Dynamic bandwidth allocation method and system in SDN (Software Defined Network) environment
By treating aggregated physical links as virtual links and constructing a linear programming model in the SDN environment, the problems of low resource utilization and local overload in bandwidth allocation in the SDN environment are solved, achieving efficient dynamic adjustment and global optimization, and reducing hardware costs.
Patent Information
- Application Number
- CN202511917845.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-18
- Publication Date
- 2026-02-13
AI Technical Summary
Traditional network bandwidth allocation methods suffer from low resource utilization and local overload when faced with dynamically changing traffic demands, especially in SDN environments where there is a lack of efficient dynamic adjustment and global optimization solutions.
In an SDN environment, multiple physical links with the same source node are aggregated into virtual links using a linear programming model. Combined with the traffic demand matrix, an optimal bandwidth allocation scheme is constructed. The SDN controller generates flow table rules for traffic forwarding, thereby achieving the global optimal allocation of network resources.
It improves network resource utilization, alleviates local congestion, enhances overall load balancing capabilities, reduces hardware costs, and supports adaptive response to real-time traffic demands.
Smart Images
Figure CN121530933A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of bandwidth allocation technology, and in particular to a dynamic bandwidth allocation method and system in an SDN environment. Background Technology
[0002] With the rapid development of emerging technologies such as cloud computing, big data, and artificial intelligence, network traffic is experiencing explosive growth, and traffic patterns are becoming increasingly complex and varied. Traditional network bandwidth allocation methods mainly employ static configuration, pre-allocating fixed bandwidth resources to network links. This approach exhibits significant limitations when facing dynamically changing traffic demands. When network traffic is unevenly distributed in time and space, fixed bandwidth allocation often leads to severe overload of some link resources while other link resources remain largely idle, resulting in low overall network resource utilization.
[0003] To address this issue, dynamic bandwidth allocation technology has emerged. This technology aims to dynamically adjust the bandwidth resource allocation of network links based on real-time traffic demands, thereby optimizing network resource utilization and improving network performance. Currently, dynamic bandwidth allocation technology is mainly applied to reconfigurable network environments, such as optical networks built using optical cross-connects (OCS), or programmable networks implemented through software-defined networking (SDN) technology.
[0004] The reconfigurable network bandwidth allocation scheme based on OCS is mainly implemented in the following way: First, it collects traffic demand information among nodes in the network. Then, using mathematical optimization methods such as linear programming, it allocates bandwidth resources to logical links in the network under the constraint of a fixed total bandwidth budget. Simultaneously, this scheme typically employs heuristic algorithms such as shortest path or minimum cost path to plan routing paths for data flows. In practical applications, OCS devices can achieve dynamic reconfiguration of physical links through optical switch arrays, thereby supporting dynamic adjustment of logical link bandwidth. Although the dynamic bandwidth allocation scheme based on OCS can achieve flexible adjustment of link bandwidth, it has the following significant drawbacks: First, OCS devices are expensive, and their deployment and maintenance are difficult, limiting the application of this scheme in large-scale networks; second, the reconfiguration process of OCS usually takes milliseconds or even longer, which cannot meet the rapidly changing traffic demands; finally, the bandwidth adjustment accuracy and granularity of OCS devices are also limited by the characteristics of the physical components.
[0005] In an SDN environment, the network control plane and data plane are separated, allowing network administrators to flexibly manage and configure network resources through a centralized controller. While SDN technology provides a more convenient way to implement dynamic bandwidth allocation, how to efficiently achieve dynamic bandwidth adjustment and simultaneously optimize traffic routing in an SDN environment remains a pressing technical challenge.
[0006] Traditional bandwidth management solutions in SDN environments primarily employ the following techniques: These solutions collect network status information, including link utilization and port status, through the SDN controller. Then, based on preset policies or simple load balancing algorithms, they manage network traffic by issuing flow table rules. Common implementations include static QoS configuration based on the OpenFlow protocol and link aggregation technology. However, traditional SDN bandwidth management solutions have the following shortcomings: First, static QoS configuration cannot dynamically adjust bandwidth allocation according to real-time traffic demands. Second, while link aggregation technology provides some load balancing capabilities, it lacks a global resource optimization perspective and cannot accurately allocate bandwidth for specific traffic needs. Third, these solutions typically do not consider the overall network bandwidth budget constraints; traffic routing is usually based on simple hash algorithms or shortest path algorithms, making it difficult to achieve optimal resource allocation under limited resource conditions. Summary of the Invention
[0007] In view of the shortcomings of the existing technology, the present invention provides a dynamic bandwidth allocation method and system in an SDN environment, which solves the existing problems.
[0008] The present invention adopts the following technical solution: In a first aspect, the present invention provides a dynamic bandwidth allocation method in an SDN environment, comprising the following steps: Network topology information is obtained through the SDN controller, and the network topology information includes a set of network nodes and a set of physical links. Multiple physical links with the same source and destination nodes are logically aggregated to form a virtual link. The virtual link is responsible for dynamically adjusting the bandwidth of the corresponding multiple physical links. The SDN controller collects the real-time traffic demands of each node pair in the network and constructs a traffic demand matrix; where the source node and the destination node are a node pair. The linear programming model is solved based on multiple virtual links and the traffic demand matrix to obtain the optimal bandwidth allocation scheme. The linear programming model takes minimizing the maximum utilization rate of all physical links in the network as the objective function, and uses total bandwidth budget constraints, link bandwidth range constraints, link capacity constraints, traffic conservation constraints, demand satisfaction constraints, and maximum link utilization constraints as constraints.
[0009] Preferably, the linear programming model is as follows: ; ; ; ; ; ; in, This represents the maximum utilization of all links in the network. Indicates allocation to link bandwidth, This is the total bandwidth budget, where E represents all links in the network. and These are links Minimum and maximum bandwidth limits, Represents a stream In the link Bandwidth used Represents the set of all non-zero flows. It is a flow The required bandwidth, It is a positive number.
[0010] Preferably, the minimum bandwidth of the virtual link is the minimum bandwidth of a single physical link within the group, and the maximum bandwidth is the sum of the bandwidths of all physical links within the group.
[0011] Preferably, based on the optimal bandwidth allocation scheme, the actual activation status of the physical link is determined by the activation ratio, as shown below: ; in, It's the activation ratio. It is the bandwidth allocated by the virtual link. It represents the total bandwidth of the corresponding physical link group.
[0012] Preferred options also include: Generate corresponding SDN flow table rules based on the optimal bandwidth allocation scheme; The SDN controller distributes SDN flow table rules to switches in the network, and traffic is forwarded according to the bandwidth allocation results.
[0013] Secondly, the present invention provides a dynamic bandwidth allocation system in an SDN environment, comprising: The acquisition module is used to acquire network topology information through the SDN controller, wherein the network topology information includes a set of network nodes and a set of physical links; The aggregation module is used to logically aggregate multiple physical links with the same source and target nodes to form a virtual link. The virtual link is responsible for dynamically adjusting the bandwidth of the corresponding multiple physical links. The building module is used to collect real-time traffic demands from each node pair in the network through the SDN controller and build a traffic demand matrix; where the source node and the target node are a node pair. The solution module is used to solve the linear programming model based on multiple virtual links and the traffic demand matrix to obtain the optimal bandwidth allocation scheme. The linear programming model takes minimizing the maximum utilization rate of all physical links in the network as the objective function, and uses total bandwidth budget constraints, link bandwidth range constraints, link capacity constraints, traffic conservation constraints, demand satisfaction constraints, and maximum link utilization constraints as constraints.
[0014] Compared with the prior art, the above-mentioned at least one technical solution adopted by the present invention can achieve the following beneficial effects: This invention first logically aggregates multiple physical links with identical source and destination nodes to form a virtual link that can dynamically adjust the bandwidth allocated to these physical links. A linear programming model is then constructed with the objective function of minimizing the maximum utilization of all physical links in the network. By solving the linear programming model using a traffic demand matrix constructed from real-time traffic and the virtual link, the system can achieve globally optimal allocation of network resources under bandwidth budget constraints, effectively improving network resource utilization, alleviating local congestion, and enhancing overall load balancing capabilities. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 This is a flowchart of a dynamic bandwidth allocation method in an SDN environment according to the present invention. Detailed Implementation
[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] Example 1 This invention provides a dynamic bandwidth allocation method in an SDN environment. It optimizes network resource allocation through a linear programming model, solving the problems of insufficient flexibility and low resource utilization in traditional static bandwidth allocation. The system employs software abstraction and centralized control to achieve efficient dynamic adjustment of network bandwidth. (Refer to...) Figure 1 Specifically, it includes the following steps:
[0019] S1: Topology discovery.
[0020] The controller first discovers the network topology and obtains physical link and node information through various methods (such as manual configuration import, SDN, etc.).
[0021] The SDN controller, as the core component of the entire system, is responsible for coordinating the execution of the entire dynamic bandwidth allocation process. The controller adopts a modular design, and its main functions include network topology discovery, link aggregation, traffic monitoring, bandwidth allocation calculation, and flow table rule generation and distribution.
[0022] S2: Physical link aggregation and virtual link abstraction.
[0023] Multiple physical links with the same endpoints are logically aggregated into a single virtual link, and its bandwidth range is calculated.
[0024] To enable dynamic bandwidth adjustment in traditional SDN environments without optical cross-connectors (OCS), the system introduces physical link aggregation and virtual link abstraction technologies. This technology aggregates multiple physical links with identical endpoints into a single virtual link, and by controlling the activation state of the physical links, dynamically adjusts the bandwidth of the virtual link.
[0025] The physical link aggregation process first groups physical links with the same source and destination nodes, and then creates a corresponding virtual link for each group. The bandwidth range of the virtual link is determined by the bandwidth of the physical links within the group, with a minimum bandwidth... The minimum bandwidth and maximum bandwidth of a single link within the group. The sum of bandwidth for all links within the group is calculated using the following formula:
[0026] ; ; in, Indicates virtual link The corresponding set of physical links, Indicates physical link l The bandwidth.
[0027] After the bandwidth allocation results are determined, the system calculates the activation ratio to determine the actual activation status of the physical links. The formula for calculating the activation ratio is:
[0028] ; in, It's the activation ratio. It is the bandwidth allocated by the virtual link. This is the total bandwidth of the corresponding physical link group. When the activation ratio is greater than the preset threshold (default 0.1), the link is considered to be in an active state.
[0029] S3: Obtain the traffic matrix.
[0030] Traffic monitoring: The controller collects traffic demands between nodes in the network and constructs a traffic demand matrix. Alternatively, machine learning can be used to predict the traffic demand matrix, potentially enabling faster bandwidth adjustment responses.
[0031] S4: Linear programming model calculation.
[0032] The core of dynamic bandwidth allocation is to construct and solve a linear programming model to achieve optimal allocation of network resources. The design goal of the model is to minimize the maximum utilization of all physical links in the network, thereby achieving global load balancing.
[0033] The objective function of the model aims to minimize the maximum link utilization in the network, i.e.: ; in, This represents the maximum utilization of all links in the network.
[0034] The constraints include total bandwidth budget constraints, link bandwidth range constraints, link capacity constraints, traffic conservation constraints, demand satisfaction constraints, and maximum link utilization constraints.
[0035] Total bandwidth budget constraint: The total bandwidth allocated to all links shall not exceed the preset total bandwidth budget. ; in, Indicates allocation to link bandwidth, This is the total bandwidth budget.
[0036] Link bandwidth range constraints: The bandwidth allocation for each link must be within its minimum and maximum bandwidth limits. ; in, and These are links Minimum and maximum bandwidth limits.
[0037] Link capacity constraint: The total bandwidth occupied by all data flows through a link cannot exceed the allocated bandwidth of that link. ; in, Represents a stream In the link Bandwidth used This represents the set of all non-zero flows.
[0038] Demand must satisfy the constraint: the sum of the bandwidth allocated to each flow across all paths must equal its flow demand. ; in, It is a flow The required bandwidth.
[0039] Maximum link utilization constraint: The utilization of any link must not exceed the maximum utilization variable. ; in, It is a small positive number used to avoid division by zero errors.
[0040] Based on multiple virtual links and a traffic demand matrix, a linear programming model is used to solve for the optimal bandwidth allocation scheme. The linear programming optimization model simultaneously optimizes virtual link bandwidth allocation and data flow routing.
[0041] Physical link activation / deactivation policy: Dynamically determine the activation status and bandwidth allocation of physical links based on the virtual link bandwidth allocation results.
[0042] S5: Generate and distribute flow tables.
[0043] Flow table rule generation and distribution: Based on the bandwidth allocation results, corresponding SDN flow table rules are generated and distributed to the switching devices in the network via the SDN controller API. Flow table rules contain matching conditions and action instructions, used to control the forwarding path and bandwidth of traffic.
[0044] Periodic adjustment: The system periodically collects network status and traffic data through independent monitoring and allocation threads, and recalculates the bandwidth allocation scheme to achieve dynamic adjustment.
[0045] The system is implemented in Python, utilizing the Gurobi optimizer to handle linear programming problems and the NumPy library for matrix operations and data processing. The linear programming model is solved using efficient algorithms provided by Gurobi, enabling the solution of bandwidth allocation problems in large-scale networks within a reasonable timeframe.
[0046] The system maintains detailed link status information, including current bandwidth usage, maximum bandwidth, and utilization rate. When a high-utilization link is detected (utilization exceeding 80%), the system issues a warning and attempts to adjust bandwidth allocation in the next allocation cycle to alleviate network congestion. For ease of debugging and monitoring, the system also implements comprehensive logging capabilities, recording network status changes, bandwidth allocation results, and abnormal events. Furthermore, the system supports generating detailed network status reports, providing a holistic view of network operation and assisting network administrators in decision-making and optimization.
[0047] This invention aims to provide a method and system for achieving efficient dynamic bandwidth allocation in an SDN environment without expensive hardware, reducing the hardware cost of dynamic bandwidth allocation and avoiding reliance on costly OCS devices. It achieves coordinated optimization of bandwidth allocation and routing planning, improving resource allocation efficiency. It maximizes network resource utilization within the total bandwidth budget constraint. It provides a flexible physical link management mechanism, supporting dynamic activation / deactivation. It enables adaptive response to changes in real-time traffic demand.
[0048] Example 2 Based on the same concept, the present invention also provides a dynamic bandwidth allocation system in an SDN environment, including an acquisition module, an aggregation module, a construction module, and a solution module.
[0049] The acquisition module is used to obtain network topology information through the SDN controller. The network topology information includes the set of network nodes and the set of physical links.
[0050] The aggregation module is used to logically aggregate multiple physical links with the same source and target nodes to form a virtual link. The virtual link is responsible for dynamically adjusting the bandwidth of the corresponding multiple physical links.
[0051] The building module is used to collect real-time traffic demands from each node pair in the network through the SDN controller and build a traffic demand matrix; where the source node and the target node are a node pair.
[0052] The solution module is used to solve the linear programming model based on multiple virtual links and traffic demand matrix to obtain the optimal bandwidth allocation scheme. The linear programming model takes minimizing the maximum utilization of all physical links in the network as the objective function, and uses total bandwidth budget constraints, link bandwidth range constraints, link capacity constraints, traffic conservation constraints, demand satisfaction constraints, and maximum link utilization constraints as constraints.
[0053] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0054] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A dynamic bandwidth allocation method in an SDN environment, characterized in that, Includes the following steps: Network topology information is obtained through the SDN controller, and the network topology information includes a set of network nodes and a set of physical links. Multiple physical links with the same source and destination nodes are logically aggregated to form a virtual link. The virtual link is responsible for dynamically adjusting the bandwidth of the corresponding multiple physical links. The SDN controller collects the real-time traffic demands of each node pair in the network and constructs a traffic demand matrix; where the source node and the destination node are a node pair. The linear programming model is solved based on multiple virtual links and the traffic demand matrix to obtain the optimal bandwidth allocation scheme. The linear programming model takes minimizing the maximum utilization rate of all physical links in the network as the objective function, and uses total bandwidth budget constraints, link bandwidth range constraints, link capacity constraints, traffic conservation constraints, demand satisfaction constraints, and maximum link utilization constraints as constraints.
2. The dynamic bandwidth allocation method in an SDN environment as described in claim 1, characterized in that, The linear programming model is shown below: ； ； ； ； ； ； in, This represents the maximum utilization of all links in the network. Indicates allocation to link bandwidth, This is the total bandwidth budget, where E represents all links in the network. and These are links Minimum and maximum bandwidth limits, Represents a stream In the link Bandwidth used Represents the set of all non-zero flows. It is a flow The required bandwidth, It is a positive number.
3. The dynamic bandwidth allocation method in an SDN environment as described in claim 1, characterized in that, The minimum bandwidth of the virtual link is the minimum bandwidth of a single physical link within the group, and the maximum bandwidth is the sum of the bandwidths of all physical links within the group.
4. The dynamic bandwidth allocation method in an SDN environment as described in claim 1, characterized in that, Based on the optimal bandwidth allocation scheme, the actual activation status of the physical link is determined by the activation ratio, as shown below: ； in, It's the activation ratio. It is the bandwidth allocated by the virtual link. It represents the total bandwidth of the corresponding physical link group.
5. The dynamic bandwidth allocation method in an SDN environment as described in claim 1, characterized in that, Also includes: Generate corresponding SDN flow table rules based on the optimal bandwidth allocation scheme; The SDN controller distributes SDN flow table rules to switches in the network, and traffic is forwarded according to the bandwidth allocation results.
6. A dynamic bandwidth allocation system in an SDN environment, characterized in that, include: The acquisition module is used to acquire network topology information through the SDN controller, wherein the network topology information includes a set of network nodes and a set of physical links; The aggregation module is used to logically aggregate multiple physical links with the same source and target nodes to form a virtual link. The virtual link is responsible for dynamically adjusting the bandwidth of the corresponding multiple physical links. The building module is used to collect real-time traffic demands from each node pair in the network through the SDN controller and build a traffic demand matrix; where the source node and the target node are a node pair. The solution module is used to solve the linear programming model based on multiple virtual links and the traffic demand matrix to obtain the optimal bandwidth allocation scheme. The linear programming model takes minimizing the maximum utilization rate of all physical links in the network as the objective function, and uses total bandwidth budget constraints, link bandwidth range constraints, link capacity constraints, traffic conservation constraints, demand satisfaction constraints, and maximum link utilization constraints as constraints.