Computer-implemented method for configuring a hierarchical network architecture

Abstract

The present invention relates to a computer-implemented method for configuring a hierarchical network architecture, wherein the network architecture (20) includes at least a first layer and a second layer, wherein the first layer includes a first controller, and wherein the second layer includes a plurality of subnets and includes a second controller for each subnet, wherein the method includes the steps of: determining an end-to-end configuration for the first layer by the first controller (S10); allocating a resource budget for each subnet according to the end-to-end configuration by the first controller (S12); communicating the allocated resource budget to each second controller by the first controller (S14); and creating a resource plan for the subnet by each second controller, and configuring the network architecture when the resource budgets transmitted to the second controllers are sufficient (20, S18), or transmitting a response from the second controllers to the first controller when the resource budgets transmitted to the second controllers are insufficient (S20).

Description

Technical Field

[0001] This invention relates to a computer implementation method for configuring a hierarchical network architecture. Background Technology

[0002] Deterministic Networking (DetNet) is a technology specifically developed to meet the demands of industrial communication. In the networked factories of Industry 4.0, reliable and time-critical data transmission plays a crucial role as production processes become increasingly automated and networked. DetNet ensures that data packets are transmitted with predictable latency, high reliability, and without loss, thereby ensuring optimal interaction between machines and control systems.

[0003] DetNet's technical implementation is based on a combination of various components and methods. The core elements are deterministic queuing mechanisms, time synchronization, and precise resource allocation within the network. Techniques such as Time-Sensitive Networking (TSN) and reserved bandwidth ensure that critical data flows are not affected by unpredictable latency or network overload. Furthermore, redundant paths and forwarding mechanisms enable enhanced fault tolerance, ensuring uninterrupted communication even in the event of connection failures.

[0004] The primary benefits of DetNet lie in industrial applications where real-time communication is essential. In smart factories, robots, machines, and control systems must work synchronously to execute production processes efficiently and without errors. DetNet enables smooth and precise control, particularly in areas such as process automation, control of autonomous vehicles in production workshops, or real-time monitoring of critical systems. Through deterministic data transmission, production failures can be avoided, and the efficiency of networked systems can be improved, ultimately leading to improved product quality and lower operating costs.

[0005] Using DetNet and TSN simultaneously within a network architecture presents various challenges, primarily stemming from the different concepts and use cases of the two technologies. TSN is mainly based on ensuring time-critical communication at the data link layer and achieving deterministic transmission within a local area network, while DetNet focuses on end-to-end guarantees for time-limited data streams across larger, often heterogeneous networks. Integrating the two schemes requires precise coordination of mechanisms for time synchronization, resource reservation, and error handling, as inconsistencies in QoS guarantees can otherwise arise.

[0006] A core issue lies in the interoperability of time control mechanisms. TSN relies on precise time synchronization within a local area; however, when packets are forwarded through the DetNet backbone, which can have variable latency, delays or jitter can occur, making it difficult to adhere to strict time presets. Furthermore, TSN uses specific mechanisms for prioritizing and buffering data streams, which are not always seamlessly harmonized with the methods used by DetNet for bandwidth reservation and packet forwarding. Misconfiguration or insufficient coordination of these techniques can lead to a loss of deterministic characteristics, thereby impacting the reliability of the entire system.

[0007] Another issue arises due to differing network resource requirements. TSNs typically operate within separate segments of industrial networks, reserving bandwidth and buffers there, while DetNets must scale across larger network infrastructures, leading to increased administrative overhead and more complex routing decisions. This necessitates continuous orchestration of all participating network nodes to ensure critical data flows are delivered with guaranteed latency and without packet loss. Incorrect or inefficient reservation strategies in this application can create bottlenecks or even disrupt less prioritized data flows.

[0008] Additionally, error handling presents a challenge because TSN and DetNet employ different redundancy and recovery mechanisms. TSN relies on mechanisms such as Frame Replication and Elimination for Reliability (FRER), while DetNet relies on a combination of redundant paths and explicit error correction. If these methods are not optimally coordinated, unintended packet loss or inefficient use of network resources can occur. Therefore, successful coexistence of TSN and DetNet requires a well-thought-out architecture that integrates both technologies, preserving their respective advantages without their mechanisms interfering with each other.

[0009] In an environment where TSN and DetNet coexist, it is disadvantageous if TSN and DetNet are configured separately by their respective control devices without knowledge of the additional technologies used. This can lead to undesirable effects and poor performance. Therefore, the objective of this invention is to propose a method for configuring a co-architecture consisting of DetNet and TSN, which eliminates the aforementioned disadvantages and problems.

[0010] Therefore, this invention proposes a common controller that organizes the entire end-to-end QoS.

[0011] This task is addressed through the subject matter of the independent claims. Summary of the Invention

[0012] According to a first aspect of the invention, the task is solved by a computer-implemented method for configuring a hierarchical network architecture (20), wherein the network architecture (20) includes at least a first layer and a second layer, wherein the first layer includes a first controller, and wherein the second layer includes a plurality of subnets and a second controller for each subnet, wherein the method includes the following steps: - Determine the end-to-end configuration for the first layer using the first controller (S10). - The resource budget is allocated for each subnet by the first controller according to the end-to-end configuration (S12). - The allocated resource budget is communicated to each of the second controllers via the first controller (S14); and - Create a resource plan for the subnet through each second controller, and configure the network architecture when the resource budget transmitted to the second controller is sufficient (20, S18), or transmit a response from the second controller to the first controller when the resource budget transmitted to the second controller is insufficient (S20).

[0013] According to the invention, the first controller organizes the entire end-to-end configuration. Depending on the network architecture, this configuration may already be known to the first controller; otherwise, the first controller must first detect the configuration.

[0014] Explicit end-to-end paths are required in DetNet to provide stable forwarding services and ensure that DetNet services are unaffected when the network topology changes. To implement explicit paths in the DetNet architecture, they must meet the application's Service Level Agreement (SLA) requirements, which include limited bandwidth, maximum end-to-end latency, maximum end-to-end latency variation, and maximum packet loss rate. In distributed network systems, an IGP with Constrained Shortest Path First (CSPF) can be used to compute a set of feasible paths for DetNet services. In centralized network systems, the controller can compute paths that meet DetNet requirements based on network information collected within the DetNet domain.

[0015] If the configuration is known, allocate a resource budget for each second controller in the second layer to its corresponding subnet.

[0016] There are various types of resources in the network that can be allocated to a second controller, such as packet processing resources, buffer resources, and output port bandwidth. The network resources required by a specific DetNet service are determined by SLA requirements and network capacity.

[0017] In addition, the second layer may include multiple subnets arranged in parallel within the hierarchical network architecture, but not directly connected to each other in a communicative manner.

[0018] The resource budget allocated to the second controller is transmitted from the first controller to the corresponding second controller. Each second controller then uses its allocated resource budget to determine a resource plan for its individual subnet. The resource plan should be recalculated when the network topology, application traffic, or requirements change. If the resource budget allocated to the second controller is sufficient and a resource plan has been created for each subnet, the hierarchical network architecture is configured. If necessary, the second controller can be configured to transmit a corresponding response to the first controller after successful resource plan creation, allowing the first controller to view the network architecture as configured.

[0019] If the resource budget of the second controller is insufficient for its corresponding subnet, the relevant second controller will transmit a response to the first controller.

[0020] The first controller can coordinate multiple second controllers in this way, each responsible for its corresponding subnet. This joint use of the TSN and DetNet layers can be advantageous in scenarios with high requirements for network determinism. Using the aforementioned method, a configuration for the architecture of the DetNet controller is presented, in which the joint configuration of the TSN and DetNet layers is handled. In this way, end-to-end determinism can be guaranteed even in more complex networks with multiple subnets. This solves the objective of the present invention.

[0021] In one implementation, the first layer is a DetNet layer, and the first controller is a DetNet controller.

[0022] In one implementation, the second layer is a TSN layer, and the second controller is a TSN controller.

[0023] In one implementation, the method is repeated when the resource budget for one of the second controllers is insufficient.

[0024] The first controller can reallocate the resource budget based on the response of at least one second controller, where prior allocation is insufficient in at least one case.

[0025] In one implementation, the response includes an inquiry into a larger resource budget, particularly into the minimum size of the resource budget.

[0026] By querying for larger resource budgets, the first controller can more efficiently reallocate resource budgets. In particular, when the response includes a minimum size of the resource budget, the first controller can directly address the corresponding requirements of the second controller, whose subnet requires a higher resource budget.

[0027] In one implementation, the response includes a description of the actual time budget required, wherein the response is also transmitted to the first controller even if the resource budget allocated to the corresponding second controller is sufficient.

[0028] If the second controller transmits the actual required resource budget to the first controller, the first controller can optimize resource allocation. If, even if the resource budget allocated to the second controller exceeds the requirements of the corresponding subnet, a response is also transmitted to the first controller, then the resources that are not needed can be deducted from the second controller.

[0029] In one implementation, when the resource budget for one of the second controllers is insufficient, data traffic within the network for one or more second controllers is restricted, wherein a higher resource budget is allocated to the unrestricted second controllers.

[0030] If the resources available for the entire network are insufficient to configure all subnets, priority can be given to each subnet or the application running therein, so that the resource budget required by each second controller is limited to a minimum, or even provided entirely to the remaining second controllers.

[0031] In one implementation, the following steps occur before the end-to-end configuration for the first layer is determined: - Identify the network topology and possible resource-bound entities in the first layer.

[0032] Resource-bound entities can be data streams, particularly video or audio streams, but they can also be continuously detected measurement data. All continuously existing data streams known in advance in the network can be taken into account when allocating resource budgets. This can make allocation more efficient.

[0033] In one implementation, the first controller includes a module for identifying the network topology and possible resource-bound entities in the first layer.

[0034] The first controller may include multiple modules, each performing different tasks. During the configuration of the second controller by the configuration engine, its own modules may exist to identify application and resource-bound entities within the network architecture.

[0035] In one implementation, the first layer is placed on top of the second layer.

[0036] In this implementation, the first controller already has an end-to-end network view. Therefore, in this case, no additional information exchange is required between the first and second controllers during the setup phase.

[0037] In one implementation, the second layer is placed on top of the first layer, and the following steps occur before the end-to-end configuration is determined: - Transmitting subnet information from each second controller to the first controller, wherein the subnet information includes: - Regarding the data traffic requirements within the subnet; and - The topology of the subnet.

[0038] In this implementation where the second layer is placed above the first layer, the second controller must inform the first controller of the data traffic in its subnet and its corresponding requirements. Furthermore, the second controller must specify the DetNet topology within its layer, meaning that the second controller specifies the gateway nodes in the DetNet layer.

[0039] In another implementation, the second controller may also designate preferred gateway nodes for existing TSN endpoints within its subnet. Here, the second controller considers the topology and performance within the respective subnet. In DetNet terminology, these gateway nodes are referred to as DetNet edge nodes.

[0040] In one implementation, the resource budget is specifically a time budget for transmission.

[0041] For example, a fixed bandwidth can be assigned to each second controller within the network architecture, so that each terminal device must transmit its data within a defined time window.

[0042] In one implementation, the network architecture maps to a network within an industrial facility, and the subnet includes multiple control units of equipment for machines, parts of machines, or groups of machines.

[0043] In this implementation, the first layer may represent the topmost or upper-level network layer, which represents the interface with an external network, particularly an intranet or the Internet. The second layer is the layer of equipment and machines within the industrial facility. Here, each subnet may be multiple (control) devices of a machine, a part of a complex machine, or a group of machines, which are interconnected within the subnet.

[0044] Therefore, industrial facilities can correspond to factories, production workshops or production lines, or individual facilities.

[0045] In another aspect, the present invention relates to a network architecture for industrial facilities, wherein the network architecture includes a first layer and a second layer. The first layer and the second layer include means for performing the methods as described above.

[0046] In another aspect, the present invention relates to a computer program having program code for performing the method as described above when the computer program is executed on a computer.

[0047] In another aspect, the present invention relates to a computer-readable data carrier having program code of a computer program for performing the methods described above when the computer program is executed on a computer.

[0048] In summary, this invention describes a computer-implemented method for configuring a hierarchical network architecture, a network architecture, a computer program with program code, and a computer-readable data carrier.

[0049] The described design and improvement schemes can be combined with each other arbitrarily.

[0050] Other possible designs, improvements, and implementations of the present invention include combinations of features of the invention not explicitly mentioned in the preceding or subsequent descriptions of the embodiments. Attached Figure Description

[0051] The accompanying drawings should facilitate a further understanding of embodiments of the invention. The drawings illustrate embodiments and, in conjunction with the description, are used to explain the principles and concepts of the invention.

[0052] Taking into account the accompanying drawings, many advantages among other embodiments and the mentioned advantages are apparent. The elements shown in the drawings are not necessarily shown to scale.

[0053] Figure 1 The arrangement of the DetNet layer and the TSN layer is illustrated schematically. Figure 2 The diagram schematically illustrates communication within a network architecture according to an embodiment of the present invention. Figure 3The structure of a network architecture according to an embodiment of the present invention is schematically illustrated; and Figure 4 The flow of a method according to one embodiment is illustrated schematically.

[0054] In the accompanying drawings, unless otherwise stated, the same reference numerals denote the same or functionally identical elements, components, or assemblies. Detailed Implementation

[0055] Figure 1 The arrangement of the DetNet and TSN layers relative to each other is illustrated schematically. Here, the first layer 10 is placed above the second layer 12. Each layer has its own network components in sub-architectures 14, 16, and 18.

[0056] In the network architecture TSN-over-DetNet, the first layer 10 will be the DetNet layer, and the second layer 12 will be the TSN layer. Correspondingly, sub-architecture 14 will be the DetNet part of the network architecture, and sub-architectures 16 and 18 will be subnets with TSN structures, respectively.

[0057] In the alternative DetNet-over-TSN network architecture, the opposite will be true. Sub-architecture 14 will be a subnet with a TSN controller, and the first layer 10 will be the TSN layer. Furthermore, correspondingly, sub-architectures 16 and 18 will be DetNet networks, and the second layer 12 will be the DetNet layer.

[0058] Figure 2 The diagram schematically illustrates communication within a network architecture 20 according to one embodiment. Network architecture 20 includes a DetNet controller 22 and two TSN controllers 24 and 26.

[0059] DetNet controller 22 is configured to communicate with TSN controllers 24 and 26. TSN controllers 24 and 26 cannot communicate directly with each other.

[0060] In the first step S10, module 28, used for identifying network activity and applications, identifies the topology and the resource-binding entities active within it, such as applications and data flows. To this end, terminal devices 32 and 34 located in the subnets of TSN controllers 24 and 26 report their resource requirements to TSN controllers 24 and 26, which in turn notify module 28 of these resource requirements.

[0061] If the topology and applications within network architecture 20 are known, DetNet controller 22 can determine a resource budget for TSN controllers 24 and 26. For this task, DetNet controller 22 has a configuration engine 30. Configuration engine 30 notifies the TSN controllers of the resource budget allocated to them, which is available for use by the TSN controllers to configure their respective subnets.

[0062] Next, TSN controllers 24 and 26 can check whether the resource budget allocated to them is sufficient, or whether the TSN controller needs more resources. If the resources are insufficient, the TSN controller reports this back to the configuration engine 30, so that the DetNet controller 22 can re-determine the resource budget and allocate it to the TSN controller.

[0063] Figure 3 An industrial facility with a network architecture 20 according to one embodiment of the present invention is illustrated schematically and exemplary. The network architecture 20 has a DetNet controller 22 and two TSN controllers 24 and 26.

[0064] Each TSN controller 24 and 26 has its own subnet assigned to it. The subnet of TSN controller 24 includes terminal devices 36, 38, 40, 42, 44, and 46, which are interconnected via two switches 48 and 50. The subnet of TSN controller 26 includes terminal devices 52, 54, 56, 58, 60, and 62, which are interconnected via switches 64 and 66.

[0065] DetNet controller 22 forms the interface for end devices 36, 38, 40, 42, 44, 46, 52, 54, 56, 58, 60, and 62 to connect to external networks. For this purpose, the DetNet controller includes two DetNet edge nodes 68 and 70 and a DetNet router 72.

[0066] DetNet router 72 is connected to backbone network 74, which communicates with DetNet controller 22 via DetNet edge node 76. Finally, the facility can connect to edge cloud or edge cloud server 78, which also communicates with DetNet controller 22 via DetNet edge node 76.

[0067] For example, four applications 80, 82, 84, and 86 can be executed within network architecture 20. Application 80, for example, has a cycle time of 1 ms and a payload of 300 B. This application is executed by terminal device 38. Application 82 has a cycle time of 500 μs and a payload of 500 B. This application is executed by terminal device 44. Application 84 has a cycle time of 4 ms and a payload of 1000 B. This application is executed by terminal device 56. And application 86 has a cycle time of 1 ms and a payload of 1500 B. This application is executed by terminal device 58.

[0068] Furthermore, the two background applications 88 and 90 in the backbone network 74 are executed by nodes within the DetNet network. Background application 88 may be, for example, a video stream with a bandwidth requirement of 40 Mbps, while background application 90 may be a video stream with a bandwidth requirement of 20 Mbps.

[0069] DetNet controller 22 can detect background applications 88 and 90 through modules (not shown here) used to identify network activity and applications.

[0070] Through TSN controller 24, DetNet controller 22 can obtain information about applications 80 and 82, and through TSN controller 26, DetNet controller 22 can obtain information about applications 84 and 86. It is also determined which DetNet edge nodes 68, 70, and 76 execute the applications. Applications 80 and 82 are run through DetNet edge node 68, and applications 84 and 86 are run through DetNet edge node 70. All applications 80, 82, 84, and 86 communicate through DetNet edge node 76.

[0071] DetNet controller 22 uses this knowledge to determine resource budgets for TSN controllers 24 and 26 in the form of defined times. To do this, it first determines how much of the available resources must be used for background applications 88 and 90. The remaining resources are then allocated to TSN controllers 24 and 26.

[0072] For example, the following resources can be allocated to TSN controllers 24 and 26: 250 μs is allocated by DetNet edge node 76 and 750 μs by DetNet edge node 68 for application 80. 250 μs is allocated by DetNet edge node 76 and 250 μs by DetNet edge node 68 for application 82. 500 μs is allocated by DetNet edge node 76 and 3.5 ms by DetNet edge node 70 for application 84. 500 μs is allocated by DetNet edge node 76 and 0.5 ms by DetNet edge node 70 for application 86.

[0073] TSN controllers 24 and 26 can distribute the resources within their subnets. If the resources are insufficient to execute all applications within the subnet or to enable their communication with other users in the network, TSN controller 24 or 26, which is not providing sufficient resources to its subnet, reports this back to DetNet controller 22. DetNet controller 22 can then begin to re-determine or allocate resource budgets for TSN controllers 24 and 26. This process is repeated until TSN controllers 24 and 26 have allocated sufficient resources to their subnets, or some of applications 80, 82, 84, and / or 86 have been shut down, allowing the remaining applications 80, 82, 84, and / or 86 to execute.

[0074] Figure 4 The flowchart of a method for configuring a network architecture according to one embodiment is illustrated schematically.

[0075] The method begins in step S10, where the DetNet controller identifies the end-to-end configuration for the DetNet layer. This end-to-end configuration includes paths to prevent network disruptions when the network topology changes.

[0076] In step S12, a resource budget is allocated to each subnet in the TSN layer. This allocation is not yet a specific resource assignment, as that is performed later by the corresponding TSN controller for the subnet.

[0077] In step S14, the DetNet controller transfers the resource budget to the corresponding TSN controller. The resource budget does not need to be evenly distributed here. The resource budget can be adapted to the requirements of the application running in the corresponding subnet, so that a TSN controller running a particularly resource-intensive application in its subnet receives a larger budget compared to a DetNet controller that does not require such a amount of resources from the outset.

[0078] In step S16, each TSN controller checks whether the resource budget allocated to it is sufficient to create an effective resource plan for its own subnet.

[0079] If each TSN controller in the TSN controllers can create a resource plan for its own network with the resource budget allocated to it, then the subnet or the entire network architecture is configured accordingly in step S18.

[0080] If one of the TSN controllers cannot supply sufficient network resources to its subnet in step S16, that is, if the resource budget allocated to it is insufficient to meet the requirements of its subnet, then the TSN controller transmits a response to the DetNet controller in step S20.

[0081] Next, the DetNet controller begins to reallocate the resource budget according to step S12. Here, if the subnet's requirements are already included in the response of the relevant TSN controller, the DetNet controller may specifically consider those requirements.

[0082] Furthermore, the TSN controller can be configured to also respond to the DetNet controller after the step of creating its resource plan, so as to notify the DetNet controller of any excess resources if necessary, or to signal the successful creation of the resource plan.

Claims

1. A computer-implemented method for configuring a hierarchical network architecture (20), wherein the network architecture (20) includes at least a first layer and a second layer, wherein the first layer includes a first controller, and wherein the second layer includes a plurality of subnets and includes a second controller for each subnet, wherein the method includes the following steps: - Determine the end-to-end configuration for the first layer using the first controller (S10). - The resource budget is allocated for each subnet by the first controller according to the end-to-end configuration (S12). - The allocated resource budget is communicated to each of the second controllers through the first controller (S14). as well as - Create a resource plan for the subnet through each second controller, and configure the network architecture when the resource budget transmitted to the second controller is sufficient (20, S18), or transmit a response from the second controller to the first controller when the resource budget transmitted to the second controller is insufficient (S20).

2. The computer-implemented method according to claim 1, wherein the first layer is a DetNet layer, and wherein the first controller is a DetNet controller (22).

3. The computer-implemented method according to claim 1 or 2, wherein the second layer is a TSN layer, and wherein the second controller is a TSN controller (24, 26).

4. The computer-implemented method according to any one of the preceding claims, wherein the method is repeated when the resource budget for one of the second controllers is insufficient.

5. The computer-implemented method according to any one of the preceding claims, wherein the response includes an inquiry for a larger resource budget, in particular a minimum size of the resource budget.

6. The computer-implemented method according to any one of the preceding claims, wherein the response includes a description of the resource budget actually needed, wherein the response is also transmitted to the first controller even if the resource budget allocated to the corresponding second controller is sufficient.

7. A computer-implemented method according to any of the preceding claims, wherein when the resource budget for one of the second controllers is insufficient, data traffic to one or more second controllers within the network architecture (20) is restricted, wherein a higher resource budget is allocated to the unrestricted second controller.

8. A computer-implemented method according to any one of the preceding claims, wherein the following steps occur before determining the end-to-end configuration for the first layer: - Identify the network topology and possible resource-bound entities in the first layer.

9. The computer-implemented method of claim 8, wherein the first controller includes a module for identifying network topology and possible resource-binding entities in the first layer.

10. The computer-implemented method according to any one of the preceding claims, wherein the first layer is placed on top of the second layer.

11. The computer-implemented method according to any one of claims 1 to 9, wherein the second layer is placed above the first layer, and wherein the following steps occur before the end-to-end configuration is determined: - Transmitting subnet information from each of the second controllers (24, 26) to the first controller, wherein the subnet information includes: - Requirements for data traffic within the subnet; and - The topology of the subnet.

12. The computer-implemented method according to any one of the preceding claims, wherein the network architecture (20) maps to a network within an industrial facility, and The subnet includes multiple control units of equipment (32, 34) of a machine, part of a machine, or group of machines.

13. A network architecture (20) for an industrial facility, wherein the network architecture (20) includes a first layer and a second layer, wherein the first layer and the second layer include means for performing the method according to any one of the preceding claims.

14. A computer program having program code for performing the method according to any one of claims 1 to 12 when the computer program is executed on a computer.

15. A computer-readable data carrier having program code of a computer program for performing the method according to any one of claims 1 to 12 when the computer program is executed on a computer.

Images