A Method, Device and Storage Medium for Concurrent Access Link Planning in an SDP System

Abstract

The present invention discloses a method, device, and storage medium for concurrent access link planning in an SDP system. The method comprises: S1, obtaining information about communication devices and SDP access gateway devices on a resource end in an access network, numbering the communication devices and SDP access gateway devices, and constructing a communication topology diagram between the SDP access gateway and the resource end based on the numbered devices; S2, generating a population comprising a plurality of link individuals based on the communication topology diagram, wherein a single link individual represents an access link, and the access node value range of the access link is the device number range; S3, optimizing the population using the access time of the link individual as an objective function; and S4, selecting a plurality of link individuals within the population with small objective function values as concurrent access links. The present invention can simultaneously plan multiple different optimal and suboptimal access links for user end access to the resource end, thereby improving the efficiency of concurrent access processing.

Description

Technical Field

[0001] The present invention relates to the field of SDP zero trust, and in particular to a method, device and storage medium for concurrent access link planning of an SDP system. Background Art

[0002] Under traditional security systems, enterprises employ a traditional network segmentation and isolation security model, using perimeter protection devices such as firewalls to demarcate the internal and external networks. Based on the perception that "internal networks are trustworthy, external networks are untrustworthy," this leads to over-trust in internal networks. Once the internal network lacks control, attacks launched from within are defenseless. Furthermore, external attacks or Trojans, once able to successfully penetrate the perimeter protection gate and enter the internal network, can conduct various illegal operations and steal information. Furthermore, with the advancement of the national digital transformation strategy, distributed offices such as remote work and off-site branches have proliferated, blurring the physical boundaries of networks. Traditional security architectures based on internal and external network boundaries are no longer able to meet the demands of digital transformation.

[0003] SDP (Software Defined Perimeter) is a new generation of network security technology architecture based on the concept of zero-trust security. It uses identity as the cornerstone, business security access, continuous trust assessment and dynamic access control as four key capabilities to build a zero-trust security system including SDP access gateway, SDP controller, access network and resource-side SDP access gateway to isolate all users from enterprise resources. Figure 1 As shown, users can access the access network only after passing identity authentication and accessing the SDP access gateway. Then, they can securely access specific resources in the resource center through the access network and SDP access gateway.

[0004] In order to enable users to access the enterprise resource center with high speed and stability in a remote network environment, it is particularly important to plan the access link from the user, namely the SDP access gateway, to the enterprise resource center in the SDP system. In the existing technology, an optimal access link from the SDP access gateway to the enterprise resource center is usually planned. However, in the case of high concurrent access, only one or a small number of similar access links are planned, which will cause access congestion and other problems. Planning multiple optimal or suboptimal access links is conducive to improving the efficiency of concurrent access processing. Summary of the Invention

[0005] The purpose of the present invention is to solve the above-mentioned problems in the prior art and provide a method, device and storage medium for concurrent access link planning in an SDP system. When high concurrent access requests occur, multiple different optimal and suboptimal access links can be planned simultaneously to facilitate user end access to resource end, thereby improving the efficiency of concurrent access processing.

[0006] In a first aspect, the present invention provides a method for concurrent access link planning of an SDP system, the method comprising:

[0007] S1, obtain information about communication devices and SDP access gateway devices on the access network, number the communication devices and SDP access gateway devices, and build a communication topology diagram between the SDP access gateway and the resource end based on the numbered devices;

[0008] S2, generating a population including a plurality of link individuals based on the communication topology graph, wherein a single link individual represents a feasible access link between the SDP admission gateway and the resource end, and the access link is formed by sequentially connecting a plurality of access nodes, and the access node value range is a device number range;

[0009] S3, takes the access time of individual links as the objective function to optimize the population;

[0010] S4, select several different link individuals with small objective function values in the population as concurrent access links.

[0011] Preferably, the step of optimizing the population in S3 includes:

[0012] S31, setting the similarity threshold of link individuals within the population;

[0013] S32, iteratively performing full-member optimization on the population until the number of full-member optimizations on the population reaches a first iteration threshold;

[0014] In a single full-member optimization of a population, each link individual in the population is selected for optimization. The steps for optimizing the selected single link individual include:

[0015] A1, other link individuals whose similarity with the selected link individual is not less than its similarity threshold are regarded as neighbors of the selected link individual;

[0016] A2: Determine whether the selected link individual has neighbors. If not, go to A3; if yes, go to A4.

[0017] A3, generate a virtual link individual whose similarity to the selected link individual is not less than its similarity threshold as the source link individual, and go to A5;

[0018] A4, randomly select a link individual from the neighbors of the selected link individual as the source link individual, and go to A5;

[0019] A5, using the source link individual to generate a variant link individual, and using the variant link individual to perform cross-mutation on the selected link individual to obtain an experimental link individual;

[0020] A6, judging whether the objective function value of the selected link individual is less than the objective function value of the experimental link individual, if so, go to A8, if not, go to A7;

[0021] A7, updating the selected link individual with the experimental link individual;

[0022] A8, the selected link individual remains unchanged.

[0023] Preferably,

[0024] The S31 further includes: setting the age of the link individuals in the population;

[0025] Said A7 further comprises: resetting the individual age of the selected link and going to A9;

[0026] Said A8 further comprises: increasing the individual age of the selected link, and going to A9;

[0027] The A9 includes: judging whether the age of the selected link individual is less than an age threshold, if so, the selected link individual remains unchanged, if not, performing secondary optimization on the selected link individual.

[0028] Preferably, the step of performing secondary optimization on the selected individual links includes:

[0029] Setting a temporary similarity threshold for the selected link individual, wherein the temporary similarity threshold is greater than the similarity threshold initially set;

[0030] Generate a virtual link individual whose similarity to the selected link individual is not less than its temporary similarity threshold as the source link individual;

[0031] Using the source link individual to generate a variant link individual, and using the variant link individual to perform cross-mutation on the selected link individual to obtain an experimental link individual;

[0032] Determine whether the objective function value of the selected link individual is less than the objective function value of the experimental link individual. If so, the selected link individual remains unchanged; if not, the selected link individual is updated with the experimental link individual and the age of the selected link individual is reset.

[0033] Preferably, the S3 further includes:

[0034] S33, setting a similarity clustering threshold, dividing the link individuals in the population into several sub-populations by similarity clustering, and the similarity between the link individual with the smallest objective function value and the remaining link individuals in the sub-population is not less than the similarity clustering threshold;

[0035] S34, setting the population capacity of the subpopulation. For subpopulations whose number of link individuals is greater than the population capacity, sort the link individuals in the subpopulation in ascending order according to the objective function value. The link individuals with the lowest order that exceed the population capacity die.

[0036] S35, obtaining the total number of dead individuals in all subpopulations, regenerating a number of link individuals according to the total number of dead individuals, and ensuring that the similarity between the regenerated link individuals and the link individual with the smallest objective function value in any subpopulation is less than a similarity clustering threshold;

[0037] S36, reset the similarity threshold and age for the reborn link individuals;

[0038] S37, iteratively perform full optimization on the regenerated link individuals until the number of full optimizations performed on the regenerated link individuals reaches a second iteration threshold.

[0039] Preferably, in a single full-staff optimization of reborn link individuals, each reborn link individual is selected for optimization separately; after the selected single reborn link individual is optimized, an exclusion mechanism is added, and the exclusion mechanism includes: judging whether the similarity between the selected reborn link individual and the link individual with the smallest objective function value in any sub-population is less than the similarity clustering threshold; if so, the selected reborn link individual remains unchanged; if not, the selected reborn link individual is allowed to die and be reborn again, and the similarity between the reborn link individual and the link individual with the smallest objective function value in any sub-population is less than the similarity clustering threshold.

[0040] Preferably, the S3 further includes:

[0041] S38, cancel the exclusion mechanism, and perform full optimization on the regenerated link individual iterations until the number of iterations reaches the third iteration threshold.

[0042] In a second aspect, the present invention provides an SDP system concurrent access link planning device, comprising a memory and a computer program stored thereon, and a processor, wherein the processor implements the above-mentioned SDP system concurrent access link planning method when executing the computer program.

[0043] In a third aspect, the present invention provides a storage medium storing computer-executable instructions. When the computer-executable instructions are loaded and executed by a processor, the above-mentioned SDP system concurrent access link planning method is implemented.

[0044] In summary, the present invention has the following beneficial effects:

[0045] 1. Set each link individual to select or generate a source link individual within a similarity threshold, use the source link individual to generate a variant link individual, and then cross-mutate with the corresponding link individual to generate an experimental link individual. If the experimental link individual is better than the corresponding link individual, the experimental link individual is used to update the link individual. Using this local mutation strategy, the link individuals in the population are repeatedly iterated to optimize, so that the link individuals in the population can change into multiple different optimal and suboptimal access links.

[0046] 2. Set an age for the link individuals to increase the probability of link individual changes during population iteration, avoid link individuals stopping changing before changing to the optimal or suboptimal access link, and increase the probability of link individuals changing to the optimal or suboptimal access link.

[0047] 3. By setting similarity clustering and population capacity, a portion of poor link individuals are selected to die and reborn, and an exclusion mechanism is added for iterative optimization to avoid the link individuals in the population being too similar and to increase the degree of difference between the link individuals to find more different optimal and suboptimal access links.

[0048] When a high concurrent access request occurs, the present invention can simultaneously plan multiple different optimal and suboptimal access links to facilitate the user end to access the resource end, thereby improving the concurrent access processing efficiency.

[0049] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only 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.

[0051] Figure 1 This is a diagram of the existing SDP system architecture.

[0052] Figure 2 The figure is a flow chart of a method according to a specific embodiment of the present invention.

[0053] Figure 3 The figure is a schematic diagram of a communication topology diagram according to a specific embodiment of the present invention.

[0054] Figure 4 The figure is a flow chart of optimizing a population according to a specific embodiment of the present invention.

[0055] Figure 5 The figure is a flow chart of optimizing a single link entity according to a specific embodiment of the present invention.

[0056] Figure 6 The figure is a flowchart of optimizing a single regenerated link entity by adding an exclusion mechanism according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages disclosed in the embodiments of the present invention clearer, the embodiments of the present invention are further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present invention and are not intended to limit the embodiments of the present invention. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. Examples of the embodiments are shown in the accompanying drawings, where the same or similar numbers throughout represent the same or similar elements or elements with the same or similar functions.

[0058] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or are inherent to these processes, methods, products or devices.

[0059] like Figure 1 As shown, the existing SDP system usually includes an SDP access gateway, an access network, an SDP controller, and an SDP access gateway. A local user center is configured with an SDP access gateway, which is a unified access point for users in the user center. A resource center is configured with several SDP access gateways, which are secure access points for the resource center. When a user at a certain user center generates an access request, the user terminal passes the SDP access gateway authentication, and then goes online through the corresponding SDP access gateway to initiate a knock-on authentication to the SDP controller - single packet authentication, namely SPA authentication. After receiving the SPA authentication request from the user terminal, the SDP controller performs identity authentication and authentication. After authentication, the SDP controller intelligently and dynamically plans the access link, and informs the user terminal of the access link including the SDP access gateway to be connected through an encrypted channel. At the same time, after performing SPA authentication with the corresponding SDP access gateway, it informs the information of the access user terminal; after receiving the response from the SDP controller, the user terminal initiates a connection request to the corresponding SDP access gateway through the access network according to the access link information, and at the same time initiates an SPA identity authentication request again; the SDP access gateway receives the user terminal connection request and verifies the user terminal information. After verification, the user terminal establishes a secure connection with the SDP access gateway, and then accesses the resource center resources. The SDP access gateway controls the connection between the user terminal and specific resources of the resource center based on the control policy information.

[0060] In order to enable users to access the resource center with high speed and stability in a remote network environment, it is particularly important to plan the access link from the user terminal, namely the SDP access gateway, to the resource center. In the existing technology, an optimal access link from the SDP access gateway to the resource center is usually planned. However, in the case of high concurrent access, only one or a small number of similar access links are planned, which will cause access congestion and other problems.

[0061] To solve the above problems, the present application provides a concurrent access link planning method for an SDP system. When high concurrent access requests occur, multiple different optimal and suboptimal access links can be planned simultaneously to facilitate user end access to resource end, thereby improving concurrent access processing efficiency.

[0062] like Figure 2 As shown, the method of the present application includes step S1, obtaining information of the communication equipment and the SDP access gateway device of the resource end in the access network, numbering the communication equipment and the SDP access gateway device, and constructing a communication topology diagram between the SDP access gateway and the resource end based on the numbering device.

[0063] In this step, for the communication devices in the access network, obtain their quantity, unique identification, and directly connectable devices and other information. The directly connectable device information of a communication device includes the information of the directly connectable SDP access gateway, the directly connectable other communication devices, and the directly connectable SDP access gateway device. The connection relationship between the SDP access gateway, all communication devices, and the SDP access gateway device can be determined through the directly connectable device information of each communication device; for the resource-side SDP access gateway device, obtain its quantity, unique identification and other information.

[0064] Usually, the unique identification of communication equipment and SDP access gateway equipment is an ID address or MAC code. In order to simplify the device identification and thus simplify the communication topology, they are simply and uniquely numbered based on the number M of communication equipment and SDP access gateway devices. For example, they can be numbered 1, 2,..., M, and then a communication topology can be constructed including M numbered devices and the connection relationships between them, and their connection relationships with the SDP access gateway.

[0065] like Figure 2 As shown, the method of the present application also includes step S2, which generates a population including several link individuals based on the communication topology diagram. A single link individual represents a feasible access link between an SDP access gateway and a resource end. The access link is composed of several access nodes connected in sequence, and the value range of the access node is the device number range.

[0066] In this step, the number of link individuals in the population can be expressed as Np, and the j|j∈(1,2,……,Np)th link individual can be expressed as , Represents a link individual The yth visited node, The value range is the device number range, through The value of the link individual The specific device selected by the yth access node; Represents a link individual The number of visited nodes.

[0067] Combined with step S1, in order to conveniently distinguish the communication devices that can be directly connected to the SDP admission gateway and the SDP access gateway devices, that is, the first access node and the last access node of the link individual, they are numbered B1, B2, ..., BM1 based on the number M1 of communication devices that can be directly connected to the SDP admission gateway, and C1, C2, ..., CM2 based on the number M2 of the remaining communication devices, and D1, D2, ..., DM3 based on the number M3 of SDP access gateway devices. The resulting communication topology is as follows Figure 3 As shown. Correspondingly, in step S2, The specific value range is: link individual First access node , the last visited node .

[0068] In this step, Np link individuals can be randomly generated to form a population, or they can be generated using the following method: randomly select a communication device that can be directly connected to the SDP access gateway as the first access node, and randomly select an SDP access gateway as the last access node, and then generate a better link individual through the depth-first method; repeat the above steps Np times to generate Np link individuals to form a population.

[0069] like Figure 2 As shown, the method of the present application further includes step S3, which optimizes the population by taking the access time of the link individuals as the objective function.

[0070] In this step, the objective function can be expressed as , Indicates the transmission time of access request between devices, which is usually related to communication performance and can be regarded as fixed. The access request needs to be forwarded from the user end to the SDP access gateway, and then through the access node in the link individual, including the communication device and the SDP access gateway device, to reach the resource end. Therefore, the link individual The total transmission time is ; Represents a link individual The forwarding time required for the yth access node to receive an access request and send an access request is related to the service status of the access node and can be obtained by summing the time required for the access node to currently process services and wait for services to be processed.

[0071] like Figure 4 As shown, the step of optimizing the population in this step specifically includes S31, setting a similarity threshold of link individuals in the population; S32, iteratively optimizing the population until the number of times the population is optimized reaches a first iteration threshold.

[0072] In step S31, a unified and reasonable similarity threshold can be set for each link individual as needed. In some embodiments of the present application, the similarity threshold can be set to the average similarity of the link individuals in the population, which can be expressed as . Represents the link individuals in the population and The similarity between and The number of identically visited nodes is used as a measure, e.g. , ,in ,but and The similarity is 3.

[0073] In step S32, in a single full-member optimization of the population, each link individual in the population is selected for optimization, such as Figure 5 As shown, the steps of optimizing the selected single link individual include:

[0074] A1, other link individuals whose similarity with the selected link individual is not less than its similarity threshold are regarded as neighbors of the selected link individual;

[0075] A2: Determine whether the selected link individual has neighbors. If not, go to A3; if yes, go to A4.

[0076] A3, generate a virtual link individual whose similarity to the selected link individual is not less than its similarity threshold as the source link individual, and go to A5;

[0077] A4, randomly select a link individual from the neighbors of the selected link individual as the source link individual, and go to A5;

[0078] A5, using the source link individual to generate a variant link individual, and using the variant link individual to perform cross-mutation on the selected link individual to obtain an experimental link individual.

[0079] In step A5, the method of using source link individuals to generate variant link individuals can adopt the mutation method commonly used in existing genetic algorithm-based path planning. In some embodiments of the present application, source link individuals are used to generate variant link individuals, that is, an access node of the source link individual is randomly selected for mutation to obtain a variant link individual.

[0080] Specifically, if the first access node of the source link individual, i.e., the first access node, is selected for mutation, then based on the communication topology diagram, it is determined whether there are other numbered devices outside the source link individual that can be substituted as the first access node and directly connected to at least one access node on the source link individual except the first access node; if so, the first access node of the source link individual is replaced with a random numbered device in the set of replaceable numbered devices, and an access node that can be directly connected to the new first access node is selected on the source link individual as the boundary point, and then the access nodes between the new first access node and the boundary point are deleted, and the source link individual is updated as the mutated link individual; if not, the source link individual is directly used as the mutated link individual.

[0081] For example, The corresponding source link individual is represented as Based on the communication topology graph, it is determined that there is another numbered device B6 that can replace the first access node and directly connect to the second, third, and fifth access nodes of the source link individual, that is, directly connect to C3, C4, and C8. In addition, based on the communication topology graph, it is determined that there is another numbered device B2 that can replace the first access node and directly connect to the second, fourth, and fifth access nodes of the source link individual, that is, directly connect to C3, C6, and C8. Then, select one from the set of replaceable numbered devices (B2, B6). If B2 is selected to replace the first access node of the source link individual, and one of the access nodes that can be directly connected to the new first access node, that is, B2, is selected on the source link individual as the boundary point, if the fourth access node of the source link individual is selected as the boundary point, then the access nodes between the new first access node and the boundary point, that is, the fourth access node, on the source link individual are deleted, and the resulting mutated link individual is .

[0082] Similarly, if the last access node, i.e., the tail access node, of the source link individual is selected for mutation, then based on the communication topology diagram, it is determined whether there are other numbered devices outside the source link individual that can be substituted as the last access node and directly connected to at least one access node other than the last access node on the source link individual; if so, the last access node of the source link individual is replaced with a random numbered device from the set of replaceable numbered devices, and an access node that can be directly connected to the new last access node is selected on the source link individual as the boundary point, and then the access nodes between the boundary point and the new last access node are deleted, and the source link individual is updated as the mutated link individual; if not, the source link individual is directly used as the mutated link individual.

[0083] Likewise Corresponding source link individual For example, based on the communication topology graph, it is determined that only D4 can be substituted as the last access node and directly connected to the third to last and sixth to last access nodes of the source link individual. Then D4 is selected to replace it as the last access node of the source link individual, and an access node that can be directly connected to the new last access node, D4, is selected on the source link individual, such as the sixth to last access node, as the boundary point. The access nodes between the boundary point on the source link individual and the new last access node are deleted, and the resulting mutated link individual is .

[0084] If an access node other than the first and last access nodes on the source link individual is selected for mutation, all access nodes before the selected access node on the source link individual are taken as the front node set, and all access nodes after the selected access node are taken as the back node set; then, based on the communication topology diagram, it is determined whether there are other numbered devices outside the source link individual, which can be substituted as the selected access node, directly connected to at least one access node in the front node set and directly connected to at least one access node in the back node set; if so, the access node selected by the source link individual is replaced with a random numbered device in the set of replaceable numbered devices, and an access node that can be directly connected to the new selected node is selected from the front node set as the front boundary point, and an access node that can be directly connected to the new selected node is selected from the back node set as the back boundary point, and then the access nodes between the front boundary point and the selected node and between the selected node and the back boundary point on the source link individual are deleted, and the source link individual is updated as the mutated link individual; if not, the source link individual is directly used as the mutated link individual.

[0085] Likewise Corresponding source link individual For example, if the second access node is selected for mutation, the front node set is (B1) and the back node set is (C4, C6, C8, ..., C78, C85, C96, C97, C103, D1). Based on the communication topology, it is determined that only the numbered device C5 can be substituted as the second access node, which is directly connected to the first access node B1 and can only be directly connected to the back access node C8. Then the second access node of the source link individual is replaced with C5, and B1 is selected as the front boundary point and C8 as the back boundary point. The final mutated link individual is .

[0086] In step A5, the method of using mutated link individuals to cross-mutate the selected link individuals can adopt the cross-method commonly used in existing genetic algorithm-based path planning. In some embodiments of the present application, the method of using mutated link individuals to cross-mutate the selected link individuals to generate experimental link individuals is: if the mutated link individual and the selected link individual have the same access node, then select an access node that the mutated link individual and the selected link individual have as the intersection point, replace all access nodes after the intersection point of the selected link individual with all access nodes after the intersection point of the mutated link individual, and replace all access nodes after the intersection point of the mutated link individual with all access nodes after the intersection point of the selected link individual, and replace all access nodes after the intersection point of the mutated link individual with all access nodes after the intersection point of the selected link individual, to obtain two cross-over individuals, and select one of the cross-over individuals as the experimental link individual. If the mutated link individual and the selected link individual do not have the same access nodes, then find the shortest connection route between the mutated link individual and the selected link individual based on the communication topology graph, select one of the shortest connection routes as the cross-link, connect all the access nodes before the intersection of the selected link individual and the cross-link with all the access nodes after the intersection of the mutated link individual and the cross-link through the cross-link to form a post-cross individual, connect all the access nodes before the intersection of the mutated link individual and the cross-link with all the access nodes after the intersection of the selected link individual and the cross-link through the cross-link to form another post-cross individual, and select one of the post-cross individuals as the experimental link individual.

[0087] For example, suppose the selected link individual is , the corresponding mutation link individual is , there are the same access nodes C7 and C78, then randomly select one of them, such as C7 as the intersection point, and get two individuals after the intersection 、 , randomly select one of them as the experimental link individual, such as the experimental link individual is If the selected link individual is , the mutation link individual is , which does not have the same access node. The shortest connection routes include direct connection between C2 and C5, direct connection between C4 and C5, and direct connection between C78 and C106. Then one of the direct connection routes such as C4 and C5 is randomly selected as the cross connection, and the two individuals after the cross are 、 , randomly select one of them as the experimental link individual, such as the experimental link individual is .

[0088] like Figure 5 As shown, the step of optimizing the selected single link individual further includes A6, judging whether the objective function value of the selected link individual is less than the objective function value of the experimental link individual, if so, going to A8, if not, going to A7;

[0089] A7, update the selected link individual with the experimental link individual, that is, ; A8, the selected link individual remains unchanged.

[0090] In the above step S3, each link individual is set to select or generate a source link individual within a similarity threshold, and the source link individual is used to generate a variant link individual, which is then cross-mutated with the corresponding link individual to generate an experimental link individual. If the experimental link individual is better than the corresponding link individual, the link individual is updated with the experimental link individual. By using such a local mutation strategy, the link individuals in the population are repeatedly iterated and optimized, so that the link individuals in the population can be changed into multiple different optimal and suboptimal access links.

[0091] Furthermore, in order to increase the probability of link individual changes during population iteration, avoid link individual changes before changing to the optimal or suboptimal access link, and increase the probability of link individual changes to the optimal or suboptimal access link, in some embodiments of the present application, an age is set for link individuals in the population. When the link individual changes, the age is reset, and when the link individual does not change, the age increases. When the age of the link individual exceeds the age threshold, the link individual is optimized twice. Specifically, if Figure 4 As shown, step S31 also includes: setting the age of the link individuals in the population, the link individuals The age can be expressed as , Can be uniformly set to 1; Figure 5 As shown, A7 also includes resetting the individual age of the selected link, that is, ; A8 also includes increasing the individual age of the selected link, which can make After A7 and A8 are executed, the process goes to A9, the age determination step.

[0092] Specifically, if Figure 5 As shown, A9 includes: judging whether the age of the selected link individual is less than the age threshold, if so, the selected link individual remains unchanged, if not, performing secondary optimization on the selected link individual.

[0093] Specifically, the steps of performing secondary optimization on the selected individual links include:

[0094] Set a temporary similarity threshold for the selected link individual, which is greater than the initial similarity threshold. , where N represents the iteration threshold, which in this embodiment represents the first iteration threshold, Indicates the age threshold.

[0095] Generate a virtual link individual whose similarity to the selected link individual is not less than its temporary similarity threshold as the source link individual, use the source link individual to generate a variant link individual, and use the variant link individual to perform cross-mutation on the selected link individual to obtain an experimental link individual. The method of generating variant link individuals and experimental link individuals in this step is the same as the aforementioned step A5 and is not described in detail here.

[0096] Determine whether the objective function value of the selected link individual is less than the objective function value of the experimental link individual. If so, the selected link individual remains unchanged; if not, the selected link individual is updated with the experimental link individual and the age of the selected link individual is reset, i.e. .

[0097] Furthermore, in order to avoid the link individuals in the population being too similar and to increase the degree of difference between the link individuals so as to find more different optimal and suboptimal access links, in some embodiments of the present application, for example, Figure 4 As shown, the step of optimizing the population in step S3 also includes S33, setting a similarity clustering threshold, dividing the link individuals in the population into several sub-populations by similarity clustering, and the similarity between the link individual with the smallest objective function value in the sub-population and the remaining link individuals is not less than the similarity clustering threshold.

[0098] In this step, the similarity clustering threshold can be set to a reasonable threshold as needed, or the same method as in step S31 can be adopted to set it to the average similarity of the link individuals in the population at this time.

[0099] In some embodiments of the present application, the specific steps of similarity clustering include:

[0100] Sort the link individuals in the population into a sequence according to the objective function value from small to large;

[0101] The first link individual of the sequence is extracted as the benchmark individual, and the link individuals whose similarity with the benchmark individual is not less than the similarity clustering threshold are also extracted from the remaining link individuals of the sequence to form a subpopulation with the benchmark individual, and then the sequence is updated;

[0102] Determine whether there is only one link individual left in the sequence. If so, the remaining link individual forms a subpopulation and the similarity clustering ends. If not, return to the previous step.

[0103] like Figure 4 As shown, the step of optimizing the population in step S3 also includes S34, setting the population capacity of the sub-population, and for the sub-population whose number of link individuals is greater than the population capacity, sorting the link individuals in it from small to large according to the objective function value, and the link individuals with a lower sorting that exceed the population capacity die.

[0104] In this step, the population capacity of the subpopulation can be set to a suitable value according to the number of subpopulations and the total number of link individuals, for example, it can be set to Np / 2k, where k represents the number of subpopulations.

[0105] like Figure 4 As shown, the step of optimizing the population in step S3 also includes S35, obtaining the total number of dead individuals in all sub-populations, regenerating a number of link individuals according to the total number of dead individuals, and the similarity between the regenerated link individuals and the link individual with the smallest objective function value in any sub-population is less than the similarity clustering threshold.

[0106] like Figure 4 As shown, the step of optimizing the population in step S3 further includes S36, resetting the similarity threshold and age for the reborn link individuals;

[0107] In this step, the setting method of the age of the reborn link individuals is the same as that in the aforementioned step S31, which will not be described in detail here; the setting method of the similarity threshold of the reborn link individuals is similar to that in the aforementioned step S31, which is the average similarity of the reborn link individuals, that is, DN represents the number of reborn link individuals, that is, the total number of dead individuals. represents the i-th reborn link individual, i∈(1,2,……,DN).

[0108] like Figure 4 As shown, the step of optimizing the population in step S3 further includes S37, iteratively performing full optimization on the regenerated link individuals until the number of full optimization times for the regenerated link individuals reaches a second iteration threshold.

[0109] In the single full optimization of the reborn link individuals, each reborn link individual is selected for optimization separately, such as Figure 6 As shown, the method for optimizing a single reborn link individual is similar to the method for optimizing the link individual in the aforementioned step S32. The similarities are not described in detail here. The difference is that an exclusion mechanism is added after the optimization of the single reborn link individual. In some embodiments of the present application, the exclusion mechanism is added after step A9. The exclusion mechanism includes: judging whether the similarity between the selected reborn link individual and the link individual with the smallest objective function value in any subpopulation is less than the similarity clustering threshold. If so, the selected reborn link individual remains unchanged. If not, the selected reborn link individual is allowed to die and then be reborn again. Similarly, the similarity between the reborn link individual and the link individual with the smallest objective function value in any subpopulation is less than the similarity clustering threshold.

[0110] It should be noted that in this embodiment, on the basis of the iterative optimization of the population, the reborn link individuals are also iteratively optimized. Therefore, when A9 switches to the secondary optimization operation in this embodiment, in the calculation formula of the temporary similarity threshold, the iterative threshold N is the sum of the number of times the population performs full optimization and the number of times the reborn link individuals perform full optimization.

[0111] In order to avoid the link individuals in the population being too similar, steps S33-S37 select a part of the poor link individuals to be reborn after death, and add an exclusion mechanism to perform iterative optimization to find more different optimal and suboptimal access links. After sufficient iterations, some of the reborn link individuals may change to new optimal or suboptimal access links, but it is possible that due to the absence of other optimal or suboptimal access links and the influence of the exclusion mechanism, the objective function values of some reborn link individuals are not optimized. In order to optimize the objective function values of all reborn link individuals, in some embodiments of the present application, such as Figure 4 As shown, the step of optimizing the population in step S3 also includes S38, canceling the exclusion mechanism, and iteratively optimizing all regenerated link individuals until the number of iterations reaches a third iteration threshold.

[0112] Similarly, in this embodiment, on the basis of iterative optimization of the population and adding an exclusion mechanism to allow the reborn link individuals to iterate and perform full-member optimization, the exclusion mechanism is cancelled, and the reborn link individuals are iterated and optimized for all again. Therefore, when A9 switches to the secondary optimization operation in this embodiment, in the calculation formula of the temporary similarity threshold, the iteration threshold N is the sum of the number of times the population performs full-member optimization, the number of times the link individuals reborn after adding the exclusion mechanism perform full-member optimization, and the number of times the link individuals reborn after canceling the exclusion mechanism perform full-member optimization.

[0113] like Figure 2 As shown, the method of the present application further includes step S4, selecting a number of different link individuals with small objective function values in the population as concurrent access links.

[0114] After step S3, most of the link individuals in the population have achieved objective function value optimization, and several of the link individuals have changed to different optimal or suboptimal access links. The method for selecting concurrent access links can be: the regenerated link individuals are also divided into several subpopulations through similarity clustering, and the subpopulations obtained in steps S33 and S34 are added, and the link individuals with the smallest objective function value in each subpopulation are extracted, sorted from small to large by the objective function value, and several link individuals with the highest order are selected as concurrent access links. In addition, the method for selecting concurrent access links can also be: the final population is re-divided into several subpopulations through similarity clustering, and the link individuals with the smallest objective function value in each subpopulation are extracted, sorted from small to large by the objective function value, and several link individuals with the highest order are selected as concurrent access links.

[0115] After the concurrent access links are selected, the concurrent access requests from the user end through the SDP admission gateway can be equally distributed to several concurrent access links to access the resource end. Of course, concurrent access requests can also be distributed based on the size difference of the objective function values of several concurrent access links.

[0116] An embodiment of the present application also provides an apparatus for concurrent access link planning in an SDP system, comprising a memory and a processor, which may be connected via a bus or other means. The memory may be used to store software programs, computer programs, and modules, such as the programs / modules corresponding to the aforementioned method for concurrent access link planning in an SDP system; the processor implements the aforementioned method for concurrent access link planning in an SDP system by executing the computer programs and modules in the memory.

[0117] The processor may be a central processing unit, a digital signal processor, an application-specific integrated circuit, a field programmable gate array, etc., and the memory may be a high-speed random access memory, a non-volatile memory, etc.

[0118] The present application also provides a storage medium storing computer-executable instructions. When loaded and executed by a processor, the computer-executable instructions implement the aforementioned SDP system concurrent access link planning method. The storage medium may be a combination of one or more of a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, and the like.

[0119] It should be noted that the order of the embodiments of the present invention described above is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The above description is of specific embodiments of this specification, and other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0120] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0121] Those skilled in the art will appreciate that all or part of the steps in the above embodiments can be implemented by hardware or by programming the relevant hardware. The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for concurrent access link planning in an SDP system, characterized in that: The method comprises the following steps: S1, obtain information about communication devices and SDP access gateway devices on the access network, number the communication devices and SDP access gateway devices, and build a communication topology diagram between the SDP access gateway and the resource end based on the numbered devices; S2, generating a population including a plurality of link individuals based on the communication topology graph, wherein a single link individual represents a feasible access link between the SDP admission gateway and the resource end, and the access link is formed by sequentially connecting a plurality of access nodes, and the access node value range is a device number range; S3, takes the access time of individual links as the objective function to optimize the population; S4, select several different link individuals with small objective function values in the population as concurrent access links; The steps of optimizing the population in S3 include: S31, setting the similarity threshold of link individuals within the population; S32, iteratively performing full-member optimization on the population until the number of full-member optimizations on the population reaches a first iteration threshold; in a single full-member optimization of the population, each link individual in the population is selected for optimization separately, and the steps of optimizing the selected single link individual include: A1, other link individuals whose similarity with the selected link individual is not less than its similarity threshold are regarded as neighbors of the selected link individual; A2: Determine whether the selected link individual has neighbors. If not, go to A3; if yes, go to A4. A3, generate a virtual link individual whose similarity to the selected link individual is not less than its similarity threshold as the source link individual, and go to A5; A4, randomly select a link individual from the neighbors of the selected link individual as the source link individual, and go to A5; A5, using the source link individual to generate a variant link individual, and using the variant link individual to perform cross-mutation on the selected link individual to obtain an experimental link individual; A6, judging whether the objective function value of the selected link individual is less than the objective function value of the experimental link individual, if so, go to A8, if not, go to A7; A7, updating the selected link individual with the experimental link individual; A8, the selected link individual remains unchanged.

2. The method for concurrent access link planning of an SDP system according to claim 1, characterized in that: The S31 further includes: setting the age of the link individuals in the population; Said A7 further comprises: resetting the individual age of the selected link and going to A9; Said A8 further comprises: increasing the individual age of the selected link, and going to A9; The A9 includes: judging whether the age of the selected link individual is less than an age threshold, if so, the selected link individual remains unchanged, if not, performing secondary optimization on the selected link individual.

3. The method for concurrent access link planning of an SDP system according to claim 2, wherein: The step of performing secondary optimization on the selected link individuals comprises: A temporary similarity threshold is set for the selected link individual, wherein the temporary similarity threshold is greater than the initial similarity threshold; a virtual link individual having a similarity with the selected link individual not less than the temporary similarity threshold is generated as the source link individual; Using the source link individual to generate a variant link individual, and using the variant link individual to perform cross-mutation on the selected link individual to obtain an experimental link individual; Determine whether the objective function value of the selected link individual is less than the objective function value of the experimental link individual. If so, the selected link individual remains unchanged; if not, the selected link individual is updated with the experimental link individual and the age of the selected link individual is reset.

4. A method for concurrent access link planning in an SDP system according to any one of claims 1 to 3, characterized in that: Said S3 further comprises: S33, setting a similarity clustering threshold, dividing the link individuals in the population into several sub-populations by similarity clustering, and the similarity between the link individual with the smallest objective function value and the remaining link individuals in the sub-population is not less than the similarity clustering threshold; S34: Set the population capacity of the subpopulation. For subpopulations with a number of link individuals greater than the population capacity, sort the link individuals in the subpopulation from smallest to largest according to the objective function value. The link individuals with a lower order that exceed the population capacity die. S35: Obtain the total number of dead individuals in all subpopulations. Regenerate a number of link individuals based on the total number of dead individuals. The similarity between the regenerated link individuals and the link individual with the smallest objective function value in any subpopulation is less than the similarity clustering threshold. S36, reset the similarity threshold and age for the reborn link individuals; S37, iteratively perform full optimization on the regenerated link individuals until the number of full optimizations performed on the regenerated link individuals reaches a second iteration threshold.

5. A method for planning concurrent access links in an SDP system according to claim 4, characterized in that: In a single full-member optimization of reborn link individuals, each reborn link individual is selected for optimization separately; After optimizing the selected single reborn link individual, an exclusion mechanism is added. The exclusion mechanism includes: judging whether the similarity between the selected reborn link individual and the link individual with the smallest objective function value in any subpopulation is less than the similarity clustering threshold. If so, the selected reborn link individual remains unchanged. If not, the selected reborn link individual is allowed to die and be reborn again. The similarity between the reborn link individual and the link individual with the smallest objective function value in any subpopulation is less than the similarity clustering threshold.

6. A method for planning concurrent access links in an SDP system according to claim 5, characterized in that: Said S3 further comprises: S38, cancel the exclusion mechanism, and perform full optimization on the regenerated link individual iterations until the number of iterations reaches the third iteration threshold.

7. A concurrent access link planning device for an SDP system, comprising a memory, a computer program stored therein, and a processor, characterized in that: When the processor executes the computer program, the method for concurrent access link planning of an SDP system according to any one of claims 1 to 6 is implemented.

8. A storage medium, characterized in that: The storage medium stores computer-executable instructions, which, when loaded and executed by a processor, implement a method for concurrent access link planning for an SDP system according to any one of claims 1 to 6.

Images