Source returning method and device, storage medium and computer program product

Abstract

The embodiment of the invention provides a back-to-source method and device, a storage medium and a computer program product. The back-to-source method comprises the steps that multiple back-to-source link time delays, multiple back-to-source link packet loss rates and back-to-source bandwidth of each edge node sent by each edge node are received; obtaining the capacity of a plurality of back-to-source nodes; wherein the plurality of link delays and the plurality of packet loss rates are generated after each edge node performs dial test on the plurality of back-to-source nodes; according to the time delays of the plurality of back-to-source links, the packet loss rates of the plurality of back-to-source links, the back-to-source bandwidth of each edge node and the capacities of the plurality of back-to-source nodes, generating a back-to-source node decision result by using a particle swarm algorithm through a constructed target function and a corresponding constraint condition; the back-to-source decision result is used for deciding a target back-to-source node of each edge node for back-to-source; and issuing a node identifier of the corresponding target back-to-source node to each edge node according to the back-to-source node decision result.

Description

Technical Field

[0001] The present application relates to the field of networks, and in particular to a back-to-source method and device, a storage medium, and a computer program product. Background Art

[0002] A Content Delivery Network (CDN) is a new type of network content service system built on the Internet Protocol (IP) network. It provides content distribution and services based on the efficiency, quality, and order of content access and application. A CDN system can return required resources to clients as quickly as possible, and it can also satisfy the vast majority of client requests through internal caching, thereby protecting the origin server from being overwhelmed by excessive back-to-source requests. A common CDN system includes edge nodes for directly processing client requests, and back-to-source nodes for interacting with the origin server to access the origin server, obtain resources, and cache them. In scenarios where edge nodes return to the source through back-to-source nodes, the back-to-source node is typically selected based on the principle of low latency in the edge node's back-to-source link. However, when back-to-source nodes are selected based on low latency, multiple edge nodes may choose the same back-to-source node, leading to back-to-source node capacity overload and low accuracy in back-to-source node selection, which in turn reduces service quality. Summary of the Invention

[0003] The embodiments of the present application provide a back-to-source method and apparatus, a storage medium, and a computer program product, which can improve the accuracy of back-to-source node selection and enhance service quality.

[0004] The technical solution of this application is achieved as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for returning to the source, the method comprising:

[0006] Receive multiple back-to-source link delays, multiple back-to-source link packet loss rates, and back-to-source bandwidth of each edge node sent by each edge node; and obtain the capacity of multiple back-to-source nodes; wherein the multiple link delays and the multiple packet loss rates are generated after each edge node dials the multiple back-to-source nodes respectively;

[0007] Based on the multiple back-to-source link delays, the multiple back-to-source link packet loss rates, the back-to-source bandwidth of each edge node, and the capacity of the multiple back-to-source nodes, a back-to-source node decision result is generated using a constructed objective function and corresponding constraints and a particle swarm algorithm; the back-to-source decision result is used to determine the target back-to-source node for each edge node to back-source.

[0008] According to the back-to-origin node decision result, the node identifier of the corresponding target back-to-origin node is sent to each edge node.

[0009] In a second aspect, an embodiment of the present application provides a back-to-source device, comprising:

[0010] A receiving unit is configured to receive multiple back-to-source link delays, multiple back-to-source link packet loss rates, and the back-to-source bandwidth of each edge node sent by each edge node; and obtain the capacity of multiple back-to-source nodes; wherein the multiple link delays and the multiple packet loss rates are generated after each edge node dials the multiple back-to-source nodes for testing;

[0011] A decision unit, configured to generate a back-to-source node decision result using a particle swarm algorithm and a constructed objective function and corresponding constraints based on the multiple back-to-source link delays, the multiple back-to-source link packet loss rates, the back-to-source bandwidth of each edge node, and the capacity of the multiple back-to-source nodes; the back-to-source decision result is used to determine a target back-to-source node for each edge node to back-source.

[0012] A configuration sending unit is configured to send a node identifier of a corresponding target back-to-source node to each edge node according to the back-to-source node decision result.

[0013] In a third aspect, an embodiment of the present application proposes a back-to-source device, which includes: a processor, a memory, and a communication bus; the processor implements the above-mentioned back-to-source method when executing a running program stored in the memory.

[0014] In a fourth aspect, an embodiment of the present application provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned return-to-source method is implemented.

[0015] In a fifth aspect, an embodiment of the present application proposes a computer program product, including a computer program, which implements the above-mentioned back-to-source method when executed by a processor.

[0016] The embodiment of the present application provides a return-to-source method and device, storage medium, and computer program product, the method comprising: receiving multiple return-to-source link delays, multiple return-to-source link packet loss rates, and the return-to-source bandwidth of each edge node sent by each edge node; and obtaining the capacity of multiple return-to-source nodes; wherein the multiple link delays and multiple packet loss rates are generated after each edge node dials multiple return-to-source nodes respectively; based on the multiple return-to-source link delays, the multiple return-to-source link packet loss rates, the return-to-source bandwidth of each edge node, and the capacity of multiple return-to-source nodes, a return-to-source node decision result is generated by constructing an objective function and corresponding constraints and using a particle swarm algorithm; the return-to-source decision result is used to determine the target return-to-source node for each edge node to return to; based on the return-to-source node decision result, the node identifier of the corresponding target return-to-source node is sent to each edge node. Using the above implementation scheme, the return-to-source link delay, packet loss rate, and the capacity of the return-to-source node are considered at the same time, the return-to-source node selection problem is solved by the objective function and corresponding constraints, and the particle swarm algorithm is used to obtain the optimal solution, and an approximately globally optimal return-to-source node selection strategy is formulated, which improves the accuracy of the return-to-source node selection and improves the service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of a back-to-source method provided in an embodiment of the present application;

[0018] Figure 2 A structural diagram of an exemplary back-to-source device provided in an embodiment of the present application;

[0019] Figure 3 A schematic diagram of an exemplary particle swarm algorithm process provided in an embodiment of the present application;

[0020] Figure 4 A schematic diagram of an exemplary back-to-origin node selection strategy provided in an embodiment of the present application;

[0021] Figure 5 A schematic diagram of the structure of a return-to-source device provided in an embodiment of the present application Figure 1 ;

[0022] Figure 6 A schematic diagram of the structure of a return-to-source device provided in an embodiment of the present application Figure 2 . DETAILED DESCRIPTION

[0023] In order to enable a more detailed understanding of the features and technical contents of the embodiments of the present application, the implementation of the embodiments of the present application is described in detail below with reference to the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present application.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0025] In the following description, reference is made to "some embodiments," which describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. It should also be noted that the terms "first, second, and third" in the embodiments of the present application are only used to distinguish similar objects and do not represent a specific ordering of the objects. It is understood that "first, second, and third" may be interchanged in a specific order or sequential order where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0026] Currently, consistent hashing algorithms can be used to select origin nodes, ensuring that identical resource requests are routed to the same origin node. This can reduce the origin rate, thereby reducing origin server bandwidth, while also improving CDN node storage utilization. However, this method doesn't consider network quality when selecting origin nodes, which can lead to poor service quality, such as high first-packet latency and lag.

[0027] Currently, it is also possible to select the source node based on the dial-up latency, thereby ensuring that the latency of the source link of each edge node is minimized. This can reduce the latency of the source link, thereby reducing the first packet latency and improving the CDN distribution quality. However, this method does not consider the capacity of the source node, which can easily cause some source nodes to be overloaded, leading to large-scale poor service quality. It also does not consider the network link packet loss rate. Latency is only part of the network quality. In the actual distribution of services such as on-demand, live broadcast, and download, the network packet loss rate is also very important. Considering only the network latency will result in low accuracy in the selection of the source node and poor service quality.

[0028] This application embodiment provides a method for returning to the source, such as Figure 1 As shown, the method may include:

[0029] S101. Receive multiple back-to-source link delays, multiple back-to-source link packet loss rates, and the back-to-source bandwidth of each edge node sent by each edge node; and obtain the capacities of the multiple back-to-source nodes; wherein the multiple link delays and multiple packet loss rates are generated after each edge node dials multiple back-to-source nodes for testing.

[0030] The embodiment of the present application provides a method for returning to the source, which is applied to a device for returning to the source. Figure 2Both the origin node and the edge node are CDN nodes. The origin device includes a decision-making unit and a configuration delivery unit. The decision-making module is used to receive multiple origin link delays, multiple origin link packet loss rates, and each edge node's origin bandwidth from each edge node. Each edge node is equipped with a dialing test module, which uses the dialing test module to test multiple origin nodes and generate multiple origin link delays and multiple origin link packet loss rates.

[0031] In an embodiment of the present application, the edge node dials the back-to-source node a preset number of times in each decision cycle. The dialing content is the back-to-source link delay and the back-to-source link packet loss rate. When the decision cycle arrives, the edge node calculates the average back-to-source link delay and the average back-to-source link packet loss rate within the decision cycle, and then uploads the average back-to-source link delay and the average back-to-source link packet loss rate to the back-to-source device. It can be seen that each back-to-source link delay and back-to-source link packet loss rate received by the back-to-source device is the average back-to-source link delay and the average back-to-source link packet loss rate within the master-side cycle.

[0032] In an embodiment of the present application, the edge node can send multiple back-to-source link delays, multiple back-to-source link packet loss rates, and the back-to-source bandwidth of each edge node to the back-to-source device via the Hypertext Transfer Protocol (HTTP). The specific selection can be based on actual conditions and is not specifically limited in the embodiment of the present application.

[0033] S102. Based on the delays of multiple back-to-source links, the packet loss rates of multiple back-to-source links, the back-to-source bandwidth of each edge node, and the capacity of multiple back-to-source nodes, a back-to-source node decision result is generated using a constructed objective function and corresponding constraints and a particle swarm algorithm. The back-to-source decision result is used to determine the target back-to-source node for each edge node to back-source.

[0034] In an embodiment of the present application, after receiving multiple return-to-source link delays, multiple return-to-source link packet loss rates, and the return-to-source bandwidth of each edge node sent by each edge node, the decision unit of the return-to-source device generates a return-to-source node decision result based on the multiple return-to-source link delays, multiple return-to-source link packet loss rates, the return-to-source bandwidth of each edge node, and the capacity of multiple return-to-source nodes by constructing an objective function and corresponding constraints, and using a particle swarm algorithm.

[0035] In an embodiment of the present application, the back-to-source node decision problem is modeled as a large-scale integer programming problem based on the back-to-source link delay, the back-to-source link packet loss rate, the back-to-source bandwidth of the edge node, and the capacity of the back-to-source node. First, it is necessary to construct an objective function and corresponding constraints. Specifically, an objective function including the back-to-source link delay characteristic term and the back-to-source bandwidth characteristic term is constructed; a back-to-source node capacity constraint is constructed; a back-to-source link packet loss rate constraint is constructed; and a decision variable constraint is constructed to constrain each edge node to return to the source through one back-to-source node at the same time.

[0036] In the embodiment of the present application, the objective function can be referred to formula (1).

[0037]

[0038] Among them, i represents the number of the edge node, j represents the number of the back-to-source node, n represents a total of n edge nodes, m represents a total of m back-to-source nodes, flow i Indicates the return bandwidth of the i-th edge node, rtt i,j represents the return link delay from the i-th edge node to the j-th return source node, x i,j It is a 0-1 decision variable, indicating whether the i-th edge node is back-sourced through the j-th back-source node.

[0039] In the embodiment of the present application, the capacity constraint of the back-to-source node is shown in formula (2), which indicates that the capacity of each back-to-source node cannot exceed the limit.

[0040]

[0041] In the embodiment of the present application, the packet loss rate constraint for the back-to-source link is shown in Formula (3), which indicates that a network link with a packet loss rate less than a preset packet loss rate threshold, such as 0.1%, must be selected. It should be noted that the preset packet loss rate threshold can also have other values, and 0.1% is only one optional preset packet loss rate threshold proposed in this application. The specific value can be selected based on actual circumstances, and this embodiment of the present application does not impose any specific limitations.

[0042] loss i,j ≤0.1% (3)

[0043] In the embodiment of the present application, the decision variable constraints refer to formulas (4) and (5):

[0044]

[0045] x i,j ={0,1} (5)

[0046] In the embodiments of the present application, the large-scale 0-1 integer programming problem is an NP-hard (non-deterministic polynomial) problem. In practical applications, the search space for exact solution algorithms such as branch-and-bound is too large, and the solution efficiency is difficult to meet the requirements of practical applications in existing networks. In the embodiments of the present application, an approximate value is solved using a heuristic algorithm - the particle swarm algorithm.

[0047] In an embodiment of the present application, particle swarm parameters are initialized, including: number of iterations, particle position, speed, individual optimal fitness value, and global optimal fitness value; the fitness value of each particle is calculated using the first back-to-source link delay, first back-to-source packet loss rate, first back-to-source bandwidth, and first back-to-source node capacity corresponding to the initialized particle position through a constructed objective function and corresponding constraints; the fitness value is used to update the initialized individual optimal fitness value and the initialized global optimal fitness value; and the individual optimal fitness difference and the global optimal fitness difference before and after the update are determined; if the number of iterations, or the individual optimal fitness difference and the global optimal fitness difference meet the corresponding termination conditions, the updated global optimal fitness value is output as the back-to-source node decision result.

[0048] It should be noted that the particle swarm parameters also include particle swarm size, particle dimension, inertia weight, learning factor and iterative compensation range, etc., which can be selected and initialized according to actual conditions, and the embodiments of this application do not make specific limitations.

[0049] In an embodiment of the present application, the termination condition may include that the number of iterations reaches a maximum number of iterations, or the difference between the individual optimal fitness and the global optimal fitness between two iterations is less than a preset threshold.

[0050] Furthermore, if the number of iterations, the difference between the individual optimal fitness, and the global optimal fitness difference do not meet the corresponding termination conditions, the number of iterations is increased by one, and the speed and particle position are iteratively updated; the fitness value calculation process is executed using the second back-to-source link delay, the second back-to-source packet loss rate, the second back-to-source bandwidth, and the capacity of the second back-to-source node corresponding to the updated particle position, through the constructed objective function and the corresponding constraints; until the number of iterations, or the difference between the individual optimal fitness before and after the update, and the global optimal fitness difference meet the corresponding termination conditions, the updated global optimal fitness value is output as the back-to-source node decision result.

[0051] In an embodiment of the present application, the update speed is calculated using the first value obtained by subtracting the individual current position from the individual historical optimal position and the second value obtained by subtracting the global current position from the global historical optimal position. Then, the update speed is added to the current position of the particle to determine the position of the particle in the next round, that is, the updated particle position.

[0052] It should be noted that if it is detected that the updated global optimal fitness value does not meet the constraint conditions, the packet loss rate constraint of the back-to-source link is adjusted, and the global optimal fitness value as the back-to-source node decision result is re-searched until the updated global optimal fitness value meets the constraint conditions.

[0053] Based on the above embodiment, the process of particle swarm algorithm is shown, see Figure 3 .

[0054] 1. Initialize the particle swarm parameters, including particle swarm size, particle dimension, number of iterations, inertia weight, learning factor, and iteration step range.

[0055] 2. Initialize the position and velocity of each particle, as well as the individual historical optimal position, the global historical optimal position, the individual historical optimal fitness value, and the global historical optimal fitness value.

[0056] 3. Update the velocity and position of each particle.

[0057] 4. Calculate the fitness value of each particle.

[0058] 5. Update the individual historical optimal fitness value and individual historical optimal position of each particle.

[0059] 6. Update the global historical optimal fitness value and the global historical optimal position.

[0060] 7. Update other parameters, such as inertia weight, number of iterations, etc.

[0061] 8. Determine whether the end condition is met. The end condition is that the number of iterations reaches the maximum number, or the individual historical optimal fitness value and the global historical optimal fitness value between two iterations are less than the preset threshold.

[0062] 9. If the end condition is not met, execute 3.

[0063] 10. If the end condition is met, the last position of the global history is output.

[0064] S103: According to the back-to-origin node decision result, the node identifier of the corresponding target back-to-origin node is issued to each edge node.

[0065] In the embodiment of the present application, the configuration sending unit is responsible for sending the node identifier of the corresponding target back-to-source node to each edge node according to the back-to-source node decision result, and the edge node returns to the source through the corresponding target back-to-source node.

[0066] For example, the back-to-source link delay and the back-to-source link packet loss rate at a certain moment and the back-to-source bandwidth of the edge node are shown in Table 1.

[0067] Table 1

[0068]

[0069] The available capacity of back-to-source node 1 is 600 Gbps, and the available capacity of back-to-source node 2 is 220 Gbps.

[0070] Based on the back-to-source method proposed in the embodiment of this application, the final back-to-source node selection strategy is as follows: Figure 4 As shown, edge nodes 1, 2, 3, and 4 return to source node 1, while edge nodes 5 and 6 return to source node 2. Since this application not only considers the latency of the return link but also the capacity of the return node, although the latency of edge node 4 returning to source node 2 is lower than that of returning to source node 1, due to the limited capacity of return node 2, the decision unit sets the return strategy of edge node 1 to return to source node 1, avoiding overload of return node 2 and thus avoiding problems of poor service quality or failure.

[0071] It is understandable that by simultaneously considering the return link delay, packet loss rate and capacity of the return node, the return node selection problem is solved through the objective function and corresponding constraints using the particle swarm algorithm to find the optimal solution, and a return node selection strategy that is approximately globally optimal is formulated, which improves the accuracy of return node selection and enhances service quality.

[0072] Based on the above embodiments, the present application provides a return-to-source device 1. Figure 5 As shown, the device 1 includes:

[0073] The receiving unit 10 is configured to receive multiple back-to-source link delays, multiple back-to-source link packet loss rates, and the back-to-source bandwidth of each edge node sent by each edge node; and obtain the capacity of the multiple back-to-source nodes; wherein the multiple link delays and the multiple packet loss rates are generated after each edge node dials the multiple back-to-source nodes for testing;

[0074] A decision unit 11 is configured to generate a back-to-source node decision result using a particle swarm algorithm and a constructed objective function and corresponding constraints based on the multiple back-to-source link delays, the multiple back-to-source link packet loss rates, the back-to-source bandwidth of each edge node, and the capacity of the multiple back-to-source nodes; the back-to-source decision result is used to determine a target back-to-source node for each edge node to back-source.

[0075] The configuration sending unit 12 is configured to send the node identifier of the corresponding target back-to-source node to each edge node according to the back-to-source node decision result.

[0076] Optionally, the device further comprises: a construction unit;

[0077] The construction unit is further used to construct an objective function including a return-to-source link delay characteristic item and a return-to-source bandwidth characteristic item; construct a return-to-source node capacity constraint; construct a return-to-source link packet loss rate constraint; and construct a decision variable constraint to constrain each edge node to return to the source through one return-to-source node at the same time.

[0078] Optionally, the decision unit 11 is also used to initialize the particle swarm parameters, which include: number of iterations, particle position, speed, individual optimal fitness value and global optimal fitness value; using the first back-to-source link delay, first back-to-source packet loss rate, first back-to-source bandwidth and capacity of the first back-to-source node corresponding to the initialized particle position, calculate the fitness value of each particle through the constructed objective function and corresponding constraints; use the fitness value to update the initialized individual optimal fitness value and the initialized global optimal fitness value; and determine the individual optimal fitness difference and the global optimal fitness difference before and after the update; if the number of iterations, or the individual optimal fitness difference and the global optimal fitness difference meet the corresponding termination conditions, the updated global optimal fitness value is output as the back-to-source node decision result.

[0079] Optionally, the decision unit 11 is further used to increase the number of iterations by one and iteratively update the speed and particle position if the number of iterations, the individual optimal fitness difference, and the global optimal fitness difference do not meet the corresponding termination conditions; use the second back-to-source link delay, the second back-to-source packet loss rate, the second back-to-source bandwidth, and the capacity of the second back-to-source node corresponding to the updated particle position, and execute the fitness value calculation process through the constructed objective function and corresponding constraints; until the number of iterations, or the individual optimal fitness difference before and after the update, and the global optimal fitness difference meet the corresponding termination conditions, the updated global optimal fitness value is output as the back-to-source node decision result.

[0080] Optionally, the decision unit 11 is further configured to adjust the packet loss rate constraint of the back-to-source link and re-search the global optimal fitness value as the back-to-source node decision result if it is detected that the updated global optimal fitness value does not meet the constraint condition.

[0081] The embodiment of the present application provides a return-to-source device, which receives multiple return-to-source link delays, multiple return-to-source link packet loss rates, and the return-to-source bandwidth of each edge node sent by each edge node; and obtains the capacity of multiple return-to-source nodes; wherein the multiple link delays and multiple packet loss rates are generated after each edge node dials multiple return-to-source nodes respectively; based on the multiple return-to-source link delays, multiple return-to-source link packet loss rates, the return-to-source bandwidth of each edge node, and the capacity of multiple return-to-source nodes, a return-to-source node decision result is generated by constructing an objective function and corresponding constraints and using a particle swarm algorithm; the return-to-source decision result is used to decide the target return-to-source node for each edge node to return to; based on the return-to-source node decision result, the node identifier of the corresponding target return-to-source node is sent to each edge node. It can be seen that the return-to-source device proposed in this embodiment takes into account the return-to-source link delay, packet loss rate, and the capacity of the return-to-source node at the same time, solves the return-to-source node selection problem through the objective function and corresponding constraints, and uses the particle swarm algorithm to obtain the optimal solution, formulates an approximate global optimal return-to-source node selection strategy, improves the accuracy of return-to-source node selection, and improves service quality.

[0082] Figure 6 A schematic diagram of the structure of a return-to-source device 1 provided in an embodiment of the present application Figure 2 In practical applications, based on the same disclosure concept of the above embodiments, Figure 6 As shown, the back-to-source device 1 of this embodiment includes: a processor 13 , a memory 14 and a communication bus 15 .

[0083] In a specific embodiment, the processor 13 may be at least one of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a CPU, a controller, a microcontroller, and a microprocessor. It is understood that for different devices, the electronic device used to implement the functions of the processor may also be other, and this embodiment does not specifically limit this.

[0084] In the embodiment of the present application, the communication bus 15 is used to realize the connection and communication between the processor 13 and the memory 14; when the processor 13 executes the running program stored in the memory 14, the following back-to-source method is implemented:

[0085] Receive multiple return-to-source link delays, multiple return-to-source link packet loss rates, and the return-to-source bandwidth of each edge node sent by each edge node; and obtain the capacity of multiple return-to-source nodes; wherein the multiple link delays and the multiple packet loss rates are generated after each edge node dials the multiple return-to-source nodes respectively; based on the multiple return-to-source link delays, the multiple return-to-source link packet loss rates, the return-to-source bandwidth of each edge node, and the capacity of the multiple return-to-source nodes, generate a return-to-source node decision result by constructing an objective function and corresponding constraints, and using a particle swarm algorithm; the return-to-source decision result is used to determine the target return-to-source node for each edge node to return to; based on the return-to-source node decision result, send the node identifier of the corresponding target return-to-source node to each edge node.

[0086] Furthermore, the processor 13 is further configured to construct an objective function including a back-to-source link delay characteristic item and a back-to-source bandwidth characteristic item; construct a back-to-source node capacity constraint; construct a back-to-source link packet loss rate constraint; and construct a decision variable constraint to constrain each edge node to simultaneously return to the source through one back-to-source node.

[0087] Furthermore, the link from the edge node to the source node is used as a particle.

[0088] Furthermore, the processor 13 is further configured to initialize particle swarm parameters, including the number of iterations, particle position, speed, individual optimal fitness value, and global optimal fitness value; calculate the fitness value of each particle using the first back-to-source link delay, first back-to-source packet loss rate, first back-to-source bandwidth, and first back-to-source node capacity corresponding to the initialized particle position, through a constructed objective function and corresponding constraints; use the fitness value to update the initialized individual optimal fitness value and the initialized global optimal fitness value; and determine the individual optimal fitness difference and the global optimal fitness difference before and after the update; if the number of iterations, or the individual optimal fitness difference and the global optimal fitness difference meet the corresponding termination conditions, output the updated global optimal fitness value as the back-to-source node decision result.

[0089] Furthermore, the processor 13 is further configured to, if the number of iterations, the individual optimal fitness difference, and the global optimal fitness difference do not satisfy corresponding termination conditions, increase the number of iterations by one, and iteratively update the speed and particle position; use the second back-to-source link delay, the second back-to-source packet loss rate, the second back-to-source bandwidth, and the capacity of the second back-to-source node corresponding to the updated particle position, and execute a fitness value calculation process through the constructed objective function and corresponding constraints; until the number of iterations, or the individual optimal fitness difference before and after the update, and the global optimal fitness difference satisfy corresponding termination conditions, output the updated global optimal fitness value as the back-to-source node decision result.

[0090] Furthermore, the processor 13 is further configured to adjust the packet loss rate constraint of the back-to-source link and re-search for the global optimal fitness value as the back-to-source node decision result if it is detected that the updated global optimal fitness value does not satisfy the constraint condition.

[0091] An embodiment of the present application provides a storage medium having a computer program stored thereon. The computer-readable storage medium stores one or more programs. The one or more programs can be executed by one or more processors and applied to a back-to-source device. The computer program implements the back-to-source method described above.

[0092] Based on the above embodiments, embodiments of the present application provide a computer program product, including a computer program. The computer program can be executed by one or more processors, and the computer program implements the above-mentioned back-to-source method.

[0093] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0094] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present disclosure, or the part that contributes to the relevant technology, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling an image display device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present disclosure.

[0095] The above description is merely a preferred embodiment of the present application and is not intended to limit the scope of protection of the present application.

Claims

1. A back-to-source method, characterized in that: The method comprises: Receive multiple back-to-source link delays, multiple back-to-source link packet loss rates, and back-to-source bandwidth of each edge node sent by each edge node; and obtain the capacity of multiple back-to-source nodes; wherein the multiple link delays and the multiple packet loss rates are generated after each edge node dials the multiple back-to-source nodes respectively; Based on the multiple back-to-source link delays, the multiple back-to-source link packet loss rates, the back-to-source bandwidth of each edge node, and the capacity of the multiple back-to-source nodes, a back-to-source node decision result is generated using a constructed objective function and corresponding constraints and a particle swarm algorithm; the back-to-source decision result is used to determine the target back-to-source node for each edge node to back-source. According to the back-to-origin node decision result, the node identifier of the corresponding target back-to-origin node is sent to each edge node.

2. The method according to claim 1, characterized in that Before generating a back-to-source node decision result based on the multiple back-to-source link delays, the multiple back-to-source link packet loss rates, the back-to-source bandwidth of each edge node, and the capacity of the multiple back-to-source nodes by constructing an objective function and corresponding constraints and using a particle swarm algorithm, the method further includes: Construct an objective function that includes the characteristic terms of the return link delay and the return bandwidth; Establish capacity constraints for back-to-origin nodes; Establishing packet loss rate constraints on the return link; Construct decision variable constraints to constrain each edge node to return to the source through a back-to-source node at the same time.

3. The method according to claim 1, characterized in that Taking the links from edge nodes to source nodes as particles, generating a source node decision result based on the multiple source link delays, the multiple source link packet loss rates, the source bandwidth of each edge node, and the capacity of the multiple source nodes by constructing an objective function and corresponding constraints and using a particle swarm algorithm, includes: Initialize the particle swarm parameters, which include: number of iterations, particle position, speed, individual optimal fitness value and global optimal fitness value; The fitness value of each particle is calculated using the first return-to-source link delay, first return-to-source packet loss rate, first return-to-source bandwidth, and first return-to-source node capacity corresponding to the initialization particle position, through the constructed objective function and corresponding constraints. Using the fitness value to update the initialized individual optimal fitness value and the initialized global optimal fitness value; and determining the individual optimal fitness difference and the global optimal fitness difference before and after the update; If the number of iterations, or the individual optimal fitness difference and the global optimal fitness difference meet corresponding termination conditions, the updated global optimal fitness value is output as the back-to-source node decision result.

4. The method according to claim 3, characterized in that After determining the individual optimal fitness difference before and after the update and the global optimal fitness difference, the method further includes: If the number of iterations, the individual optimal fitness difference, and the global optimal fitness difference do not meet the corresponding termination conditions, the number of iterations is increased by one, and the speed and particle position are iteratively updated; The fitness value calculation process is performed using the second back-to-source link delay, second back-to-source packet loss rate, second back-to-source bandwidth, and second back-to-source node capacity corresponding to the updated particle position, through the constructed objective function and corresponding constraints; Until the number of iterations, or the difference between the individual optimal fitness before and after the update and the global optimal fitness difference meet the corresponding termination conditions, the updated global optimal fitness value is output as the decision result of the back-to-source node.

5. The method according to claim 3 or 4, characterized in that After outputting the updated global optimal fitness value as the back-to-source node decision result, the method further includes: If it is detected that the updated global optimal fitness value does not meet the constraint condition, the packet loss rate constraint of the back-to-source link is adjusted, and the global optimal fitness value as the decision result of the back-to-source node is searched again.

6. A return-to-source device, characterized in that: The device comprises: A receiving unit is configured to receive multiple back-to-source link delays, multiple back-to-source link packet loss rates, and the back-to-source bandwidth of each edge node sent by each edge node; and obtain the capacity of multiple back-to-source nodes; wherein the multiple link delays and the multiple packet loss rates are generated after each edge node dials the multiple back-to-source nodes for testing; A decision unit, configured to generate a back-to-source node decision result using a particle swarm algorithm and a constructed objective function and corresponding constraints based on the multiple back-to-source link delays, the multiple back-to-source link packet loss rates, the back-to-source bandwidth of each edge node, and the capacity of the multiple back-to-source nodes; the back-to-source decision result is used to determine a target back-to-source node for each edge node to back-source. A configuration sending unit is configured to send a node identifier of a corresponding target back-to-source node to each edge node according to the back-to-source node decision result.

7. The device according to claim 6, characterized in that The device further comprises: a construction unit; The construction unit is further used to construct an objective function including a return-to-source link delay characteristic item and a return-to-source bandwidth characteristic item; construct a return-to-source node capacity constraint; construct a return-to-source link packet loss rate constraint; and construct a decision variable constraint to constrain each edge node to return to the source through one return-to-source node at the same time.

8. The device according to claim 6, characterized in that The decision unit is further configured to initialize particle swarm parameters, including the number of iterations, particle position, speed, individual optimal fitness value, and global optimal fitness value; calculate the fitness value of each particle using the first back-to-source link delay, first back-to-source packet loss rate, first back-to-source bandwidth, and first back-to-source node capacity corresponding to the initialized particle position, through a constructed objective function and corresponding constraints; use the fitness value to update the initialized individual optimal fitness value and the initialized global optimal fitness value; and determine the individual optimal fitness difference and the global optimal fitness difference before and after the update; if the number of iterations, or the individual optimal fitness difference and the global optimal fitness difference meet the corresponding termination conditions, output the updated global optimal fitness value as the back-to-source node decision result.

9. A return-to-source device, characterized in that: The device includes: a processor, a memory, and a communication bus; when the processor executes the running program stored in the memory, the method according to any one of claims 1 to 5 is implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

11. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the computer program implements the method according to any one of claims 1 to 5.

Images