Method and system for balancing intelligent traffic data load by using intelligent network card

Abstract

The invention belongs to the field of intelligent traffic data management, and provides a method and system for balancing intelligent traffic data load by using an intelligent network card, and the balancing method comprises the steps: selecting the intelligent network card supporting link aggregation, and configuring a link aggregation group; a link through which the data flow passes is determined through the load balancing algorithm, and according to the determined data link, the data link comprises a source IP address, a target IP address, a source port number and a target port number, so that the data flow realizes data flow priority-based load balancing at the data peak moment on different physical links; through the intelligent network card at a data processing receiving end, according to the relationship between the data processing capability and the data processing demand, the intelligent network card allocates different physical links to the network connection request by dynamically adjusting a load balancing strategy, so that load balancing is realized.

Description

Technical Field

[0001] The present invention belongs to the field of smart traffic data management, and specifically relates to a method and system for balancing smart traffic data load by using an intelligent network card. Background Art

[0002] Today's smart transportation systems involve a variety of sensors, communication devices, and smart devices, such as traffic cameras, vehicle-mounted sensors, GPS, smart traffic signal controllers, and smart traffic management centers. The data generated by these devices involves information on traffic flow, vehicle speed, vehicle location, road conditions, and traffic events. The hyperscale data center for smart transportation refers to a centralized platform for collecting, processing, and analyzing the massive amounts of data generated by the smart transportation system. The platform efficiently collects and processes this data and converts it into useful information to help traffic managers make better decisions and improve road safety and traffic efficiency. To achieve this goal, smart transportation data centers usually use big data technologies and cloud computing technologies, such as distributed storage, streaming computing, data mining, and machine learning.

[0003] Smart traffic data has the characteristics of large data volume, prominent peak value, diverse data types, high data real-time requirements, strong spatiotemporal characteristics of data, and high data quality requirements. The sources of smart traffic data are extensive, including sensors, cameras, vehicle-mounted equipment and other channels, and the data volume is large. At the same time, traffic data needs to be collected and processed in real time so that it can respond and adjust to traffic conditions in a timely manner. As a network accelerator, smart network cards can delegate network processing and data processing capabilities to smart network cards, thereby reducing the load pressure on the server and improving the throughput and response speed of the system. In the ultra-large-scale data center of smart transportation, smart network cards can use their powerful network processing and computing capabilities to allocate peak workloads and provide better support for load balancing in data centers.

[0004] Since the smart traffic data channel is customized, pipeline parallelism, data parallelism and request parallelism can be used simultaneously to reduce data processing delay and improve data throughput. According to the dependencies between traffic data and traffic control, pipeline parallelism is used for dependencies, and data parallelism is used for non-dependencies. Then, request parallelism is used to maintain the load balance of pipeline parallelism and data parallelism between processing units. Instruction-based GPUs and CPUs have limited use of pipeline parallelism. Although the instruction processing core is pipelined, its depth is limited. Due to the high communication overhead between multiple cores, the efficiency of using multiple cores to form a pipeline is often low.

[0005] In order to solve the load balancing problem of smart transportation decision-making data, it is necessary to provide a method and system for balancing the load of smart transportation data using a smart network card. At the data processing receiving end, the smart network card can dynamically adjust the load balancing strategy according to the relationship between data processing capability and data processing requirements, and allocate network connection requests to different physical links to achieve load balancing. Summary of the invention

[0006] In order to solve the problems raised in the above background technology, a method and system for balancing the smart traffic data load using a smart network card is proposed. At the data processing receiving end, the smart network card dynamically adjusts the load balancing strategy according to the relationship between data processing capability and data processing requirements, and distributes network connection requests to different physical links to achieve load balancing.

[0007] To achieve the above-mentioned purpose, the first aspect of the present invention provides a method for balancing the load of smart traffic data using a smart network card, comprising: a method for balancing the load of smart traffic data using a smart network card, comprising the steps of: S1, select a smart network card (23) that supports link aggregation, and select a supported link aggregation group through a link aggregation protocol; S2, configuring a link aggregation group: in the settings of the smart network card (23), creating a link aggregation group, specifying a number or name of the link aggregation group, adding physical links or network interfaces to the group, and each link aggregation group has the same speed and link type; S3, configure the load balancing algorithm: determine the link through which the data traffic passes through by the load balancing algorithm, and according to the determined data link, the data link includes: source IP address, target IP address, source port number, target port number, so that the data traffic can be load balanced based on the data traffic priority at the data peak time on different physical links; S4, configuring the link aggregation protocol: configuring the parameters of the link aggregation protocol, including the mode and priority, according to the data link determined by the algorithm, so that the link aggregation protocol configuration between the smart network card (23) and the load balancer (21) / switch (22) is consistent; S5, verification and testing: After the configuration is completed, verification and testing are performed to check the status of the smart network card (23) and the status of the link aggregation group to ensure that all physical links are correctly aggregated together. By sending traffic and monitoring the traffic conditions of each link, it is verified whether the load balancing algorithm is working properly; if not, restart S3 to recalculate.

[0008] Among them, the preferred solution is: the step S3 includes: Step S31: Establishing the objective function and constraints of load balancing based on data traffic priority: (A1) In formula A1, the set of priorities is: , For the The priority of the data traffic request is Level, Level or Level, among which The priority of level is greater than Level, Level greater than level, It indicates the The impact of bandwidth response requests for individual data flows on the traffic diversion capacity of the entire intelligent transportation system after being responded to at peak times. The higher the priority, the greater the impact on traffic diversion capacity. When allocating bandwidth to data flows during load balancing, it should be responded to first. for Whether the data flow request is responded to, the set is: , the value range is 0 or 1, the value range is 0 means no response, the value range is 1 means response, that is, Each data flow request is responded to 1 or not responded to 0 at this point in time; Also need to meet: The set of constraints is: , not only satisfying The constraints on data traffic demand must also be met is the link bandwidth load constraint, where The constraint is the data flow demand. For any data flow demand, the required bandwidth must fully meet the data flow demand before it can be responded to. The link bandwidth load constraint means that the total load required by any link bandwidth allocated to each data flow cannot exceed the load of the link. At the same time, the allocated bandwidth cannot be allocated repeatedly at a certain time. The first restriction condition is: , (A2) In formula A2, , for The total bandwidth required for each data flow; for The total bandwidth available through each data link; The second restriction condition is: (A3) In formula A3, ; ; For the The data traffic is divided into The decision variable for the size of the load bandwidth provided by the data link, with a value range of 0 to ,; Step S32: Calculate the load balancing operator. The formula is: , (A4), The load balancing operator Based on the load, data traffic request and data traffic request priority, the load balancing operator can be determined comprehensively to maximize the load capacity when the load capacity is insufficient. The load balancing operator is obtained by the ratio of high priority and low bandwidth data traffic. Calculate its ability to solve traffic diversion while balancing load consumption; Step S33: in accordance with The values ​​of are sorted in descending order, and the sorting is recorded ,Record and The subscript correspondence of each element ( ),make The corresponding response parameter value for ; Step S34: First All elements of are assigned to 0, and then , ; Step S35: Determination ,if (A5), then , ,otherwise , ,when When, go to step S36, otherwise, repeat step S35; Step S36: Recorded and The subscript correspondence of each element ( ),Will The corresponding variable assignments are mapped to ; Step S37: According to middle The assignment determines the corresponding Whether the data flow request is responded to, the bandwidth according to Order, give middle The data traffic request depends on Assign values ​​in sequence. Then call the next data flow and assign a value. Then call the next link bandwidth pair Please assign values ​​to all parts until all The bandwidth requested by the data traffic is allocated and the output .

[0009] Among them, the preferred solution is: or , where the priority of level 9 is greater than level 7, the priority of level 7 is greater than level 5, the priority of level 5 is greater than level 3, and level 3 is greater than level 1, indicating that the priority of level 9 is greater than level 7, the priority of level 7 is greater than level 5, and the priority of level 5 is greater than level 3, and level 3 is greater than level 1. The bandwidth response request of each data flow will affect the traffic diversion capacity of the entire intelligent transportation system after being responded to at the peak time. The greater the priority, the greater the impact on the traffic diversion capacity. When allocating bandwidth to data flow in load balancing, it should be responded to first. The value selected by the priority level is based on the load balancing operator. The degree of distinction is large enough.

[0010] Among them, the preferred solution is: , the priority of the 500th level is greater than the 300th level, and the priority of the 300th level is greater than the 100th level, The value selected by the priority level is based on the load balancing operator The degree of distinction is large enough.

[0011] Among them, the preferred solution is: step 5 also includes: Step S51: Select data traffic to generate different test tools; Step S52: configuring the data traffic generator and the target host to ensure that the link between the data traffic generator and the target host is the link to be monitored; Step S53: Setting data traffic parameters: Setting required data traffic parameters in the traffic generation tool, including: priority, traffic type, and data packet size; Step S54: Start the data traffic generator: start the test tool on the data traffic generator and the target host to start generating data traffic; Step S55: Monitor the link data traffic: After collecting data traffic data, analyze and compare the traffic distribution, bandwidth utilization, and delay indicators on the link to evaluate the load balancing effect and link performance: is the flow distribution, Indicates the traffic distribution on each link. (A6), is the bandwidth utilization, for The bandwidth available on the data link, for The actual bandwidth used by the data link, (A7), Indicates the delay index on each link, which is the link The average delay of data transmission, Indicates The receiving time of the data packet, Indicates The sending time of the data packet, the load balancing effect is (A8), the load balancing effect refers to the effect of the load balancing algorithm in balancing traffic distribution, improving bandwidth utilization and reducing latency indicators, where A function that calculates the load balancing effect based on traffic distribution, bandwidth utilization, and latency indicators.

[0012] Among them, the preferred solution is: the data testing tool is: iPerf, Netperf or TCPBench testing tool to generate TCP or UDP.

[0013] The present invention also includes a smart network card applied to a smart traffic data load balancing system, comprising: a data flow collection terminal collects required data, sends a data response request of data flow type, data packet size, source IP address of the data, and source port number to a data access terminal, and a load balancer or switch of the data access terminal. The load balancer or switch uses the load balancing algorithm of the smart network card therein to decide whether to send the data response request to a specific corresponding server or virtual machine, and pushes the response data packet and data flow type to a target data processor of a target IP address for processing.

[0014] Among them, the preferred scheme is: the data stream collection terminal includes: electric police data collection terminal, UWB data collector, drone data collection terminal, FRID data collector, video collection terminal, positioning and navigation data collection terminal, traffic signal control data collection terminal, microwave data collection terminal and checkpoint data collection terminal, and defines the priority of the collected data according to the data collection period, and sends the data response request of the collected data flow type, data packet size, source IP address and source port number to the load balancer or switch.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention solves the problem of congestion in intelligent traffic data transmission and processing. Priority levels are used to process traffic decision data; 2. When solving the intelligent traffic decision data, the first restriction condition of the present invention is: Due to the constraints of data flow demand and overall data bandwidth limitation, there is a transmission conflict problem, that is, the data bandwidth cannot meet the simultaneous transmission of data. The balanced allocation method solves the transmission problem; 3. In order to ensure the integrity of the decision data, the second restriction condition of the present invention is adopted: Due to the constraint of link bandwidth load, in the process of processing data packet transmission, a data packet must be considered as a whole, so as to solve the decision problem of whether the data bandwidth is equal to the total amount of traffic decision packets. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0017] Figure 1 A flow chart of a method for balancing smart traffic data load using a smart network card proposed by the present invention; Figure 2 The present invention proposes an intelligent network card that is applied to a balancing framework of smart traffic data load. DETAILED DESCRIPTION

[0018] 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.

[0019] Figure 1 is a flow chart of the intelligent traffic data load balancing method of the present invention, such as Figure 1 As shown: The method includes: S1, select the smart network card that supports link aggregation: select the supported link aggregation through the Link Aggregation Control Protocol (LACP); S2, configure link aggregation group: in the settings of the smart NIC, create a link aggregation group (LAG), specify the number or name of the aggregation group, add physical links or network interfaces to the group, and each aggregation group has the same speed and link type; S3, configure the load balancing algorithm: determine the link through which the data traffic passes through the load balancing algorithm, and determine the source IP address, target IP address, source port number, and target port number of the data based on the determined data link, so that the data traffic can be load balanced based on the data traffic priority at the data peak time on different physical links; S4, configure the link aggregation protocol: according to the data link determined by the algorithm, configure the parameters of the link aggregation protocol, including mode and priority, so that the link aggregation protocol configuration between the smart network card and the switch is consistent; S5, verification and testing: After the configuration is completed, verification and testing are performed to check the status of the smart network card and the status of the link aggregation group to ensure that all physical links are correctly aggregated together. By sending traffic and monitoring the traffic conditions of each link, verify whether the load balancing algorithm is working properly.

[0020] Wherein: the step S3 further includes: Step S31: Establishing the objective function and constraints of load balancing based on data traffic priority: (A1) Wherein, formula A1 represents the set of n data flow priorities that are responsive or not, The set of priorities selected in the present invention is For the The priority of the data traffic request is Level, Level or Level, among which, , as well as is a natural constant. For example, when Level 5 has a higher priority than level 3, and level 3 has a higher priority than level 1. The bandwidth response request of each data flow will have an impact on the traffic diversion capacity of the entire intelligent transportation system after being responded to at the peak time. The greater the priority, the greater the impact on the traffic diversion capacity. When allocating bandwidth to data flow in load balancing, it should be responded to first. Of course, in order to achieve more accurate priority, it is not limited to being divided into the three levels mentioned above, and can be divided into more levels. For example, The value selected for the priority level is for the later load balancing operator. The distinction is large enough.

[0021] for Whether the data flow request is responded to, the set is: , the value range is 0 or 1, the value range is 0 means no response, the value range is 1 means response, that is, A data traffic request is responded to 1 or not responded to 0 at this point in time.

[0022] , (A2) Among them, formula A2 is the first restriction condition: any The data flow restriction conditions are: The bandwidth required for each data flow is zero or The bandwidth provided by the data link is greater than or equal to The bandwidth required for data traffic, where , For the The bandwidth required for each data flow; for The bandwidth available on the data link; (A3) Among them, formula A3 is the second restriction condition: any The data traffic is divided into The decision variable for the size of the load bandwidth provided by the data link, where ; ;

[0023] For the The data traffic is divided into The decision variable for the size of the load bandwidth provided by the data link, with a value range of 0 to , is a set of constraints, and the total constraints are: The constraint set not only satisfies The constraints on data traffic demand must also be met is the link bandwidth load constraint, where The data traffic demand constraint means that for any data traffic demand, the required bandwidth must fully meet the data traffic demand in order to be responded to. The link bandwidth load constraint means that the total load required by any link bandwidth allocated to each data flow cannot exceed the load of the link, and the allocated bandwidth cannot be allocated repeatedly at a certain time. Step S32: Calculate the load balancing operator. The formula is: , ……(A4), The load balancing operator For the Data priority and The ratio of data traffic to determine the load balancing operator can be fully reflected in the case of insufficient load capacity. The load balancing operator is obtained by the ratio of high priority and low bandwidth data traffic. To calculate its ability to solve traffic diversion and balance load consumption; Step S33: Initial , based on Sort by descending order of the values, and record In this way, the corresponding solution is to solve the problem of traffic diversion capacity while balancing the load consumption. Arrange and record and The subscript correspondence of each element ( ),make The corresponding response parameter value for ; Step S33: First All elements of are assigned 0; Step S34: , ; Step S35: Determination ,if (A5), then , ,otherwise , ,when When, go to step S36, otherwise, repeat step S35; Step S36: Recorded and The subscript correspondence of each element ( ),Will The corresponding variable assignments are mapped to ; Step S37: According to middle The assignment determines the corresponding Whether the data flow request is responded to, the bandwidth according to Order, give middle The data traffic request depends on Assign values ​​in sequence. Then call the next data flow and assign a value. Then call the next link bandwidth pair Please assign values ​​to all parts until all The bandwidth requested by the data traffic is allocated and the output .

[0024] In order to understand step 31 to step 37 in more detail, the first embodiment is used to describe them in detail. In this embodiment, if i=10, it means that there are 10 data flow requests that need to be processed. At this time, the priority set of the data flow request is , for Whether the data flow request is responded to, the value range is 0 or 1, where the value range is 0 for no response and the value range is 1 for response, that is, A data traffic request is responded to or not responded to at this point in time. For the The bandwidth required by each data flow. In this embodiment, i=10, and the bandwidths requested by the 10 data flows are , for Data traffic from The data link provides a decision variable for the size of the load bandwidth, and the value range of the decision variable is 0 to , for The bandwidth that the data link can provide, in this embodiment, when j=3, , is a constraint set, which is represented by the following two-element array: The data traffic demand constraint means that for any data traffic demand, the required bandwidth must fully meet the data traffic demand in order to be responded to. The link bandwidth load constraint means that the total load required by any link bandwidth allocated to each data flow cannot exceed the load of the link, and the allocated bandwidth cannot be allocated repeatedly at a certain time. According to step S32: according to and The ratio of is used to calculate the load balancing operator. In this embodiment, and but, ,Based on the load, data traffic request and the priority of data traffic request, ,the load balancing operator can be fully determined to maximize the ,performance when the load capacity is insufficient. By responding to ,highpriority and low-bandwidth data traffic, the load consumption can be balanced while solving ,the traffic diversion capacity; Step S33: According to the present embodiment Sort in descending order, remember ,Record and The subscript correspondence of each element ( ), so, according to the corresponding relationship make Corresponding for ; First All elements of are assigned the value 0. ; Step S34: , ; In this embodiment, step S35 is performed: first determining ,Compare (A5), then , , then judge +330=390, ; Calculate in sequence, when ,judge when ,judge , obviously processed data In the collection, when processing the 7th data, that is In this embodiment, Processing of data The total link bandwidth is 1000, which cannot handle the total load of 1050 required by data traffic. When , repeat step S35, thus obtaining the above sorted The data flow request response set is ; Step S36: Recorded and The subscript correspondence of each element ( ),Will The corresponding variable assignments are mapped to ; Step S37: According to middle The assignment determines the corresponding Whether the data flow request is responded to, the bandwidth according to Order, give middle The data traffic request depends on Assign values ​​in sequence. Then call the next data flow and assign a value. Then call the next link bandwidth pair Please assign values ​​to all parts until all The bandwidth allocation for data traffic requests is completed. According to the data in this implementation, the output .

[0025] In order to understand step 31 to step 37 in more detail, the second embodiment is used to describe them in detail. In this embodiment, if i=5, it means that there are 5 data flow requests that need to be processed. At this time, the priority set of the data flow request is , for Whether the data flow request is responded to, the value range is 0 or 1, where the value range is 0 for no response and the value range is 1 for response, that is, The data traffic requests are responded to or not responded to at this time point. In this embodiment, i=5, and the bandwidths of the 5 data traffic requests are respectively , for Data traffic from The data link provides a decision variable for the size of the load bandwidth, and the value range of the decision variable is 0 to , for The bandwidth that the data link can provide, in this embodiment, when j=4, , is a constraint set, which is represented by the following two-element array: The data traffic demand constraint means that for any data traffic demand, the required bandwidth must fully meet the data traffic demand in order to be responded to. The link bandwidth load constraint means that the total load required by any link bandwidth allocated to each data flow cannot exceed the load of the link, and the allocated bandwidth cannot be allocated repeatedly at a certain time. According to step S32: according to and The ratio of is used to calculate the load balancing operator. In this embodiment, and but, ,Based on the load, data traffic request and the priority of data traffic request, ,the load balancing operator can be fully determined to maximize the ,performance when the load capacity is insufficient. By responding to ,highpriority and low-bandwidth data traffic, the load consumption can be balanced while solving ,the traffic diversion capacity; Step S33: According to the present embodiment Sort in descending order, remember ,Record and The subscript correspondence of each element ( ), so, according to the corresponding relationship make Corresponding for ; First All elements of are assigned the value 0. ; Step S34: , ; In this embodiment, step S35 is performed: first determining ,Compare (A5), then , , then judge +250=370, ; Calculate in sequence, when ,judge , obviously processed data In the collection, when processing the third data, that is In this embodiment, Processing of data The total link bandwidth is 380, which cannot handle the total load of 430 required by data traffic. When , repeat step S35, thus obtaining the above sorted The data flow request response set is ; Step S36: Recorded and The subscript correspondence of each element ( ),Will The corresponding variable assignments are mapped to ; Step S37: According to middle The assignment determines the corresponding Whether the data flow request is responded to, the bandwidth according to Order, give middle The data traffic request depends on Assign values ​​in sequence. Then call the next data flow and assign a value. Then call the next link bandwidth pair Please assign values ​​to all parts until all The bandwidth allocation for data traffic requests is completed. According to the data in this implementation, the output Step S51: Select data traffic to generate different network performance test data iPerf, Netperf or TCPBench to generate TCP or UDP.

[0026] By using the verification calculation of the first embodiment and the second embodiment, the method of the present invention can calculate the load balancing operator proposed by the present invention through the priority level and the bandwidth ratio when processing the decision data with priority through steps 31 to 37. , and based on Will Arrange the data in order to be processed And comprehensive consideration In order to meet the link bandwidth load carrying capacity, the transmission of decision data is reasonably allocated, and finally the data is distributed to different branches in turn.

[0027] Among them, step S52 of the present invention: configure the data traffic generator and the target host to ensure that the link between the data traffic generator and the target host is the link to be monitored; Step S53: Setting data traffic parameters: Setting required data traffic parameters in the traffic generation tool, including: priority, traffic type, and data packet size; Step S54: Start the data traffic generator: start iPerf, Netperf or TCPBench on the data traffic generator and the target host to start generating data traffic; Step S55: Monitor the link data traffic: After collecting data traffic data, analyze and compare the traffic distribution, bandwidth utilization, and delay indicators on the link to evaluate the load balancing effect and link performance: is the flow distribution, Indicates the traffic distribution on each link. (A6), is the bandwidth utilization, for The bandwidth available on the data link, for The actual bandwidth used by the data link, (A7), Indicates the delay index on each link, which is the link The average delay of data transmission, Indicates The receiving time of the data packet, Indicates The sending time of the data packet, the load balancing effect is (A8), the load balancing effect refers to the effect of the load balancing algorithm in balancing traffic distribution, improving bandwidth utilization and reducing latency indicators, where A function that calculates the load balancing effect based on traffic distribution, bandwidth utilization, and latency indicators.

[0028] The bandwidth utilization rate of the first embodiment in step S55 is: =975 / 1000=97.5%, and the bandwidth utilization of implementation 2 is: =370 / 380=97.3%, the bandwidth utilization rate is relatively high.

[0029] Figure 2 The present invention is a schematic diagram of the structure of the intelligent network card applied to the intelligent traffic data load balancing architecture, which adopts the load balancing method as described above to perform the intelligent traffic data load balancing architecture, including: a data stream access terminal 20, the data stream access terminal 20 includes a load balancer 21 or a switch 22, an intelligent network card 23, a server 24 or a virtual machine 25, and the data stream acquisition terminal 10 includes: an electric police data acquisition terminal 11, a UWB data collector 12, a drone data acquisition terminal 13, a FRID data collector 14, a video acquisition terminal 15, a positioning and navigation data acquisition terminal 16, a traffic signal control data acquisition terminal 17, a microwave data acquisition terminal 18, and a checkpoint data acquisition terminal 19.

[0030] Among them, the electric police data acquisition terminal 11, UWB data collector 12, drone data acquisition terminal 13, FRID data collector 14, video acquisition terminal 15, positioning and navigation data acquisition terminal 16, traffic signal control data acquisition terminal 17, microwave data acquisition terminal 18, and checkpoint data acquisition terminal 19 define priorities for the collected data at the same time or according to the data collection period, and send data response requests with data flow type, data packet size, source IP address of the data, and source port number to the load balancer 21 or switch 22. The balancer 21 or switch 22 uses the load balancing algorithm of the above-mentioned smart network card 23 to decide to send the data response request to the specific corresponding server 24 or virtual machine 25, and pushes the response data packet and data flow type to the target data processor 30 of the target IP address for processing.

[0031] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention solves the problem of congestion in intelligent traffic data transmission and processing. Priority levels are used to process traffic decision data; 2. When solving the intelligent traffic decision data, the first restriction condition of the present invention is: Due to the constraints of data flow demand and overall data bandwidth limitation, there is a transmission conflict problem, that is, the data bandwidth cannot meet the simultaneous transmission of data. The balanced allocation method solves the transmission problem; 3. In order to ensure the integrity of the decision data, the second restriction condition of the present invention is adopted: Due to the constraint of link bandwidth load, in the process of processing data packet transmission, a data packet must be considered as a whole, so as to solve the decision problem of whether the data bandwidth is equal to the total amount of traffic decision packets.

[0032] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned wireless terminal can refer to the corresponding process in the above-mentioned method embodiment, and will not be repeated here.

[0033] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for balancing smart traffic data load using a smart network card, characterized in that: Includes steps: S1, select a smart network card (23) that supports link aggregation, and select a supported link aggregation group through a link aggregation protocol; S2, configuring a link aggregation group: in the settings of the smart network card (23), creating a link aggregation group, each link aggregation group having the same speed and link type; S3, configure the objective function and constraints of data traffic priority load balancing: through the load balancing algorithm, determine the data traffic on different physical links when the data peak is based on the data traffic priority load balancing; S4, configuring the link aggregation protocol: configuring the parameters of the link aggregation protocol according to the data link determined by the algorithm, so that the link aggregation protocol configuration between the smart network card (23) and the load balancer (21) / switch (22) is consistent; S5, verification and testing: After the configuration is completed, verification and testing are performed to check the status of the smart network card (23) and the status of the link aggregation group, ensure that all physical links are correctly aggregated together, send traffic and monitor the traffic status of each link, and verify whether the load balancing algorithm is working properly; if not, restart S3 for reallocation.

2. The method for balancing smart traffic data load using a smart network card as claimed in claim 1, characterized in that: The step S3 comprises: Step S31: Establishing the objective function and constraints of load balancing based on data traffic priority: （A1） In formula A1, the set of priorities is: , For the The priority of each data traffic request, for Whether the data flow request is responded to; The constraints are: ,in, The data traffic demand is constrained by the bandwidth required to meet the data traffic demand; The link bandwidth load constraint means that the total load required by any link bandwidth allocated to each data flow should not exceed the load of the link; Step S32: Calculate the load balancing operator. The formula is: ， （A4）， In formula A4, , for The total bandwidth required for each data flow. Step S33: in accordance with The values ​​of Re-order, sort postscript ,Record and The subscript correspondence of each element ( ),make The corresponding response parameter value for ; Step S34: First All elements of are assigned to 0, and then , ; Step S35: Determination ,if (A5), then , ,otherwise , ,when When, go to step S36, otherwise, repeat step S35; Step S36: Recorded and The subscript correspondence of each element ( ),Will The corresponding variable assignments are mapped to ; Step S37: According to middle The assignment determines the corresponding Whether the data flow request is responded to, the bandwidth according to Order, give middle The data traffic request depends on Assign values ​​in sequence. Then call the next data flow and assign a value. Then call the next link bandwidth pair Please assign values ​​to all parts until all The bandwidth requested by the data traffic is allocated and the output is .

3. The method for balancing smart traffic data load using a smart network card as claimed in claim 2, characterized in that: Constraints: , in, , (A2) In formula A2, , for The total bandwidth required for each data flow; for The total bandwidth that can be provided by the data link; （A3） In formula A3, ; ; For the The data traffic is divided into The decision variable for the size of the load bandwidth provided by the data link, with a value range of 0 to .

4. The method for balancing smart traffic data load using a smart network card as claimed in claim 2, characterized in that: Said or .

5. The method for balancing smart traffic data load using a smart network card as claimed in claim 2, characterized in that: Said .

6. The method for balancing smart traffic data load using a smart network card as claimed in claim 1, characterized in that: Step 5 also includes: Step S51: Select data traffic to generate different test tools; Step S52: configuring the data traffic generator and the target host to ensure that the link between the data traffic generator and the target host is the link to be monitored; Step S53: Setting data traffic parameters: Setting required data traffic parameters in the traffic generation tool, including: priority, traffic type, and data packet size; Step S54: Start the data traffic generator: start the test tool on the data traffic generator and the target host to start generating data traffic; Step S55: Monitor the link data traffic: After collecting data traffic data, analyze and compare the traffic distribution, bandwidth utilization, and delay indicators on the link.

7. The method for balancing smart traffic data load using a smart network card as claimed in claim 6, characterized in that: The data testing tool is: iPerf, Netperf or TCPBench testing tool to generate TCP or UDP.

8. The smart network card as claimed in claim 1 is applied to a smart traffic data load balancing system, characterized in that , including: a data flow collection terminal (10) collects required data, sends a data response request including data flow type, data packet size, source IP address of the data, and source port number to a data access terminal (20), a load balancer (21) or a switch (22) of the data access terminal (20), and the load balancer (21) or the switch (22) uses the load balancing algorithm of the smart network card (23) therein to determine whether to send the data response request to a specific corresponding server (24) or a virtual machine (25), and pushes the response data packet and data flow type to a target data processor (30) of a target IP address for processing.

9. The intelligent network card as claimed in claim 8 is applied to a smart traffic data load balancing system, characterized in that: The data stream collection terminal (10) comprises: an electric police data collection terminal (11), a UWB data collector (12), a drone data collection terminal (13), a RFID data collector (14), a video collection terminal (15), a positioning navigation data collection terminal (16), a traffic signal control data collection terminal (17), a microwave data collection terminal (18) and a checkpoint data collection terminal (19), and defines a priority for the collected data according to the data collection period, and sends a data response request for the collected data flow type, data packet size, source IP address and source port number to a load balancer (21) or a switch (22).

Images