Dynamic cooperative multi-path routing control method and device

Abstract

The invention provides a dynamic collaborative multi-path routing control method and device. The method comprises the following steps: determining a health score of each substream based on a performance index of the substream at the current moment; based on a first difference value between the health score of each substream and the health score mean value of all the substream, determining the weight adjustment amount of each substream at the current moment by adopting a PID control technology; and determining the weight of each substream at the next moment based on the weight adjustment amount of each substream at the current moment, and determining a multi-path routing control scheme at the next moment based on the weight of each substream at the next moment. According to the invention, rapid and stable switching of paths is realized through a dynamic cooperative routing control method.

Description

Technical Field

[0001] The present invention relates to the field of network communication technology, and in particular to a dynamic collaborative multi-path routing control method and device. Background Art

[0002] With the rapid development of 5G networks, industrial Internet, and integrated air-space-ground communication systems, multipath routing technology, as a core means to improve network resource utilization and transmission reliability, has become an important research direction in modern network architecture.

[0003] Multipath routing builds heterogeneous transmission paths by aggregating multiple network cards and multiple protocol stack resources on devices, achieving bandwidth aggregation, redundant fault tolerance, and load balancing in scenarios such as data centers, mobile terminals, and the Internet of Things.

[0004] However, in a dynamic network environment, the path status is highly time-varying. Traditional routing strategies based on fixed weights or single-dimensional indicators are difficult to adapt to complex business needs, resulting in frequent path switching, increased throughput fluctuations, and insufficient global stability, which seriously restricts the in-depth optimization of multi-path transmission performance.

[0005] Current research in multipath routing technology focuses on path selection and traffic distribution. Typical solutions include dynamic selection based on latency-sensitive minimum round-trip delay, weighted round-robin scheduling for throughput optimization, and redundant transmission mechanisms combined with network coding. The MultiPathTCP (MPTCP) protocol, proposed by the International Organization for Standardization, achieves multi-subflow coordination through coupled congestion control. However, its path management module still relies on static priority policies, making it ineffective in addressing sudden link quality degradation.

[0006] In summary, how to build a new routing control mechanism with multi-dimensional dynamic perception and global stable coordination has become a key challenge to break through the bottleneck of multi-path transmission performance. Summary of the Invention

[0007] The present invention provides a dynamic collaborative multi-path routing control method and device to address the defects of the prior art, such as the single path evaluation dimension, the rigid switching mechanism, and the insufficient global coordination capability, to realize a dynamic collaborative routing control method and achieve fast and stable path switching.

[0008] The present invention provides a dynamic collaborative multi-path routing control method, comprising: Determine the health score of each subflow based on the performance indicators of the subflow at the current moment; Based on a first difference between the health score of each subflow and the average health score of all subflows, a PID control technique is used to determine a weight adjustment amount for each subflow at the current moment; The weight of each subflow at the next moment is determined based on the weight adjustment amount of each subflow at the current moment, and the multipath routing control scheme at the next moment is determined based on the weight of each subflow at the next moment.

[0009] According to a dynamic collaborative multipath routing control method provided by the present invention, before the step of determining the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, the method further includes: Input the weight adjustment amount of the sub-flow at the current moment into the pre-constructed system energy function to obtain the energy function value of the sub-flow at the current moment; updating the proportional term weight and the differential term weight used in calculating the weight adjustment amount of the sub-flow at the next moment according to the energy function value of the sub-flow at the current moment; The relationship between the system energy function and the proportional term weight and the differential term weight is predefined.

[0010] According to a dynamic collaborative multipath routing control method provided by the present invention, the step of determining the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment specifically includes: The sigmoid function is used to determine the weight coefficient of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment and the weight adjustment amounts of other subflows in the system at the current moment.

[0011] According to a dynamic collaborative multi-path routing control method provided by the present invention, the weight adjustment amount of each sub-flow at the current moment is for: ; Where, 、 and are the weight coefficients of the proportional term, integral term and differential term respectively, represents the first difference of the i-th sub-stream, represents the synergy coefficient between the i-th subflow and the j-th subflow in the system except for it, N represents the number of subflows in the system except for the i-th subflow, represents the first difference of the j-th sub-stream, represents the coupled differential suppression factor.

[0012] According to a dynamic collaborative multipath routing control method provided by the present invention, the step of determining the multipath routing control scheme at the next moment based on the weight of each subflow at the next moment specifically includes: When the health score of the subflow is greater than a first preset threshold within a preset time period, the subflow is determined as an active subflow; When the health score of the sub-flow is less than a second preset threshold, determining the sub-flow as a disabled sub-flow; triggering a path switch for the deactivated sub-flow, and determining a sub-flow to be switched among the activated sub-flows based on the weight of each sub-flow at the next moment; The first preset threshold is greater than the second preset threshold.

[0013] According to a dynamic collaborative multi-path routing control method provided by the present invention, the first preset threshold As shown below: ; The second preset threshold As shown below: ; Where, and are the mean and standard deviation of the health scores of all sub-flows in the system at time t, is the system bandwidth utilization, m is the dispersion sensitivity coefficient, n is the dispersion tolerance coefficient, is the bandwidth utilization tolerance coefficient, is the overload protection factor.

[0014] The present invention also provides a dynamic coordinated multi-path routing control device, comprising: A determination module, configured to determine a health score of each subflow based on the performance indicators of the subflow at the current moment; an adjustment module, configured to determine a weight adjustment amount for each subflow at a current moment using a PID control technique based on a first difference between the health score of each subflow and an average health score of all subflows; The generating module is used to determine the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, and determine the multipath routing control scheme at the next moment based on the weight of each subflow at the next moment.

[0015] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the dynamic collaborative multi-path routing control method as described above is implemented.

[0016] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described dynamic collaborative multi-path routing control methods.

[0017] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any one of the above-mentioned dynamic collaborative multi-path routing control methods.

[0018] The dynamic collaborative multi-path routing control method and device provided by the present invention define the health score of each sub-flow and construct a PID controller model of the multi-path system based on the health score to introduce PID control into multi-path routing control, thereby integrating local response, integral mechanism and coupling differential mechanism to achieve real-time dynamic adjustment of sub-flow weights, so as to utilize the fast and stable adjustment characteristics of PID control for load balancing and network performance optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 This is one of the flow charts of the dynamic collaborative multi-path routing control method provided by the present invention; Figure 2 This is the second flow chart of the dynamic collaborative multi-path routing control method provided by the present invention; Figure 3 It is a structural diagram of the dynamic collaborative multi-path routing control device provided by the present invention; Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0021] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0022] The following combination Figure 1 and Figure 2 The present invention introduces a dynamic collaborative multi-path routing control method, such as Figure 1 As shown, including: Step 101, determining a health score for each subflow based on the performance index of the subflow at the current moment; Sub-flows are multiple transmission sub-flows of the system corresponding to the routing.

[0023] The performance indicators of the sub-flow are indicators used to evaluate the data transmission performance of the sub-flow, including but not limited to one or more of path delay, packet loss rate, throughput, jitter and bandwidth utilization.

[0024] The health score of each sub-flow represents the overall transmission health of the sub-flow.

[0025] Preferably, in this embodiment, the performance indicators of the subflow are determined as path delay, packet loss rate, and throughput. A health score model is constructed based on the values ​​of the above three indicators of the subflow at the current moment to determine the health score of the subflow at the current moment: ; Where, represents the health score of the subflow, 、 and They represent the normalized round-trip delay, packet loss rate, and throughput, respectively. 、 and Represent the weights corresponding to each indicator.

[0026] Among them, the weights corresponding to each indicator are predetermined according to the business type of the system and the selected performance indicators. Taking the above three performance indicators as an example, the emphasis is on delay-sensitive real-time streaming media transmission. , focusing on throughput-sensitive file transfer ,satisfy That's it.

[0027] For example, in a video transmission scenario, is 0.6, is 0.2, 0.2, as determined in the file download scenario is 0.2, is 0.2, is 0.6.

[0028] In addition, since the data scale corresponding to each performance indicator is different, it is necessary to standardize the collected values ​​of different performance indicators and then perform linear weighting to obtain the health score of the sub-flow.

[0029] In a feasible implementation, the collected values ​​of various performance indicators are directly standardized using the Z-Score method to avoid dimensional conflicts: ; Where, and are the mean and standard deviation of the indicator within the historical window. The length of the historical window is determined based on experience.

[0030] Preferably, in this embodiment, the round-trip delay, packet loss rate, and throughput data of the sub-flow are collected at a period of 1 second, and the RTT is exponentially smoothed to eliminate instantaneous jitter: ; Where, 、 and Represent the round-trip time delay after smoothing, the previous moment, and the current moment respectively.

[0031] On this basis, the smoothed RTT is normalized to calculate the health score of the sub-flow, and a more accurate health score is obtained.

[0032] Step 102: Based on a first difference between the health score of each subflow and the average health score of all subflows, a PID control technique is used to determine a weight adjustment amount for each subflow at the current moment. The control method in control theory in PID control constructs a triple mechanism of "current correction, historical leakage filling and predictive anti-shake" through the combination of three links: proportional, integral and differential, to achieve fast and stable adjustment of the system. This feature is also applicable to the control of multi-path routing. Therefore, the present invention proposes to introduce PID control into the control of multi-path routing.

[0033] Specifically, a PID controller constructs a deviation between a given value and the actual output value, and linearly combines the proportional, integral, and differential components of the deviation to form a control variable. To incorporate PID control into multipath routing, the present invention first defines a health score. This health score, which characterizes the comprehensive transmission performance of substreams, can be used to define a deviation for constructing a PID controller.

[0034] On this basis, in this embodiment, the health score of the i-th sub-flow can be The average health score of all sub-flows in the system The absolute value of the difference is taken as the first difference , build a basic PID controller.

[0035] The corresponding first difference is calculated based on the health score of the sub-flow at the current moment, and the weight adjustment amount of the sub-flow at the current moment output by the PID controller can be obtained.

[0036] Step 103: Determine the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, and determine the multipath routing control scheme at the next moment based on the weight of each subflow at the next moment.

[0037] The weight of each subflow refers to the proportion of traffic allocated to each transmission subflow in the current multipath system. A higher weight indicates a greater proportion of traffic allocated to each subflow. It can be understood that the weight of each subflow at the next moment is determined based on the weight adjustment of each subflow at the current moment. In other words, the weight adjustment value of each subflow at the current moment is mapped so that the sum of the weights of all subflows at the next moment is 1.

[0038] For example, the system includes three sub-flows, and the weight adjustment amounts calculated for the three sub-flows at the current moment are 1, 3, and 4, respectively. Then, the weights of the three sub-flows at the next moment are determined to be 0.125, 0.375, and 0.5.

[0039] Based on the above, a multi-path routing control solution can be generated based on the weight of each sub-flow at the next moment.

[0040] In a feasible implementation, a weight threshold can be set in advance. When the weight of the sub-stream at the next moment is greater than the weight threshold, it is considered that the data of the sub-stream is too much, and the data sent to the sub-stream will be sent to other sub-streams in order of weight from small to large.

[0041] The present invention defines a health score for each sub-flow and constructs a PID controller model for a multi-path system based on the health score, so as to introduce PID control into multi-path routing control. This integrates local response, integral mechanism, and coupled differential mechanism to achieve real-time dynamic adjustment of sub-flow weights, so as to utilize the fast and stable adjustment characteristics of PID control for load balancing and network performance optimization.

[0042] In the dynamic collaborative multi-path routing control method of the present invention, the weight adjustment amount of each sub-flow at the current moment is for: ; Where, 、 and are the weight coefficients of the proportional term, integral term and differential term respectively, represents the first difference of the i-th sub-stream, represents the synergy coefficient between the i-th subflow and the j-th subflow in the system except for it, N represents the number of subflows in the system except for the i-th subflow, represents the first difference of the j-th sub-stream, represents the coupled differential suppression factor.

[0043] Since traditional PID control usually only considers the control of a single device, in a multi-path routing system, a single sub-flow will also be affected by other sub-flows in the system. In order to take the influence of other sub-flows in the system into consideration as a factor in the weight adjustment amount of the single sub-flow, this embodiment provides an optimized weight adjustment amount calculation method, as shown in the above formula.

[0044] Optionally, the synergy coefficient is calculated as follows: ; Where, represents the health score of subflow j at the current moment, is the activation duration of subflow i, which can be generally understood as the usage time of subflow i, and K represents the total number of subflows in the system.

[0045] That is, the synergy coefficient The activation duration of subflow i is determined by coupling the ratio of the health score of subflow j to the total health score of the entire system, so that the activation duration is equal to Reverse change; the longer the activation time of sub-flow i is, the longer the sub-flow runs and the more stable it is. The smaller the value, the smaller the calculated weight adjustment amount; the shorter the activation time of sub-flow i, the shorter the running time of the sub-flow, and the more data it can accept. Therefore, the corresponding The larger it is, the larger the calculated weight adjustment will be.

[0046] Optionally, the coupled differential suppression factor , used to suppress the chain effect of other path changes on the current path.

[0047] Optionally, the integral term of the PID controller is a sliding window integral, and a 5-second window accumulation is selected to filter out occasional health deviations.

[0048] Optionally, the integral term of the PID controller Perform Kalman filtering and coupled differential to suppress noise: ; Where, 、 and are the differential values ​​after filtering, at the current moment, and at the next moment, respectively. The weight coefficients 0.8 and 0.2 are obtained based on the signal-to-noise ratio optimization in typical scenarios. Users can adjust the weight ratio to adapt to specific network environments. As calculation Used .

[0049] The present invention constructs a multi-distributed PID controller model for a multi-path system in the above manner, integrating local response, cross-path collaborative integration and coupled differential mechanisms. The sub-flow weights are adjusted in real time according to the path health, achieving better load balancing and network performance optimization.

[0050] In the dynamic collaborative multipath routing control method of the present invention, before the step of determining the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, the method further includes: Input the weight adjustment amount of the sub-flow at the current moment into the pre-constructed system energy function to obtain the energy function value of the sub-flow at the current moment; updating the proportional term weight and the differential term weight used in calculating the weight adjustment amount of the sub-flow at the next moment according to the energy function value of the sub-flow at the current moment; The relationship between the system energy function and the proportional term weight and the differential term weight is predefined.

[0051] Furthermore, although the weights corresponding to the proportional term, integral term, and differential term of the PID controller are 、 and It can be an empirical value obtained from multiple experiments, but in order to achieve better control effect, in this embodiment, a system energy function is constructed based on Lyapunov (Lyapunov stability), and then coupled to the weight calculation of at least one item of the PID controller.

[0052] The system energy function is used to constrain the system stability and accelerate the system to reach convergence stability. Optionally, a constructed system energy function is as follows: ; Where, is the weight change penalty factor, used to suppress drastic adjustments; is the load balancing intensity coefficient, forcing weights to be uniform; represents the weight difference between paths, represents the weight adjustment of the i-th path (i.e. the current path), is the first deviation of the i-th path at the current moment.

[0053] On this basis, the proportional term weight of the PID controller is defined Where, is the energy attenuation coefficient and the oscillation suppression strength.

[0054] Define the derivative term weight of the PID controller .

[0055] in, and are the proportional term weight and the differential term weight in the initial state, respectively. Optionally, they are set to 0.8 and 0.5, respectively, and are dynamically adjusted according to the system stability. V(t) is the system energy value at time t.

[0056] Optionally, Set to 0.1, Set to 0.3, Set to 0.5.

[0057] Determined by the above method Attenuation is achieved when the system oscillates, and damping is enhanced when the system diverges, achieving dynamic optimization of the entire system, thereby obtaining more accurate weight adjustments for sub-circuits.

[0058] In addition, in this embodiment, the integral item weight , in high discreteness scenarios ( Large) lower , suppressing the chain reaction caused by cross-path collaboration; in low-load scenarios ( <0.6) , accelerating long-term error correction.

[0059] It can be understood that based on the above definition, inputting the current moment , you can calculate the current moment , and according to the current moment Update the proportional term weight and the differential term weight to calculate at the next moment is used in the process of The overall stability of the system is taken into consideration during the calculation process to achieve dynamic optimization.

[0060] In the dynamic collaborative multipath routing control method of the present invention, the step of determining the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment specifically includes: The sigmoid function is used to determine the weight coefficient of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment and the weight adjustment amounts of other subflows in the system at the current moment.

[0061] In this embodiment, after determining the weight adjustment amount of each sub-flow at the current moment, the weight coefficient of each sub-flow at the next moment is obtained based on the normalization strategy of the sigmoid function. : ; Where, and Respectively represent the weight changes of the i-th path and the j-th path at time t, It is a temperature parameter that controls the sharpness of weight distribution.

[0062] In the dynamic collaborative multipath routing control method of the present invention, the step of determining the multipath routing control scheme at the next moment based on the weight of each subflow at the next moment specifically includes: When the health score of the subflow is greater than a first preset threshold within a preset time period, the subflow is determined as an active subflow; When the health score of the sub-flow is less than a second preset threshold, determining the sub-flow as a disabled sub-flow; triggering a path switch for the deactivated sub-flow, and determining a sub-flow to be switched among the activated sub-flows based on the weight of each sub-flow at the next moment; The first preset threshold is greater than the second preset threshold.

[0063] In this embodiment, in order to avoid system instability caused by frequent switching of sub-circuits as much as possible, all sub-circuits in the system are divided into three categories based on health scores and predetermined first and second preset thresholds.

[0064] Optionally, the first preset threshold and the second preset threshold are empirical values.

[0065] The first preset threshold is an active state threshold. If the health score of a sub-line at the current moment is greater than the first preset threshold, the sub-line is considered to carry the main data flow. Preferably, the system is forced to maintain at least two active paths.

[0066] The second preset threshold is the disabled state threshold. When the health score of the sub-circuit at the current moment is less than the disabled state threshold, the sub-circuit is immediately disabled and switched to the backup path, triggering the Lyapunov load balancing strategy.

[0067] If the health score of the sub-line at the current moment is greater than or equal to the second preset threshold and less than or equal to the first preset threshold, the sub-line is determined to be a standby line, downgraded and PID fine-tuning is triggered, and keep-alive detection packets are periodically sent.

[0068] In other words, a sub-circuit will only be deactivated and switched if it is inactive. A sub-circuit will only be considered active if its health score remains above the active threshold (i.e., above the first preset threshold for a predetermined period of time). Furthermore, a mandatory backup mechanism activates the best backup path historically based on its health when active paths are insufficient.

[0069] On this basis, optionally, when switching lines, data is sent in reverse order according to the weights of the active sub-flows at the next moment to achieve line switching. Through the above approach, the stability of the sub-line in the standby state is maintained as much as possible, thereby improving the overall stability of the system.

[0070] In the dynamic collaborative multi-path routing control method of the present invention, the first preset threshold As shown below: ; The second preset threshold As shown below: ; Where, and are the mean and standard deviation of the health scores of all sub-flows in the system at time t, is the system bandwidth utilization, m is the dispersion sensitivity coefficient, n is the dispersion tolerance coefficient, is the bandwidth utilization tolerance coefficient, is the overload protection factor.

[0071] In this embodiment, in order to obtain more accurate first preset threshold value and second preset threshold value, a dynamic calculation method of the first preset threshold value and the second preset threshold value is defined, as shown in the above formula.

[0072] Among them, the system bandwidth utilization represents the system load pressure, the dispersion sensitivity coefficient is dynamically adjusted, and the activation threshold sensitivity is increased when the health difference is large; the dispersion tolerance coefficient is used to lower the deactivation threshold when the health difference is large; the overload protection coefficient is used to significantly increase the deactivation threshold to force traffic diversion when the load exceeds 80%.

[0073] Coefficients m, n, and The default values ​​of m and n are 0.56 and 1.2 respectively, which are obtained by training with historical network data. is 0.3, The value 10 applies to 80% of typical scenarios and can be customized based on the network topology.

[0074] For example, if the load is low and the health is concentrated, assuming the parameter is 65, If the health of path A is 69, and L is 8, then the first preset threshold is calculated to be 68.8. If path A has a health of 69, then its health exceeds the activation threshold and it becomes active. If there are insufficient active paths (the system enforces the maintenance of at least two active paths), the backup path with the largest health difference is prioritized for replacement.

[0075] For example, if the health is dispersed in the overload protection scenario, assuming the parameter is 50, If the health of path B is 20 and L is 0.85, the calculated first preset threshold is 26.5. If the health of path B is 25, meaning it falls below the deactivation threshold, the deactivation action is immediately triggered, and traffic is switched to the backup path. At the same time, the system energy function is used to redistribute weights to avoid chain reactions caused by load concentration.

[0076] The present invention achieves rapid response and global optimization of path weights in dynamic network environments by constructing a multi-dimensional path state perception system, global coordinated proportional-integral-derivative control (PID), and a Lyapunov stability assurance mechanism. This effectively reduces path switching frequency and throughput fluctuations, and improves load balancing and fault recovery efficiency.

[0077] Among them, a complete process such as Figure 2 shown.

[0078] The following describes a dynamic coordinated multi-path routing control device provided by the present invention. The dynamic coordinated multi-path routing control device described below and the dynamic coordinated multi-path routing control method described above can be referenced to each other.

[0079] like Figure 3 As shown, the dynamic collaborative multi-path routing control device includes a determination module 301, an adjustment module 302 and a generation module 303; Determination module 301, for determining the health score of each sub-flow based on the performance index of the sub-flow at the current moment; Sub-flows are multiple transmission sub-flows of the system corresponding to the routing.

[0080] The performance indicators of the sub-flow are indicators used to evaluate the data transmission performance of the sub-flow, including but not limited to one or more of path delay, packet loss rate, throughput, jitter and bandwidth utilization.

[0081] The health score of each sub-flow represents the overall transmission health of the sub-flow.

[0082] Preferably, in this embodiment, the performance indicators of the subflow are determined as path delay, packet loss rate, and throughput. A health score model is constructed based on the values ​​of the above three indicators of the subflow at the current moment to determine the health score of the subflow at the current moment: ; Where, represents the health score of the subflow, 、 and They represent the normalized round-trip delay, packet loss rate, and throughput, respectively. 、 and Represent the weights corresponding to each indicator.

[0083] Among them, the weights corresponding to each indicator are predetermined according to the business type of the system and the selected performance indicators. Taking the above three performance indicators as an example, the emphasis is on delay-sensitive real-time streaming media transmission. , focusing on throughput-sensitive file transfer ,satisfy That's it.

[0084] For example, in a video transmission scenario, is 0.6, is 0.2, 0.2, as determined in the file download scenario is 0.2, is 0.2, is 0.6.

[0085] In addition, since the data scale corresponding to each performance indicator is different, it is necessary to standardize the collected values ​​of different performance indicators and then perform linear weighting to obtain the health score of the sub-flow.

[0086] An adjustment module 302 is configured to determine a weight adjustment amount for each subflow at a current moment using a PID control technique based on a first difference between the health score of each subflow and the average health score of all subflows; The control method in control theory in PID control constructs a triple mechanism of "current correction, historical leakage filling and predictive anti-shake" through the combination of three links: proportional, integral and differential, to achieve fast and stable adjustment of the system. This feature is also applicable to the control of multi-path routing. Therefore, the present invention proposes to introduce PID control into the control of multi-path routing.

[0087] Specifically, a PID controller constructs a deviation between a given value and the actual output value, and linearly combines the proportional, integral, and differential components of the deviation to form a control variable. To incorporate PID control into multipath routing, the present invention first defines a health score. This health score, which characterizes the comprehensive transmission performance of substreams, can be used to define a deviation for constructing a PID controller.

[0088] On this basis, in this embodiment, the health score of the i-th sub-flow can be The average health score of all sub-flows in the system The absolute value of the difference is taken as the first difference , build a basic PID controller.

[0089] The corresponding first difference is calculated based on the health score of the sub-flow at the current moment, and the weight adjustment amount of the sub-flow at the current moment output by the PID controller can be obtained.

[0090] The generating module 303 is configured to determine the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, and determine the multipath routing control scheme at the next moment based on the weight of each subflow at the next moment.

[0091] The weight of each subflow refers to the proportion of traffic allocated to each transmission subflow in the current multipath system. A higher weight indicates a greater proportion of traffic allocated to each subflow. It can be understood that the weight of each subflow at the next moment is determined based on the weight adjustment of each subflow at the current moment. In other words, the weight adjustment value of each subflow at the current moment is mapped so that the sum of the weights of all subflows at the next moment is 1.

[0092] For example, the system includes three sub-flows, and the weight adjustment amounts calculated for the three sub-flows at the current moment are 1, 3, and 4, respectively. Then, the weights of the three sub-flows at the next moment are determined to be 0.125, 0.375, and 0.5.

[0093] Based on the above, a multi-path routing control solution can be generated based on the weight of each sub-flow at the next moment.

[0094] In a feasible implementation, a weight threshold can be set in advance. When the weight of the sub-stream at the next moment is greater than the weight threshold, it is considered that the data of the sub-stream is too much, and the data sent to the sub-stream will be sent to other sub-streams in order of weight from small to large.

[0095] The present invention defines a health score for each sub-flow and constructs a PID controller model for a multi-path system based on the health score, so as to introduce PID control into multi-path routing control. This integrates local response, integral mechanism, and coupled differential mechanism to achieve real-time dynamic adjustment of sub-flow weights, so as to utilize the fast and stable adjustment characteristics of PID control for load balancing and network performance optimization.

[0096] Figure 4 An example of a physical structure diagram of an electronic device is shown below. Figure 4As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 may call logic instructions in the memory 430 to execute a dynamic collaborative multi-path routing control method, which includes: determining a health score for each sub-flow based on a performance indicator of the sub-flow at a current moment; determining a weight adjustment amount for each sub-flow at a current moment using a PID control technique based on a first difference between the health score of each sub-flow and the average health score of all sub-flows; determining a weight for each sub-flow at a next moment based on the weight adjustment amount for each sub-flow at a current moment, and determining a multi-path routing control scheme for a next moment based on the weight of each sub-flow at a next moment.

[0097] Furthermore, the logic instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0098] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the dynamic collaborative multi-path routing control method provided by the above methods, the method including: determining the health score of each sub-flow based on the performance index of the sub-flow at the current moment; determining the weight adjustment amount of each sub-flow at the current moment using PID control technology based on the first difference between the health score of each sub-flow and the average health score of all sub-flows; determining the weight of each sub-flow at the next moment based on the weight adjustment amount of each sub-flow at the current moment, and determining the multi-path routing control scheme at the next moment based on the weight of each sub-flow at the next moment.

[0099] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the dynamic collaborative multi-path routing control method provided by the above-mentioned methods, the method comprising: determining the health score of each sub-flow based on the performance index of the sub-flow at the current moment; determining the weight adjustment amount of each sub-flow at the current moment using PID control technology based on the first difference between the health score of each sub-flow and the average health score of all sub-flows; determining the weight of each sub-flow at the next moment based on the weight adjustment amount of each sub-flow at the current moment, and determining the multi-path routing control scheme at the next moment based on the weight of each sub-flow at the next moment.

[0100] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0101] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A dynamic collaborative multi-path routing control method, characterized in that: include: Determine the health score of each subflow based on the performance indicators of the subflow at the current moment; Based on a first difference between the health score of each subflow and the average health score of all subflows, a PID control technique is used to determine a weight adjustment amount for each subflow at the current moment; The weight of each subflow at the next moment is determined based on the weight adjustment amount of each subflow at the current moment, and the multipath routing control scheme at the next moment is determined based on the weight of each subflow at the next moment.

2. The dynamic coordinated multi-path routing control method according to claim 1, characterized in that: Before the step of determining the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, the method further includes: Input the weight adjustment amount of the sub-flow at the current moment into the pre-constructed system energy function to obtain the energy function value of the sub-flow at the current moment; updating the proportional term weight and the differential term weight used in calculating the weight adjustment amount of the sub-flow at the next moment according to the energy function value of the sub-flow at the current moment; The relationship between the system energy function and the proportional term weight and the differential term weight is predefined.

3. The dynamic coordinated multi-path routing control method according to claim 1, characterized in that: The step of determining the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment specifically includes: The sigmoid function is used to determine the weight coefficient of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment and the weight adjustment amounts of other subflows in the system at the current moment.

4. The dynamic coordinated multi-path routing control method according to claim 1, characterized in that: The weight adjustment amount of each sub-flow at the current moment for: ； Where, 、 and are the weight coefficients of the proportional term, integral term and differential term respectively, represents the first difference of the i-th sub-stream, represents the synergy coefficient between the i-th subflow and the j-th subflow in the system except for it, N represents the number of subflows in the system except for the i-th subflow, represents the first difference of the j-th sub-stream, represents the coupled differential suppression factor.

5. The dynamic coordinated multi-path routing control method according to any one of claims 1 to 4, characterized in that: The step of determining the multipath routing control scheme at the next moment based on the weight of each sub-flow at the next moment specifically includes: When the health score of the subflow is greater than a first preset threshold within a preset time period, the subflow is determined as an active subflow; When the health score of the sub-flow is less than a second preset threshold, determining the sub-flow as a disabled sub-flow; triggering a path switch for the deactivated sub-flow, and determining a sub-flow to be switched among the activated sub-flows based on the weight of each sub-flow at the next moment; The first preset threshold is greater than the second preset threshold.

6. The dynamic coordinated multi-path routing control method according to claim 5, characterized in that: The first preset threshold As shown below: ； The second preset threshold As shown below: ； Where, and are the mean and standard deviation of the health scores of all sub-flows in the system at time t, is the system bandwidth utilization, m is the dispersion sensitivity coefficient, n is the dispersion tolerance coefficient, is the bandwidth utilization tolerance coefficient, is the overload protection factor.

7. A dynamic collaborative multi-path routing control device, comprising: A determination module, configured to determine a health score of each subflow based on the performance indicators of the subflow at the current moment; an adjustment module, configured to determine a weight adjustment amount for each subflow at a current moment using a PID control technique based on a first difference between the health score of each subflow and an average health score of all subflows; The generating module is used to determine the weight of each subflow at the next moment based on the weight adjustment amount of each subflow at the current moment, and determine the multipath routing control scheme at the next moment based on the weight of each subflow at the next moment.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the dynamic collaborative multi-path routing control method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the dynamic coordinated multi-path routing control method according to any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the dynamic coordinated multi-path routing control method according to any one of claims 1 to 6 is implemented.

Images