each path Calculated by the following formula: ;in: : represents the delay of the corresponding path; : The average value of the load regulation index of all network element devices passing through the corresponding path; : Minimum link bandwidth on the corresponding path; : A positive number used to prevent division by zero errors.

3. A data communication method based on future network according to claim 2, characterized in that: Based on the optimal task splitting solution, the task is split into multiple subtasks, and these subtasks are assigned to different network element devices according to the path priority model. The specific task allocation process is implemented through the following steps: According to the size of each subtask in the optimal splitting solution, the subtask with small load is preferentially allocated to the network element device with low load; According to the path priority between each network element device and the source node, the path with the highest priority is selected, and the task subtasks are assigned to the network element devices on the corresponding path; After the subtasks are assigned to the network element devices, the execution status of the subtasks of each network element device is monitored, and the execution progress of the subtasks is adjusted in real time according to the load regulation index. When the load index of a certain network element device rises rapidly, the reallocation mechanism is triggered to transfer some unfinished subtasks to other network element devices with lower loads.

4. A data communication method based on future network according to claim 3, characterized in that: According to the task requirements and network element equipment resources, the existing subtasks are split or merged, including the following: According to the changes in the network element load index and task load index, the task splitting or merging operation is automatically triggered. When the following two situations are detected, different operation strategies will be adopted: When the task load intensity index is higher than the task load threshold and the load regulation index is lower than the load threshold, the task splitting operation will be triggered; When the task load intensity index is lower than the task load threshold and the load regulation index is higher than the load threshold, the task merging operation will be triggered.

5. A data communication system based on a future network, used to implement a data communication method based on a future network as claimed in any one of claims 1 to 4, characterized in that: include: Monitoring module, optimization module, allocation module and adjustment module; Monitoring module: monitors the usage of each network element device in real time, analyzes the network element device load and task complexity, reflects the status of tasks and network element devices, and transmits the analyzed network element device load and task complexity to the optimization module; Optimization module: A multi-objective optimization model based on genetic algorithms balances task computing requirements and network resources, obtains the optimal task splitting solution, and passes the optimal splitting solution to the allocation module; Allocation module: splits the task into multiple subtasks according to the optimal splitting scheme, and preferentially allocates the subtasks to low-load network element devices based on the network topology and delay conditions. The matching results between the task and the network element device are sent to the adjustment module. Adjustment module: Split or merge existing subtasks according to task requirements and network element equipment resources.

Images