Data center network unicast fault-tolerant path algorithm based on local security model

Abstract

The invention relates to the technical field of digital information transmission, in particular to a data center network unicast fault-tolerant path algorithm based on a local security model, which comprises the following steps of: determining a plurality of paths in a local twisted cube; judging whether a 0-1 bridge needs to be introduced and is connected to the fault-free path, if the 0-1 bridge does not need to be introduced, directly entering the next step, and if the 0-1 bridge needs to be introduced, connecting the 0-1 bridge to the fault-free path, and enabling each part of the fault-free path to be located in the determined different sub-local distortion cubes; and constructing a complete fault-free unicast transmission path based on the maximum security sub-local distortion cube of each part of the fault-free path, and enabling the data center network to carry out data transmission according to the complete fault-free unicast transmission path. According to the method provided by the invention, the fault tolerance of the data center network based on the local warping cube topology network can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of digital information transmission, and in particular to a data center network unicast fault-tolerant path algorithm based on a local security model. Background Art

[0002] As the scale of data centers continues to expand and the complexity of networks increases, network reliability and fault tolerance issues are receiving more and more attention. In data center networks, the local twisted cube structure is a common network architecture, in which the failure of a node may cause the interruption of the entire network path, thus affecting data transmission and the normal operation of the system.

[0003] In order to ensure the stability and reliability of data transmission, an algorithm with good fault tolerance is needed when a node fails. The existing technology generally uses static routing or algorithms based on global information in fault-tolerant path optimization, but these methods have high time complexity or lack sufficient robustness. When the number of faulty nodes increases, the probability of successful message transmission is very small. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a data center network unicast fault-tolerant path algorithm based on a local security model, which can improve the fault tolerance of a data center network based on a local twisted cube topology network and has important research significance and application value.

[0005] The present invention is achieved through the following technical solutions: A data center network unicast fault-tolerant path algorithm based on a local security model includes the following steps: S1: Randomly select two fault-free nodes in the local twisted cube network of the data center network as the starting point and the end point, and determine multiple paths from the starting point to the end point of the local twisted cube; S2: judging whether it is necessary to introduce a 0-1 bridge to connect the multiple paths from the starting point to the end point in the local twisted cube to the fault-free path according to the shortest path length from the starting point to the end point in the fault-free situation; if it is judged that it is not necessary to introduce a 0-1 bridge, directly proceed to the next step; if it is judged that it is necessary to introduce a 0-1 bridge, connect the 0-1 bridge to the fault-free path, and make each part of the fault-free path located in a determined different sub-local twisted cube, and then proceed to the next step; S3: using the local safety model to classify the non-faulty nodes in all sub-local twisted cubes determined by the data center network, and constructing the maximum safe sub-local twisted cubes of each part of the fault-free path according to the classification of the non-faulty nodes in all sub-local twisted cubes; S4: Construct a complete fault-free unicast transmission path based on the maximum safe sub-local twisted cube of each part of the fault-free path, and enable the data center network to transmit data according to the complete fault-free unicast transmission path.

[0006] Optimally, the two fault-free nodes as the starting point and the end point in step S1 are divided into two cases: one is that the last digits of the binary strings of the starting point and the end point are consistent, and the other is that the last digits of the binary strings of the starting point and the end point are inconsistent.

[0007] Furthermore, in step S2, if a pair of adjacent points in the local distorted cube and ,have , , then connect to , The edge between is a 0-1 bridge. , If all are fault-free nodes, then the 0-1 bridge is a fault-free 0-1 bridge. , If at least one of them is a faulty node, then the 0-1 bridge is a faulty 0-1 bridge, where express The last digit of the binary string of the point, express The last digit of the binary string.

[0008] Further, in step S2, the shortest path length from the starting point to the end point in the absence of a fault is determined according to the following method: when When ,but ; when When ,but ; when When ,but ; when When ,but ; when When ; in: represents the shortest path length from the starting point to the end point without failure, Indicates the starting point The last digit of the binary string, Indicates the end point The last digit of the binary string, Indicates that when the local distorted cube has no fault, the starting point With the end point When the last position is inconsistent, start from the starting point To the end The set of edge connections between any pair of adjacent nodes on the shortest fault-free path, It means taking the number of elements in the set in brackets. In the case of a local distorted cube without failure, Time, starting point To the end The set of edge connections between any pair of neighboring nodes on a fault-free path, In the case of a local distorted cube without failure, Time, starting point To the end The set of edge connections between any pair of adjacent nodes that appear on a fault-free path.

[0009] Furthermore, in step S2, the following method is used to determine whether a 0-1 bridge needs to be introduced into multiple paths from the starting point to the end point in the local twisted cube: when When ,but Two pairs of bridges are used to calculate the shortest path in the fault-free case. If there is a faulty node in the local twisted cube network and both pairs of bridges are fault-free 0-1 bridges, then the 0-1 bridge is introduced to connect to the fault-free path and then proceed to the next step. If there is a faulty node in the local twisted cube network and at least one of the two pairs of bridges is a faulty 0-1 bridge, then the 0-1 bridge is not introduced and the next step is directly proceeded. when When , then the 0-1 bridge is not introduced and the next step is directly entered; when When If the 0-1 bridge calculated in the above process is faulty, a heuristic algorithm is used to find a new 0-1 bridge to connect to the fault-free path and then proceed to the next step; ,like If the 0-1 bridge calculated in the above process has no fault, the 0-1 bridge is directly introduced into the fault-free path and then proceeds to the next step.

[0010] Furthermore, the method of using a heuristic algorithm to find a new 0-1 bridge connected to a fault-free path is as follows: Traverse each pair of fault-free 0-1 bridges, according to and The sum is sorted in ascending order, and then the smallest value of the fault-free 0-1 bridge is selected to connect different sub-local twisted cubes, so that the last consistent nodes in the same sub-local twisted cube are connected to the last inconsistent nodes in other sub-local twisted cubes through the 0-1 bridge, where: Indicates the node whose last bit of the 0-1 bridge is 0. Indicates the node whose last bit of the 0-1 bridge is 1.

[0011] Optimized, in step S3, based on the local safety model, the classification of non-faulty nodes in all sub-local twisted cubes includes locally unsafe nodes and locally safe nodes. If there are at least two faulty nodes among the adjacent nodes of a node, or the sum of the number of locally unsafe neighbor nodes and faulty neighbor nodes is greater than or equal to three, then the node is determined to be a locally unsafe node. If there are less than two faulty nodes among the adjacent nodes of a node, and the sum of the number of locally unsafe neighbor nodes and faulty neighbor nodes is less than three, then the node is determined to be a locally safe node. If there are locally safe neighbor nodes around the locally unsafe node, it is a locally weakly unsafe node. If there are no locally safe neighbor nodes around the locally unsafe node, it is a locally strongly unsafe node.

[0012] Optimized, for any two nodes with the same last digit, the maximum safe sub-local twisted cube is the maximum sub-local twisted cube that contains the first sub-local twisted cube or the second sub-local twisted cube, and there is at least one local safe node in this maximum sub-local twisted cube, wherein: the first sub-local twisted cube is a sub-local twisted cube whose last digits of the binary strings of the two nodes are both 0, and the second sub-local twisted cube is a sub-local twisted cube whose last digits of the binary strings of the two nodes are both 1.

[0013] Furthermore, in step S3, the following method is used to construct the maximum safe sub-local twisted cube of each part of the fault-free path: If the 0-1 bridge is not introduced, and the last digits of any two node binary strings are both 0, then select the first sub-local twisted cube whose last digits are all 0 in the local twisted cube, and then determine whether the first sub-local twisted cube has a local safety node. If there is at least one local safety node, the current first sub-local twisted cube is its maximum sub-local twisted cube; if there is no local safety node, reduce the points in the first sub-local twisted cube to obtain a new first sub-local twisted cube, until the current new first sub-local twisted cube no longer contains the first sub-local twisted cube whose last digits of the two node binary strings are both 0. If it still does not contain a local safety node, determine the maximum safe sub-local twisted cube that cannot construct a fault-free path; If the 0-1 bridge is not introduced, and the last digits of any two node binary strings are both 1, first select the second sub-local twisted cube whose last digits are all 1 in the local twisted cube, and then determine whether the second sub-local twisted cube has at least one local safety node. If there is at least one local safety node, the current second sub-local twisted cube is the maximum safety sub-local twisted cube. If there is no local safety node, reduce the points in the second sub-local twisted cube to obtain a new second sub-local twisted cube, until the current new second sub-local twisted cube no longer contains the second sub-local twisted cube whose last digits of the two node binary strings are both 1. If there is still no local safety node, it is determined that the maximum safety sub-local twisted cube of each part of the fault-free path cannot be constructed. If the 0-1 bridge is introduced, the 0-1 bridge divides the local warp cube into multiple first sub-local warp cubes and multiple second sub-local warp cubes, and constructs the maximum safe sub-local warp cube of each part of the fault-free path in each first sub-local warp cube and second sub-local warp cube according to the method without introducing the 0-1 bridge.

[0014] Furthermore, in step S4, when the last digits of the binary strings of any two nodes are consistent, a complete fault-free unicast transmission path is constructed in the determined maximum safe sub-local twisted cube according to the following method: like , then the message is sent directly from the starting point Routing to the destination , thereby building a complete trouble-free unicast transmission path; like , then the message is routed to the starting point first To the end The non-faulty neighbor on the shortest path reroutes the message to the new non-faulty neighbor. To the end If there is no fault-free neighbor on the shortest path, the message is routed to the starting point on any non-shortest path. A fault-free neighbor is found and the message is rerouted on the new fault-free neighbor, thereby constructing a complete fault-free unicast transmission path; when , messages are routed as follows to build a complete, trouble-free unicast transmission path: If the starting point If it is a local safety node, the message is routed directly to the origin To the end The starting point on the shortest path The local safety node neighbors are re-routed on the new local safety node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point If it is a local weak and unsafe node, the message is first routed to the starting point To the end The starting point on the shortest path After finding the local weakly unsafe node neighbors, the message is rerouted on the new local weakly unsafe node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local weak unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local weakly unsafe node neighbors, then from the starting point To the end Routes the message to the starting point on any non-shortest path The local safety node neighbors are re-routed on the new local safety node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point If it is a local strong unsafe node, the message is first routed to the starting point To the end The starting point on the shortest path The local weakly insecure node neighbors are then rerouted on the new local weakly insecure node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local strong unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local weakly unsafe node neighbors, then route the message to the starting point To the end The starting point on the shortest path The local strongly insecure node neighbors are then rerouted on the new local strongly insecure node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local strong unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local strong unsafe node neighbors, then from the starting point To the end Routes the message to the starting point on any non-shortest path The local weakly insecure node neighbors are then rerouted on the new local weakly insecure node neighbors, thereby building a complete fault-free unicast transmission path; When the starting point and end point When the last digits of the binary strings are inconsistent, the maximum security sub-local twisted cube determined in step S3 is a plurality of different maximum security sub-local twisted cubes. Then, in each maximum security sub-local twisted cube, a fault-free unicast transmission path is constructed respectively according to the method when the last digits of the starting and ending binary strings are consistent. Then, the fault-free unicast transmission paths constructed by different maximum security sub-local twisted cubes are connected through a 0-1 bridge to obtain a complete fault-free unicast transmission path.

[0015] Beneficial effects of the invention: The present invention provides a data center network unicast fault-tolerant path algorithm based on a local security model. It uses a 0-1 bridge to connect multiple sub-local twisted cubes. Through the construction of a local security model and a maximum security sub-local twisted cube, the security and routing strategy of each sub-local twisted cube are determined. Finally, each part of the fault-free path is effectively connected through a 0-1 bridge, thereby forming a complete fault-free unicast transmission path from the starting point to the end point. The fault tolerance of the data center network based on the local twisted cube topology network is improved, and the reliability and stability of data during the transmission process are ensured. It has important research significance and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the process of the present invention.

[0017] Figure 2 The present invention Schematic diagram. DETAILED DESCRIPTION

[0018] A data center network unicast fault-tolerant path algorithm based on a local security model includes the following steps, and its flow chart is as follows: Figure 1 As shown: S1: Randomly select two fault-free nodes in the local twisted cube network of the data center network as the starting point and the end point, and determine multiple paths from the starting point to the end point of the local twisted cube; exist Locally distorted cube In the example, all nodes can be represented as a A binary string of bits, arbitrarily selecting two fault-free nodes as the starting point and end point .

[0019] Specifically, the two fault-free nodes as the starting point and the end point can be divided into two situations: one is that the last digits of the binary strings of the starting point and the end point are consistent, and the other is that the last digits of the binary strings of the starting point and the end point are inconsistent. The situation where the last digits are consistent is further divided into two situations: the last digits are both 0 and the last digits are both 1.

[0020] S2: judging whether it is necessary to introduce a 0-1 bridge to connect the multiple paths from the starting point to the end point in the local twisted cube to the fault-free path according to the shortest path length from the starting point to the end point in the fault-free situation; if it is judged that it is not necessary to introduce a 0-1 bridge, directly proceed to the next step; if it is judged that it is necessary to introduce a 0-1 bridge, connect the 0-1 bridge to the fault-free path, and make each part of the fault-free path located in a determined different sub-local twisted cube, and then proceed to the next step; Definition of 0-1 bridge: If a pair of adjacent points in a locally distorted cube and ,have , , then connect to , The edge between is a 0-1 bridge. , If all are fault-free nodes, then the 0-1 bridge is a fault-free 0-1 bridge. , If at least one of them is a faulty node, then the 0-1 bridge is a faulty 0-1 bridge, where express The last digit of the binary string of the point, express The last digit of the binary string.

[0021] Specifically, in the case of no fault, arrive The shortest path length is , for different numbers of starting and ending points, ,in: Indicates the starting point and end point Different positions, Represents exclusive-or-and.

[0022] Define two edge connection methods: one for representing the connection between two nodes whose last digits are both 0 and the other for representing the connection between two nodes whose last digits are both 1. For two nodes whose last digits are both 0, define a bit binary string , For the A binary string with bit 1 and all other bits 0; For two nodes whose last digits are both 1, define a bit binary string ,in For the Position and A binary string with bit 1 and all other bits 0; according to Locally distorted cube The connection method further defines , .

[0023] In the absence of faults, from the starting point To the end The shortest path means choosing the least number of edges and using the least number of or and Equal. is included in The smallest subset of ;make is included in The smallest subset of ;make is included in The smallest subset of ,satisfy , , These three formulas can eliminate the starting point To the end Different positions between.

[0024] Specifically, the shortest path length from the starting point to the end point under fault-free conditions can be determined according to the following method: when When ,but ; when When ,but ; when When ,but ; when When ,but ; when When ; in: represents the shortest path length from the starting point to the end point without failure, Indicates the starting point The last digit of the binary string, Indicates the end point The last digit of the binary string, Indicates that when the local distorted cube has no fault, the starting point With the end point When the last position is inconsistent, start from the starting point To the end The set of edge connections between any pair of adjacent nodes on the shortest fault-free path, is included in the set The smallest subset of A collection of It means taking the number of elements in the set in brackets. In the case of a local distorted cube without failure, Time, starting point To the end The set of edge connections between any pair of neighboring nodes on a fault-free path, is included in the set The smallest subset of and satisfies, A collection of In the case of a local distorted cube without failure, Time, starting point To the end The set of edge connections between any pair of neighboring nodes on a fault-free path, is included in the set The smallest subset of A collection of .

[0025] by Figure 2 As an example, suppose we want to calculate arrive The shortest path without failure, where: ; ; because therefore, It indicates that arrive The connection mode of all nodes on the shortest path without failure, and all possible nodes on the shortest path just form the largest sub-local twisted cube .choose Any element in one of the ways can be selected through , , Calculation, thus achieving arrive The shortest path in the absence of failures.

[0026] same, The elements in also represent a way to connect edges to build arrive path in the absence of a fault. The characteristics of the structure, the connection between two nodes whose last digits are both 1 can be used Elements in First select , , and then use the 0-1 bridge to make Route to ,Right now After that, use , Use the 0-1 bridge again to reach the end .

[0027] therefore It is from arrive Path length in the fault-free case.

[0028] Specifically, the following method can be used to determine whether a 0-1 bridge needs to be introduced into multiple paths from the starting point to the end point in the local twisted cube: when When ,but Two pairs of bridges are used to calculate the shortest path in the fault-free case. If there is a faulty node in the local twisted cube network and both pairs of bridges are fault-free 0-1 bridges, then a 0-1 bridge is introduced to connect to the fault-free path and then proceed to the next step. If there is a faulty node in the local twisted cube network and at least one of the two pairs of bridges is a faulty 0-1 bridge, then the 0-1 bridge is not introduced. Due to the consistency of the last bit, the next step can be directly entered to construct the maximum safe sub-local twisted cube. when When , then the 0-1 bridge is not introduced and the next step is directly entered; when When If the 0-1 bridge calculated in the above process is faulty, a heuristic algorithm is used to find a new 0-1 bridge to connect to the fault-free path and then proceed to the next step; Specifically, the method of using a heuristic algorithm to find a new 0-1 bridge connected to a fault-free path is as follows: Traverse each pair of fault-free 0-1 bridges, according to and The sum is sorted in ascending order, and then the smallest value of the fault-free 0-1 bridge is selected to connect different sub-local twisted cubes, so that the last consistent nodes in the same sub-local twisted cube are connected to the last inconsistent nodes in other sub-local twisted cubes through the 0-1 bridge, where: Indicates the node whose last bit of the 0-1 bridge is 0. Indicates the node whose last bit of the 0-1 bridge is 1.

[0029] ,like If the 0-1 bridge calculated in the above process has no fault, the 0-1 bridge is directly introduced into the fault-free path and then proceeds to the next step.

[0030] like Figure 2 As shown, calculation arrive Passing the 0-1 bridge. Since the last digit of the starting point is 0, the node connection method with the last digit being 0 is used first, that is, In the next step, we need to use a 0-1 bridge to connect the nodes whose last digit is 1, that is , Indicates the bridge from one node in the bracket to another node. If the 0-1 bridge is fault-free, the message will be sent from arrive Routing and Slave arrive Routing, finally The two paths will be connected. If If a fault occurs, a heuristic algorithm is used to find a fault-free 0-1 bridge. The specific rules are as follows The values ​​are arranged in ascending order, and the 0-1 bridge with the smallest value is selected as the bridge for message routing.

[0031] S3: using the local safety model to classify the non-faulty nodes in all sub-local twisted cubes determined by the data center network, and constructing the maximum safe sub-local twisted cubes of each part of the fault-free path according to the classification of the non-faulty nodes in all sub-local twisted cubes; Based on the local safety model, the classification of non-faulty nodes in all sub-local twisted cubes includes locally unsafe nodes and locally safe nodes. If there are at least two faulty nodes among the neighboring nodes of a node, or the sum of the number of locally unsafe neighbor nodes and faulty neighbor nodes is greater than or equal to three, then the node is judged as a locally unsafe node. If there are less than two faulty nodes among the neighboring nodes of a node, and the sum of the number of locally unsafe neighbor nodes and faulty neighbor nodes is less than three, then the node is judged to be a locally safe node. If there are local safe neighbor nodes around the local unsafe node, it is a local weak unsafe node. If there are no locally safe neighbor nodes around a locally unsafe node, it is a locally strongly unsafe node. .

[0032] like Figure 2 As shown, where it is specified , , is the fault point, then middle, yes Node, more specifically, is Node, and middle, is the LS node, It represents the sub-local twisted cube formed by the two nodes with the same last digit in the brackets. If the last digits of these two nodes are both 0, it means that the first sub-local twisted cube is formed by these two nodes. If the last digits of these two nodes are both 1, it means that the second sub-local twisted cube is formed by these two nodes.

[0033] For any two nodes with the same last digit, the maximum safe sub-local twisted cube is the maximum sub-local twisted cube that contains the first sub-local twisted cube or the second sub-local twisted cube, and there is at least one local safe node in this maximum sub-local twisted cube, where: the first sub-local twisted cube is a sub-local twisted cube whose last digits of the binary strings of the two nodes are both 0, and the second sub-local twisted cube is a sub-local twisted cube whose last digits of the binary strings of the two nodes are both 1.

[0034] Specifically, the following method can be used to construct the maximum safe sub-local twisted cube of each part of the fault-free path: If the 0-1 bridge is not introduced, and the last digits of any two node binary strings are both 0, then select the first sub-local twisted cube whose last digits are all 0 in the local twisted cube, and then determine whether the first sub-local twisted cube has a local safety node. If there is at least one local safety node, the current first sub-local twisted cube is its maximum sub-local twisted cube; if there is no local safety node, reduce the points in the first sub-local twisted cube to obtain a new first sub-local twisted cube, until the current new first sub-local twisted cube no longer contains the first sub-local twisted cube whose last digits of the two node binary strings are both 0. If it still does not contain a local safety node, determine the maximum safe sub-local twisted cube that cannot construct a fault-free path; If the 0-1 bridge is not introduced, and the last digits of any two node binary strings are both 1, first select the second sub-local twisted cube whose last digits are all 1 in the local twisted cube, and then determine whether the second sub-local twisted cube has at least one local safety node. If there is at least one local safety node, the current second sub-local twisted cube is the maximum safety sub-local twisted cube. If there is no local safety node, reduce the points in the second sub-local twisted cube to obtain a new second sub-local twisted cube, until the current new second sub-local twisted cube no longer contains the second sub-local twisted cube whose last digits of the two node binary strings are both 1. If there is still no local safety node, it is determined that the maximum safety sub-local twisted cube of each part of the fault-free path cannot be constructed. If the 0-1 bridge is introduced, the 0-1 bridge divides the local warp cube into multiple first sub-local warp cubes and multiple second sub-local warp cubes, and constructs the maximum safe sub-local warp cube of each part of the fault-free path in each first sub-local warp cube and second sub-local warp cube according to the method without introducing the 0-1 bridge.

[0035] Taking into account bandwidth resource consumption and message delivery reliability, if the maximum safe sub-local twisted cube cannot be found when the message is first routed to the current node, the message will be first routed to a non-faulty neighboring node around the current node, and the maximum safe sub-local twisted cube will be re-constructed on the new node. The process of re-finding the maximum safe sub-local twisted cube can be repeated up to three times. If the maximum safe sub-local twisted cube is still not found after three attempts, the routing is considered to have failed.

[0036] S4: Construct a complete fault-free unicast transmission path based on the maximum safe sub-local twisted cube of each part of the fault-free path, and enable the data center network to transmit data according to the complete fault-free unicast transmission path.

[0037] Specifically, when the last digits of the binary strings of any two nodes are consistent, a complete fault-free unicast transmission path is constructed in the determined maximum secure sub-local twisted cube according to the following method: like , then the message is sent directly from the starting point Routing to the destination , thereby constructing a complete fault-free unicast transmission path; this ensures that a fault-free unicast transmission path can be found in the maximum security sub-local twisted cube.

[0038] like , then the message is routed to the starting point first To the end The non-faulty neighbor on the shortest path reroutes the message to the new non-faulty neighbor. To the end If there is no fault-free neighbor on the shortest path, the message is routed to the starting point on any non-shortest path. A fault-free neighbor is found and the message is rerouted on the new fault-free neighbor, thereby constructing a complete fault-free unicast transmission path; when , messages are routed as follows to build a complete, trouble-free unicast transmission path: If the starting point If it is a local safety node, the message is routed directly to the origin To the end The starting point on the shortest path The local safety node neighbors of the network are rerouted on the new local safety node neighbors, thereby building a complete fault-free unicast transmission path, because the path is always available; If the starting point If it is a local weak and unsafe node, the message is first routed to the starting point To the end The starting point on the shortest path After finding the local weakly unsafe node neighbors, the message is rerouted on the new local weakly unsafe node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local weak unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local weakly unsafe node neighbors, then from the starting point To the end Routes the message to the starting point on any non-shortest path The local safety node neighbors are re-routed on the new local safety node neighbors, thereby building a complete fault-free unicast transmission path, because this path is always available; If the starting point If it is a local strong unsafe node, the message is first routed to the starting point To the end The starting point on the shortest path The local weakly unsafe node neighbors are then rerouted on the new local weakly unsafe node neighbors, thereby building a complete fault-free unicast transmission path, because this path is always available; If the starting point It is a local strong unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local weakly unsafe node neighbors, then route the message to the starting point To the end The starting point on the shortest path The local strongly insecure node neighbors are then rerouted on the new local strongly insecure node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local strong unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local strong unsafe node neighbors, then from the starting point To the end Routes the message to the starting point on any non-shortest path The local weakly insecure node neighbors are then rerouted on the new local weakly insecure node neighbors, thereby building a complete fault-free unicast transmission path; When the starting point and end point When the last digits of the binary strings are inconsistent, the maximum security sub-local twisted cube determined in step S3 is a plurality of different maximum security sub-local twisted cubes. Then, in each maximum security sub-local twisted cube, a fault-free unicast transmission path is constructed respectively according to the method when the last digits of the starting and ending binary strings are consistent. Then, the fault-free unicast transmission paths constructed by different maximum security sub-local twisted cubes are connected through a 0-1 bridge to obtain a complete fault-free unicast transmission path.

[0039] In summary, the present invention provides a data center network unicast fault-tolerant path algorithm based on a local security model, which uses a 0-1 bridge to connect multiple sub-local twisted cubes, and determines the security and routing strategy of each sub-local twisted cube through the construction of a local security model and a maximum security sub-local twisted cube. Finally, the 0-1 bridge is used to effectively connect each part of the fault-free path, thereby forming a complete fault-free unicast transmission path from the starting point to the end point, which effectively improves the fault tolerance of the data center network based on the local twisted cube topology network, can ensure the reliability and stability of data during transmission, and has important research significance and application value.

[0040] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data center network unicast fault-tolerant path algorithm based on a local security model, characterized by: The steps include: S1: Randomly select two fault-free nodes in the local twisted cube network of the data center network as the starting point and the end point, and determine multiple paths from the starting point to the end point of the local twisted cube; S2: judging whether it is necessary to introduce a 0-1 bridge to connect the multiple paths from the starting point to the end point in the local twisted cube to the fault-free path according to the shortest path length from the starting point to the end point in the fault-free situation; if it is judged that it is not necessary to introduce a 0-1 bridge, directly proceed to the next step; if it is judged that it is necessary to introduce a 0-1 bridge, connect the 0-1 bridge to the fault-free path, and make each part of the fault-free path located in a determined different sub-local twisted cube, and then proceed to the next step; S3: using the local safety model to classify the non-faulty nodes in all sub-local twisted cubes determined by the data center network, and constructing the maximum safe sub-local twisted cubes of each part of the fault-free path according to the classification of the non-faulty nodes in all sub-local twisted cubes; S4: Construct a complete fault-free unicast transmission path based on the maximum safe sub-local twisted cube of each part of the fault-free path, and enable the data center network to transmit data according to the complete fault-free unicast transmission path.

2. According to the local security model-based unicast fault-tolerant path algorithm for data center networks according to claim 1, it is characterized by: There are two cases for the two non-faulty nodes as the starting point and the end point in step S1: one is that the last digits of the binary strings of the starting point and the end point are consistent, and the other is that the last digits of the binary strings of the starting point and the end point are inconsistent.

3. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 1, characterized in that: In step S2, if a pair of adjacent points in the local distorted cube and ,have , , then connect to , The edge between is a 0-1 bridge. , If all are fault-free nodes, then the 0-1 bridge is a fault-free 0-1 bridge. , If at least one of them is a faulty node, then the 0-1 bridge is a faulty 0-1 bridge, where express The last digit of the binary string. express The last digit of the binary string.

4. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 3 is characterized by: In step S2, the shortest path length from the starting point to the end point in the absence of a fault is determined according to the following method: when When ,but ; when When ,but ; when When ,but ; when When ,but ; when When ; in: represents the shortest path length from the starting point to the end point without failure, Indicates the starting point The last digit of the binary string, Indicates the end point The last digit of the binary string, Indicates that when the local distorted cube has no fault, the starting point With the end point When the last position is inconsistent, start from the starting point To the end The set of edge connections between any pair of adjacent nodes on the shortest fault-free path, It means taking the number of elements in the set in brackets. In the case of a local distorted cube without failure, Time, starting point To the end The set of edge connections between any pair of neighboring nodes on a fault-free path, In the case of a local distorted cube without failure, Time, starting point To the end The set of edge connections between any pair of adjacent nodes that appear on a fault-free path.

5. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 4 is characterized by: In step S2, the following method is used to determine whether a 0-1 bridge needs to be introduced into multiple paths from the starting point to the end point in the local twisted cube: when When ,but Two pairs of bridges are used to calculate the shortest path in the fault-free case. If there is a faulty node in the local twisted cube network and both pairs of bridges are fault-free 0-1 bridges, then the 0-1 bridge is introduced to connect to the fault-free path and then proceed to the next step. If there is a faulty node in the local twisted cube network and at least one of the two pairs of bridges is a faulty 0-1 bridge, then the 0-1 bridge is not introduced and the next step is directly proceeded. when When , then the 0-1 bridge is not introduced and the next step is directly entered; when When If the 0-1 bridge calculated in the above process is faulty, a heuristic algorithm is used to find a new 0-1 bridge to connect to the fault-free path and then proceed to the next step; ,like If the 0-1 bridge calculated in the above process has no fault, the 0-1 bridge is directly introduced into the fault-free path and then proceeds to the next step.

6. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 5, characterized in that: The method of using a heuristic algorithm to find a new 0-1 bridge to connect to a fault-free path is as follows: Traverse each pair of fault-free 0-1 bridges, according to and The sum is sorted in ascending order, and then the smallest value of the fault-free 0-1 bridge is selected to connect different sub-local twisted cubes, so that the last consistent nodes in the same sub-local twisted cube are connected to the last inconsistent nodes in other sub-local twisted cubes through the 0-1 bridge, where: Indicates the node whose last bit of the 0-1 bridge is 0. Indicates the node whose last bit of the 0-1 bridge is 1.

7. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 1, characterized in that: In step S3, based on the local safety model, the classification of non-faulty nodes in all sub-local twisted cubes includes locally unsafe nodes and locally safe nodes. If there are at least two faulty nodes among the adjacent nodes of a node, or the sum of the number of locally unsafe neighbor nodes and faulty neighbor nodes is greater than or equal to three, then the node is determined to be a locally unsafe node. If there are less than two faulty nodes among the adjacent nodes of a node, and the sum of the number of locally unsafe neighbor nodes and faulty neighbor nodes is less than three, then the node is determined to be a locally safe node. If there are locally safe neighbor nodes around the locally unsafe node, it is a locally weakly unsafe node. If there are no locally safe neighbor nodes around the locally unsafe node, it is a locally strongly unsafe node.

8. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 1, characterized in that: For any two nodes with the same last digit, the maximum safe sub-local twisted cube is the maximum sub-local twisted cube that contains the first sub-local twisted cube or the second sub-local twisted cube, and there is at least one local safe node in this maximum sub-local twisted cube, where: the first sub-local twisted cube is a sub-local twisted cube whose last digits of the binary strings of the two nodes are both 0, and the second sub-local twisted cube is a sub-local twisted cube whose last digits of the binary strings of the two nodes are both 1.

9. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 8, characterized in that: In step S3, the following method is used to construct the maximum safe sub-local twisted cube of each part of the fault-free path: If the 0-1 bridge is not introduced, and the last digits of any two node binary strings are both 0, then select the first sub-local twisted cube whose last digits are all 0 in the local twisted cube, and then determine whether the first sub-local twisted cube has a local safety node. If there is at least one local safety node, the current first sub-local twisted cube is its maximum sub-local twisted cube; if there is no local safety node, reduce the points in the first sub-local twisted cube to obtain a new first sub-local twisted cube, until the current new first sub-local twisted cube no longer contains the first sub-local twisted cube whose last digits of the two node binary strings are both 0. If it still does not contain a local safety node, determine the maximum safe sub-local twisted cube that cannot construct a fault-free path; If the 0-1 bridge is not introduced, and the last digits of any two node binary strings are both 1, first select the second sub-local twisted cube whose last digits are all 1 in the local twisted cube, and then determine whether the second sub-local twisted cube has at least one local safety node. If there is at least one local safety node, the current second sub-local twisted cube is the maximum safety sub-local twisted cube. If there is no local safety node, reduce the points in the second sub-local twisted cube to obtain a new second sub-local twisted cube, until the current new second sub-local twisted cube no longer contains the second sub-local twisted cube whose last digits of the two node binary strings are both 1. If there is still no local safety node, it is determined that the maximum safety sub-local twisted cube of each part of the fault-free path cannot be constructed. If the 0-1 bridge is introduced, the 0-1 bridge divides the local warp cube into multiple first sub-local warp cubes and multiple second sub-local warp cubes, and constructs the maximum safe sub-local warp cube of each part of the fault-free path in each first sub-local warp cube and second sub-local warp cube according to the method without introducing the 0-1 bridge.

10. The data center network unicast fault-tolerant path algorithm based on a local security model according to claim 4, characterized in that: In step S4, when the last digits of any two node binary strings are consistent, a complete fault-free unicast transmission path is constructed in the determined maximum safe sub-local twisted cube according to the following method: like , then the message is sent directly from the starting point Routing to the destination , thereby building a complete trouble-free unicast transmission path; like , then the message is routed to the starting point first To the end The non-faulty neighbor on the shortest path reroutes the message to the new non-faulty neighbor. To the end If there is no fault-free neighbor on the shortest path, the message is routed to the starting point on any non-shortest path. A fault-free neighbor is found and the message is rerouted on the new fault-free neighbor, thereby constructing a complete fault-free unicast transmission path; when , messages are routed as follows to build a complete, trouble-free unicast transmission path: If the starting point If it is a local safety node, the message is routed directly to the origin To the end The starting point on the shortest path The local safety node neighbors are re-routed on the new local safety node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point If it is a local weak and unsafe node, the message is first routed to the starting point To the end The starting point on the shortest path After finding the local weakly unsafe node neighbors, the message is rerouted on the new local weakly unsafe node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local weak unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local weakly unsafe node neighbors, then from the starting point To the end Routes the message to the starting point on any non-shortest path The local safety node neighbors are re-routed on the new local safety node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point If it is a local strong unsafe node, the message is first routed to the starting point To the end The starting point on the shortest path The local weakly insecure node neighbors are then rerouted on the new local weakly insecure node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local strong unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local weakly unsafe node neighbors, then route the message to the starting point To the end The starting point on the shortest path The local strongly insecure node neighbors are then rerouted on the new local strongly insecure node neighbors, thereby building a complete fault-free unicast transmission path; If the starting point It is a local strong unsafe node and the starting point cannot be found To the end The starting point on the shortest path The local strong unsafe node neighbors, then from the starting point To the end Routes the message to the starting point on any non-shortest path The local weakly insecure node neighbors are then rerouted on the new local weakly insecure node neighbors, thereby building a complete fault-free unicast transmission path; When the starting point and end point When the last digits of the binary strings are inconsistent, the maximum security sub-local twisted cube determined in step S3 is a plurality of different maximum security sub-local twisted cubes. Then, in each maximum security sub-local twisted cube, a fault-free unicast transmission path is constructed respectively according to the method when the last digits of the starting and ending binary strings are consistent. Then, the fault-free unicast transmission paths constructed by different maximum security sub-local twisted cubes are connected through a 0-1 bridge to obtain a complete fault-free unicast transmission path.

Images