An optimized deployment method for network modality coexistence in a multimodal network environment

Abstract

The present invention discloses an optimization deployment method for network mode coexistence in a multimodal network environment, and belongs to the field of multimodal intelligent networks. The method includes: modeling the deployment problem of network mode as a 0-1 integer programming problem, comprehensively considering key constraints such as multimodal network element resources, link resources and traffic forwarding delay to ensure user service quality; at the same time, through joint routing optimization, the connectivity of each network mode is ensured; in view of the non-convexity of the original problem, the present invention converts it into an equivalent 0-1 integer linear programming problem, and designs a network mode deployment and routing selection algorithm based on random rounding; the algorithm can realize efficient deployment of network mode under limited resource conditions and ensure network performance. The method proposed by the present invention effectively reduces the link load and improves the utilization rate of network resources under the premise of ensuring service quality.

Description

Technical Field

[0001] The present invention belongs to the field of multimodal intelligent networks, and specifically relates to an optimized deployment method for network mode coexistence in a multimodal network environment. Background Art

[0002] In the digital age, with the rapid development of vertical industry applications such as the Industrial Internet, Internet of Vehicles, and telemedicine, new service models continue to emerge, bringing unprecedented challenges to the network, including the urgent need for large bandwidth, high reliability, low latency, and huge connection requirements. The traditional single network technology system can no longer meet these complex and diverse needs. To solve this problem, the new concept of multimodal network environment has been proposed, which adopts the method of "separating the technology system from the network environment" to achieve the symbiosis and evolution of multiple network modes to meet the needs of different vertical application industries. This environment supports the dynamic loading and operation of multiple network modes on the same infrastructure, improving the flexibility and efficiency of the network.

[0003] However, existing technologies have failed to effectively solve the problem of network modality deployment in multimodal network environments, especially how to ensure the effective deployment of various network modalities and service quality assurance under the condition of limited resources. Although there have been studies on compilation optimization and node system design for multimodal networks, further exploration is still needed on how to optimize the deployment of network modalities to meet different network requirements and ensure service quality. In addition, the diversity of terminals in multimodal network environments also brings challenges to ensuring the connectivity of various network modalities. Therefore, in the face of the needs of diversified network modalities, how to generate network-level network modality deployment solutions in multimodal network environments, maximize the fit between network modality generation requirements and network resource efficiency, and achieve efficient deployment of network modalities in multimodal network data planes are issues that need to be urgently addressed. Summary of the invention

[0004] The purpose of the present invention is to address the deficiencies in the prior art and to propose an optimized deployment method for coexistence of network modes in a multimodal network environment.

[0005] The objective of the present invention is achieved through the following technical solution: A method for optimizing the deployment of network modal coexistence in a multimodal network environment, comprising the following steps:

[0006] Input the network topology flow set and the candidate path set for each flow;

[0007] According to the network topology traffic set and the candidate path set, a relaxation algorithm is used to process the network modal deployment and path selection problem, the value range of the path selection variable and the modal deployment variable is relaxed from 0 and 1 to the continuous interval [0, 1], each flow is allowed to select a path and modal deployment with a probability in the interval of 0-1, and a relaxed Joint Optimization of Network Modal Deployment and Routing (JMDR) problem is constructed; the relaxed JMDR problem is solved using a linear programming solver to obtain a decimal solution of the path selection variable and the modal deployment variable;

[0008] Based on the decimal solution, a path is randomly selected, and the path selection variables and modal deployment variables are rounded according to probability to determine the specific transmission path of each flow and the modal deployment on the switch; the resource consumption on the path selected for each flow is calculated, and the optimality of the selected path and modal deployment in resource utilization is ensured.

[0009] Furthermore, the linear programming solver is a Gurobi or CPLEX tool, which is used to solve the relaxed JMDR problem and output the optimal solution for each flow path and mode selection.

[0010] Furthermore, the use of the relaxation algorithm to process the network mode deployment and path selection problem is specifically as follows:

[0011] The probability of representing a network modality as its routing path;

[0012] Use a linear programming solver to solve the linear programming and define the load balancing factor , the linear programming is expressed as the following optimization objective:

[0013]

[0014] in, Representation Flow Whether to use path For routing, if If 1 is selected, this path is selected for routing. If it is 0, it means that this path is not selected for routing; Indicates the path, represents the flow set, Representation Flow Whether the network mode of is m; Indicates whether the network mode m is deployed on the switch v, m represents the network mode, M represents the network mode set; p represents the path, Representation Flow The candidate path set; v represents a switch, V represents a switch set, represents the set of all switches on path p; Indicates whether network mode m is deployed on switch v. If If it is 1, it means that the network mode is deployed on the switch. If it is 0, it means that the network mode is not deployed on the switch; Indicates the amount of resources consumed when deploying network mode m on a switch; represents the resource capacity of switch v that can be used to deploy network modes; Indicates the switch processing flow The amount of resources required; Indicates the resource capacity of switch v that can be used to configure routing; Representation Flow The flow size, represents the bandwidth capacity of link e, e represents the link, E represents the link set, represents the transmission delay of link e; represents the processing delay of switch v; Representation Flow The transmission delay threshold of Indicates the ratio of link resource consumption to link capacity;

[0015] The decimal solution of the optimization objective obtained by the above formula is: and , the integer solution is and ; Based on decimal solution Construct an integer solution to the optimization objective before relaxation , with probability Will Set to 1 to determine the routing selection; set the flow The path you choose The switch set is ;

[0016] Based on decimal solution Construct an integer solution to the optimization objective before relaxation The value of Required network modality Deployed on a switch on, and Set to 1.

[0017] Furthermore, the method of using a relaxation algorithm to process the network mode deployment and path selection problem also includes: if an overloaded switch is found, the overloaded switch set Initialize to an empty set; traverse and calculate each switch Resource consumption of network mode deployment ; If resource consumption Exceeding preset resource limits , then the switch Add to overload switch collection in; use Indicates a switch The flow collection on Modal deployment variables right The flows in are sorted in ascending order; for each flow , from the candidate path set A new routing path is selected, wherein the routing path has the largest value, and satisfy all constraints; update according to the new routing path and The value of , after updating, recalculates the resource consumption of the path ; If the resource consumption at this time Less than the preset resource limit , then exit the loop and solve the switch overload problem; otherwise, loop the above steps until a suitable path is found.

[0018] Furthermore, the randomly selecting a path includes: randomly selecting an optimal path for each flow based on a fractional solution of a path selection variable by a probability distribution method.

[0019] Furthermore, the modality deployment includes: deploying a network modality for each switch on the selected path of each flow and ensuring that the switch has sufficient resources to support the transmission requirements of the flow.

[0020] Furthermore, the method further comprises: when deploying the network mode, calculating the resource consumption of each switch and determining whether it exceeds a predetermined resource limit. If the resource consumption exceeds the limit, reselecting the path and adjusting the mode deployment to ensure efficient use of network resources.

[0021] Furthermore, the calculation of resource consumption on the path selected by each flow includes: calculating network resources such as bandwidth and delay required by switches on the path for each flow, and ensuring that the selected path and modality deployment do not exceed resource limits of the switch.

[0022] The present invention also provides an electronic device, comprising a memory and a processor, wherein the memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to execute the program data to implement the optimized deployment method for coexistence of network modes in a multimodal network environment.

[0023] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the program implements the method for optimizing the deployment of network modality coexistence in a multimodal network environment.

[0024] The beneficial effects of the present invention are as follows: the present invention performs well in load balancing, can significantly reduce the link load ratio under high traffic conditions, and maintain it at a low level, effectively avoiding network overload, and improving overall network performance and stability. The RBDR algorithm exhibits excellent adaptability and scalability under different traffic conditions in various network topologies, enabling it to optimize resource usage while maintaining efficient traffic distribution and ensure balanced distribution of link loads. In addition, the RBDR algorithm achieves stability and high performance under larger traffic scales through an efficient link load management strategy. Even if the processing time is relatively long, its overall advantages in load balancing and network performance improvement are still outstanding. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0026] Figure 1 is an example diagram of a multimodal network environment of the present invention;

[0027] Figure 2 It is a special case schematic diagram of the network mode deployment and routing joint optimization problem of the present invention;

[0028] Figure 3 It is a flow chart of the implementation of the RBDR algorithm of the present invention;

[0029] Figure 4 It is a schematic diagram of the device of the present invention. DETAILED DESCRIPTION

[0030] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0031] The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the" and "the" used in the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0032] It should be understood that although the terms first, second, third, etc. may be used in the present invention to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present invention, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0033] The present invention is described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the features of the following embodiments and implementations can be combined with each other.

[0034] like Figure 1 As shown in FIG. 1 , the data plane of the multimodal network is composed of multimodal network elements, including access-level multimodal switches and core-level multimodal switches. Each multimodal network element can carry multiple network modal programs. In a multimodal network environment, network administrators need to deploy network modal programs on multimodal network elements to meet the needs of different network modal terminals. With the continuous growth and diversification of various network applications in different vertical fields, the demand for generating network modalities has become complex and diverse. Due to the complexity of heterogeneous resources, the trade-off between the service quality of business flows and the use of multimodal network resources has become increasingly apparent, resulting in the situation where the service quality is not up to standard or multimodal network resources are wasted, which runs counter to the adaptation requirements of unified physical resources to differentiated upper-layer business performance indicators. Therefore, in the face of the needs of diversified network modalities, how to generate a network-level network modal deployment solution in a multimodal network environment, maximize the fit between the network modal generation requirements and the network resource efficiency, and realize the efficient deployment of network modalities in the multimodal network data plane is an urgent problem to be solved.

[0035] In view of the above problems, the present invention proposes an optimized deployment method for network modality coexistence in a multimodal network environment to solve the problem of joint optimization of network modality deployment and routing (JMDR problem) in the multimodal network data plane. Specifically, given the multimodal network topology and terminal distribution, as well as the traffic set, the JMDR problem includes the following two sub-problems: 1) which network modalities are deployed on each switch to complete the correct parsing and forwarding of data packets; 2) which network modalities are deployed on each switch to complete the correct parsing and forwarding of data packets; Whether it is deployed on a switch The present invention uses binary variables Indicates network mode Whether it is deployed on a switch If the network mode Deployed on a switch superior, ;otherwise, Using binary variables Representation Flow Whether to use path For routing. The routing path is ,but ;otherwise, When jointly optimizing network modality deployment and routing, the following constraints need to be considered:

[0036] Constraint 1: Routing constraint. For any flow For Only one routing path must be selected for forwarding. This constraint ensures that each flow can only select a unique path, thus avoiding the situation where the flow is distributed to multiple paths. In this way, the transmission path selection of the flow in the network is constrained to ensure that each flow has only one transmission channel.

[0037]

[0038] in, Representation Flow Whether to use path For routing, if If 1 is selected, this path is selected for routing. If it is 0, it means that this path is not selected for routing; Indicates the path, Indicates a traffic set.

[0039] Constraint 2: Network mode deployment constraint. If the network mode is flow Select Path As its forwarding path, all switches on the path need to deploy network mode . Change the path The set of switches on This constraint ensures that in the network, only when the flow Some kind of modality is needed And all switches on the path The flow can select a path only when the mode is deployed. If a switch on the path If no modal is deployed, the flow It cannot be transmitted through this path. This constraint ensures that the flow path selection matches the deployment of the network mode, ensuring that the traffic can obtain the required network services and resource support during transmission.

[0040]

[0041] in, Representation Flow Whether the network mode of is m; Indicates whether the network mode m is deployed on the switch v, m represents the network mode, M represents the network mode set; p represents the path, Representation Flow The candidate path set; v represents the switch, represents the set of all switches on path p; Indicates whether network mode m is deployed on switch v. If If it is 1, it means that the network mode is deployed on the switch. If it is 0, it means that the network mode is not deployed on the switch; Indicates the amount of resources consumed when deploying network mode m on a switch; Represents the resource capacity of switch v that can be used to deploy network modes.

[0042] Constraint 3: Switch network mode deployment resource constraints. Suppose that the network mode is deployed on the switch The resources required are When deploying network modes on a switch, the switch should not be exceeded. Resource capacity available for deploying network modalities This constraint limits the switch The total resource consumption of the network modes deployed on the switch cannot exceed the resource capacity of the switch. This ensures that the switch is deployed within its resource capacity limit, prevents resource overload, and ensures the stability and efficient operation of the network.

[0043]

[0044] Where V represents the set of switches.

[0045] Constraint 4: Switch routing configuration resource constraints. When configuring routing, the switch processes the flow The amount of resources required The resource capacity related to the switch routing configuration should not be exceeded This constraint can limit the number of The total bandwidth demand of all flows will not exceed the maximum bandwidth capacity it supports, thus avoiding switch overload and ensuring network operation.

[0046]

[0047] in, Indicates the switch processing flow The amount of resources required; Indicates the resource capacity of switch v that can be used to configure routing.

[0048] Constraint 5: Link resource constraint. This constraint limits the The total flow size of all traffic The bandwidth capacity of the link cannot be exceeded It ensures that the traffic bandwidth demand on each link does not exceed the bandwidth capacity of the link itself, thereby avoiding link congestion and ensuring the stability and efficiency of network transmission.

[0049]

[0050] in, Representation Flow The flow size, represents the bandwidth capacity of link e, e represents the link, and E represents the link set.

[0051] Constraint 6: Forwarding delay constraint. Select a routing path from the candidate paths Each link on Transmission delay and switches Processing delay The sum of the flow Maximum forwarding delay threshold This ensures that the transmission delay of each flow can meet the delay requirements, guaranteeing the quality of service and transmission performance.

[0052]

[0053] in, represents the transmission delay of link e; represents the processing delay of switch v; Representation Flow The transmission delay threshold.

[0054] Optimization goal: Link overload will increase the forwarding delay of data packets and the end-to-end communication delay, which will reduce the user's service quality. Therefore, the optimization goal considered in this invention is to achieve link load balancing. That is, on all links, the traffic load of the most congested link should be reduced as much as possible. By balancing the traffic distribution in the network and avoiding the overload of a single link, the resource utilization efficiency and performance of the entire network can be improved. The constraint goal is formalized as shown in the following formula (7).

[0055]

[0056] In summary, the JMDR problem can be formalized as shown in the following formula (8).

[0057]

[0058] Since the JMDR problem has a minimum-maximum function in the optimization objective, the JMDR problem is not an integer linear programming problem, which increases the difficulty of solving it. Therefore, before designing the solution algorithm, the present invention first converts the original problem into an integer linear programming problem. The present invention defines the load balancing factor , which represents the ratio of link resource consumption to link capacity. The JMDR problem defined in formula (8) is transformed into the following formula (9):

[0059]

[0060] in, Indicates the ratio of link resource consumption to link capacity.

[0061] In response to the JMDR problem, the present invention designs a randomized-base algorithm for network modal deployment and routing selection (RBDR), which can solve the JMDR problem defined in formula (9) in polynomial time. The RBDR algorithm flow is as follows: Figure 3 shown.

[0062] The RBDR algorithm consists of three steps. The first step is to relax the JMDR problem defined in formula (9) by replacing the variable and The range of the value of is relaxed to the continuous interval [0,1], thus obtaining a linear programming. After relaxation, the variable Indicates network mode Deployed on a switch The probability of Representation Flow Select Path as the probability of its routing path. The relaxed formula (9) is a linear programming, so it can be solved using a linear programming solver.

[0063] Suppose the decimal solution of formula (9) after relaxation is and , the integer solution is and In the second step, the RBDR algorithm is based on the fractional solution Construct the integer solution of formula (9) before relaxation , with probability Will Set to 1 to determine the routing selection. The third step is to determine the integer solution based on the second step The value of . The set of switches for the path selected in step 2 is The RBDR algorithm converts the flow Required network modality Deployed on a switch on, and Set to 1. The detailed process of the RBDR algorithm is shown in Algorithm 1.

[0064] Algorithm 1: Network mode deployment and routing selection algorithm RBDR

[0065] Input: Network topology , Traffic Collection And each flow The set of candidate paths

[0066] Output: Variable and Integer solutions of and

[0067] Step 1: Solve the relaxed JMDR problem using a linear programming solver

[0068] The variable and The value range of is relaxed to the continuous interval [0,1], thus constructing a relaxed JMDR problem;

[0069] Solving the relaxed JMDR problem yields a fractional solution { }and{ };

[0070] Step 2: Route selection

[0071] Will{ }Initialize to 0;

[0072] for :

[0073] for :

[0074] By probability Will Set to 1;

[0075] end for

[0076] end for

[0077] Step 3: Network mode deployment

[0078] Will{ }Initialize to 0;

[0079] for :

[0080] Suppose the set of switches on the path selected in step 2 is ;

[0081] for :

[0082] Will flow Required network modality Deployed on a switch superior;

[0083] Will Set to 1;

[0084] end for

[0085] end for

[0086] Since the RBDR algorithm determines the route first and then deploys the network mode, there may be insufficient resources when the network mode is deployed. Therefore, this article improves the RBDR algorithm. The improved RBDR algorithm consists of five steps, and the first three steps are the same as Algorithm 1. Use Indicates the set of overloaded switches. The set of traffic passing through is expressed as For those overloaded switches, the RBDR algorithm is based on The value pair Sort the streams in ascending order. In the flow, the RBDR algorithm selects the candidate path set Find a new routing path with the maximum value, and satisfy the constraints defined by formulas (1)-(6), and update according to the new path and Repeat the above process until the switch No overload. The detailed process of the improved RBDR algorithm is shown in Algorithm 2.

[0087] Algorithm 2: Improved RBDR algorithm

[0088] Input: Network topology , Traffic Collection And each flow The set of candidate paths 、{ }、{ }、{ }、{ }

[0089] Output: Adjusted { }and{ }

[0090] Step 1: Solve the relaxed JMDR problem using a linear programming solver

[0091] The variable and The value range of is relaxed to the continuous interval [0,1], thus constructing a relaxed JMDR problem.

[0092] Solving the relaxed JMDR problem yields a fractional solution { }and{ }.

[0093] Step 2: Route selection

[0094] First, the set { }Initialize to 0 and traverse the traffic set in sequence All traffic in , and then traverse the flow in sequence The corresponding path set All paths in ,by The probability will be Set to 1, that is, The probability selection flow The corresponding path , the traversal ends after successfully finding the path.

[0095] Step 3: Network mode deployment

[0096] First, the set { }Initialize to 0 and traverse the traffic set in sequence All traffic in , set the set of switches on the path selected in step 2 to , and then traverse the switch set in turn Switches in , will flow Required network modality Deployed on a switch On, will Set to 1, that is, in this switch Deploy network mode , the traversal ends after the modal is successfully deployed.

[0097] Step 4: Network mode deployment and routing adjustment

[0098] If an overloaded switch appears, first aggregate the overloaded switches. Initialize to , traverse the switches in the switch set , calculate the network mode deployment resource consumption of switch v , if the network mode deployment resource consumption Greater than the resource capacity of switch v that can be used to configure routing , then the switch Add to overload switch collection Start traversing the overload switch set Switches in ,use Indicates a switch The set of flows on The value pair Sort the flows in ascending order. Traverse the flow collection The flow in the candidate path set Find a new routing path with the maximum value, and satisfy the constraints defined by formulas (1)-(6), and update according to the new path and , if the network mode deployment resource consumption Smaller than the resource capacity of switch v that can be used to configure routing , then the traversal ends and a new network mode is deployed in the newly found switch set.

[0099] In order to analyze the approximate performance of the improved RBDR algorithm, the present invention firstly gives two theorems.

[0100] Theorem 1 (Chernov inequality) gives Independent variables: ,in ,make , we can get ,in is an arbitrary positive value.

[0101] Theorem 2 (Joint Boundary) Given Events: , each event by The probability of occurrence is .

[0102] After random rounding, the improved RBDR algorithm selects a routing path for each flow. The present invention proposes other theorems, which are described as follows:

[0103] Theorem 3: After random rounding, each flow will be assigned only one routing path.

[0104] flow There is a set of candidate paths The probability of each path being selected as a routing path is . So we get:

[0105]

[0106] Candidate path set The total probability of selecting the upper path is:

[0107]

[0108] After random rounding, each flow is assigned only one routing path.

[0109] The smallest network mode deployment resource capacity among all switches is recorded as , the minimum routing configuration resource capacity is recorded as .use Indicates the minimum bandwidth capacity of the link, using Representation Flow The longest forwarding delay among the candidate paths. Define the following constants:

[0110]

[0111] The present invention is in constant Based on this, the performance of network mode deployment resource consumption and link bandwidth load is analyzed.

[0112] Theorem 4: In the solution obtained by the improved RBDR algorithm, the network modal deployment resource consumption on each switch will not exceed the fractional solution times, of which is the number of switches.

[0113] The improved RBDR algorithm first The routing result of each flow is obtained by rounding, and then the network mode is deployed on the selected path to meet the normal processing and forwarding of traffic. Due to the existence of formula (2) and the optimization goal in formula (9) is to minimize , so when solving formula (9), The value will approach , from which we can conclude:

[0114]

[0115] Set flow The resource consumption of deploying the network mode on the switch is ,but The expectation is:

[0116]

[0117] Combining the switch network modal deployment resource constraints in formula (12) and formula (9), we get:

[0118]

[0119] According to the constant defined in formula (12) ,get:

[0120]

[0121] Applying Theorem 2, we can get:

[0122]

[0123] in is an arbitrarily small positive number. Assume that:

[0124]

[0125] in It is a variable related to the size of the network and can be set using a function related to the number of switches. As the network size increases The value of approaches 0. By solving the inequality in formula (18) and setting , is the number of switches in the network, we can get:

[0126]

[0127] Applying Theorem 2 based on formula (18), we get:

[0128]

[0129] According to formula (20), , which means that after random rounding, the network modal deployment resource consumption on each switch will not exceed the fractional solution times.

[0130] Theorem 5: The improved RBDR algorithm can ensure that the load of each link does not exceed the fractional solution times, of which is the number of switches. That is, for link capacity constraints, the improved RBDR algorithm can obtain Approximate ratio.

[0131] The first step of the improved RBDR algorithm will obtain a decimal solution to the relaxed JMDR problem and optimal results By random rounding, each stream Only in the candidate path set Therefore, the flow In the link The load generated on the :

[0132]

[0133] By definition, a random variable , are independent of each other. Link The expectation on can be expressed as:

[0134]

[0135] Combining formula (22) with the constant The definition of , we can get:

[0136]

[0137]

[0138] Applying the Chernoff inequality of Theorem 1, assuming is an arbitrarily small positive number, we get:

[0139] Assume the following inequality holds:

[0140]

[0141] in It has been defined in the proof of Theorem 4. By solving the inequality in formula (25) and setting , we can get:

[0142]

[0143] Finally, applying Theorem 2, we get:

[0144]

[0145] This means that after random rounding, the load on each link will not exceed the fractional solution. times.

[0146] Theorem 6: In the solution obtained by the improved RBDR algorithm, the routing configuration resource consumption on each switch will not exceed the fractional solution. times, the routing delay will not exceed the decimal solution times.

[0147] Since the proof of Theorem 6 is similar to that of Theorems 4 and 5, the proof of Theorem 6 is omitted here.

[0148] Example:

[0149] Given a Figure 2 The network topology shown has two access switches and ,as well as Core switches (denoted as Each access switch is connected to Connected to the access switch The connected terminals are denoted as , and access switches The connected terminals are denoted as ,in . Will access the switch and And core switches The network mode deployment resources and routing configuration resources are set to infinity. In addition, the link bandwidth resources between the terminal and the access switch are set to infinity, and the link bandwidth capacity between the access switch and the core switch is set to a constant. The JMDR problem instance aims to select a routing path for each flow in the network, deploy network modes and configure routing rules on the switches on the corresponding path, and the optimization goal is to achieve load balancing of the links.

[0150] In summary, the optimized deployment method for network mode coexistence in a multimodal network environment proposed by the present invention can effectively solve the problem of limited resources of multimodal network elements. Multimodal network elements are mainly composed of programmable devices. Although they can support the deployment of multiple network modes at the same time, their resources are limited. Taking the domestic Shengke switch

[12] equipped with TsingMa.MX chip as an example, although it uses user-defined fields (UDF) to support the processing of different network mode messages, each port can only be configured with two UDF rules, which makes it difficult to carry a large number of network modes at the same time. The second is to ensure the connectivity of each network mode. In a multimodal network environment, different terminals may have different network mode requirements. Therefore, when deploying network modes, it is necessary to ensure that the deployment scheme of each network mode has connectivity to achieve normal communication of the terminal. Finally, ensure the user service quality. The upper layer services of multimodal networks have differentiated network requirements. For example, industrial Internet and telemedicine have high requirements for network latency, while video services require larger network bandwidth. Therefore, when deploying network modalities, it is necessary to consider how to ensure the service quality of upper-layer applications under the premise of limited network resources.

[0151] Corresponding to the aforementioned embodiment of a method for optimizing deployment of network modal coexistence in a multimodal network environment, the present invention also provides an embodiment of a device for optimizing deployment of network modal coexistence in a multimodal network environment.

[0152] See also Figure 4 An embodiment of the present invention provides an optimized deployment device for network modality coexistence in a multimodal network environment, including one or more processors, which are used to implement an optimized deployment method for network modality coexistence in a multimodal network environment in the above embodiment.

[0153] An embodiment of an optimized deployment device for coexistence of network modes in a multimodal network environment of the present invention can be applied to any device with data processing capabilities, and the device with data processing capabilities can be a device or apparatus such as a computer. The device embodiment can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by the processor of any device with data processing capabilities in which it is located reading the corresponding computer program instructions in the non-volatile memory into the internal memory for execution. From a hardware perspective, if Figure 4 As shown, it is a hardware structure diagram of any device with data processing capability where an optimized deployment device for network mode coexistence in a multimodal network environment of the present invention is located, except Figure 4In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities in which the apparatus in the embodiments is located may also include other hardware, generally based on the actual functions of the device with data processing capabilities, which will not be described in detail.

[0154] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0155] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.

Claims

1. A method for optimizing the deployment of network modality coexistence in a multimodal network environment, characterized in that: The following steps are involved: Input the network topology flow set and the candidate path set for each flow; According to the network topology flow set and the candidate path set, a relaxation algorithm is used to process the network modal deployment and path selection problem, and the value range of the path selection variable and the modal deployment variable is relaxed from 0 and 1 to the continuous interval [0, 1], allowing each flow to probabilistically select a path and modal deployment in the interval [0, 1], and constructing a relaxed JMDR problem; using a linear programming solver to solve the relaxed JMDR problem, and obtaining a decimal solution of the path selection variable and the modal deployment variable; Based on the decimal solution, a path is randomly selected, and the path selection variables and modal deployment variables are rounded according to probability to determine the specific transmission path of each flow and the modal deployment on the switch; the resource consumption on the path selected for each flow is calculated, and the optimality of the selected path and modal deployment in resource utilization is ensured.

2. According to claim 1, the method for optimizing the deployment of network modality coexistence in a multimodal network environment is characterized in that: The linear programming solver is a Gurobi or CPLEX tool, which is used to solve the relaxed JMDR problem and output the optimal solution for each flow path and mode selection.

3. The method for optimizing the deployment of network modality coexistence in a multimodal network environment according to claim 1, characterized in that: The use of the relaxation algorithm to process the network mode deployment and path selection problem is specifically as follows: Will represent the network mode Whether deployed on a switch Variables Heliu Select path Variables The range of values ​​of is relaxed to the continuous interval [0, 1], and a linear programming is obtained; among them, the relaxed variable Indicates network mode Deployed on a switch The probability of Representation Flow Select Path as the probability of its routing path; Use a linear programming solver to solve the linear programming and define the load balancing factor , the linear programming is expressed as the following optimization objective: in, Representation Flow Whether to use path For routing, if If 1 is selected, this path is selected for routing. If it is 0, it means that this path is not selected for routing; Indicates the path, represents the flow set, Representation Flow Whether the network mode of is m; Indicates whether the network mode m is deployed on the switch v, m represents the network mode, M represents the network mode set; p represents the path, Representation Flow The candidate path set; v represents a switch, V represents a switch set, represents the set of all switches on path p; Indicates whether network mode m is deployed on switch v. If If it is 1, it means that the network mode is deployed on the switch. If it is 0, it means that the network mode is not deployed on the switch; Indicates the amount of resources consumed when deploying network mode m on a switch; represents the resource capacity of switch v that can be used to deploy network modes; Indicates the switch processing flow The amount of resources required; Indicates the resource capacity of switch v that can be used to configure routing; Representation Flow The flow size, represents the broadband capacity of link e, e represents the link, E represents the link set, represents the transmission delay of link e; represents the processing delay of switch v; Representation Flow The transmission delay threshold of Indicates the ratio of link resource consumption to link capacity; The decimal solution of the optimization objective is obtained by calculation: and , the integer solution is and ; Based on decimal solution Construct an integer solution to the optimization objective before relaxation , with probability Will Set to 1 to determine the routing selection; set the flow The path you choose The switch set is ; Based on decimal solution Construct an integer solution to the optimization objective before relaxation The value of Required network modality Deployed on a switch on, and Set to 1.

4. The method for optimizing the deployment of network modality coexistence in a multimodal network environment according to claim 3, characterized in that: The method of using a relaxation algorithm to process the network mode deployment and path selection problem also includes: if an overloaded switch is found, the overloaded switch set Initialize to an empty set; traverse and calculate each switch Resource consumption of network mode deployment ; If resource consumption If the preset resource limit is exceeded, the switch Add to overload switch collection in; use Indicates a switch The flow collection on , and according to the flow Modal deployment variables right The flows in are sorted in ascending order; for each flow , from the candidate path set A new routing path is selected from the value, and satisfy all constraints; update according to the new routing path and The value of , after updating, recalculate the resource consumption of the path ; If the resource consumption at this time Less than the preset resource limit , then exit the loop and solve the switch overload problem; otherwise, loop the above steps until a suitable path is found.

5. The method for optimizing the deployment of network modality coexistence in a multimodal network environment according to claim 1, characterized in that: The randomly selecting a path includes: randomly selecting an optimal path for each flow based on a fractional solution of a path selection variable by a probability distribution method.

6. The method for optimizing the deployment of network modality coexistence in a multimodal network environment according to claim 1, characterized in that: The modality deployment includes: deploying a network modality for each switch on the selected path of each flow and ensuring that the switch has sufficient resources to support the transmission requirements of the flow.

7. The method for optimizing the deployment of network modality coexistence in a multimodal network environment according to claim 1, characterized in that: The method further includes: when deploying the network mode, calculating the resource consumption of each switch and determining whether it exceeds a predetermined resource limit. If the resource consumption exceeds the limit, reselecting the path and adjusting the mode deployment to ensure efficient use of network resources.

8. The method for optimizing the deployment of network modality coexistence in a multimodal network environment according to claim 1, characterized in that: The calculating of the resource consumption on the path selected by each flow includes: calculating the network resources required by the switches on the path for each flow, and ensuring that the selected path and modality deployment do not exceed the resource limit of the switch.

9. An electronic device, comprising a memory and a processor, characterized in that: The memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to execute the program data to implement an optimized deployment method for network modality coexistence in a multimodal network environment as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by the processor, it implements the optimized deployment method for coexistence of network modes in a multimodal network environment as described in any one of claims 1-8.

Images