COMPUTER-IMPLEMENTED METHOD FOR ADAPTING AT LEAST ONE PREDEFINED FRAME DELAY

Abstract

The invention relates to a computer-implemented method for adapting at least one pre-defined frame delay, wherein the corresponding at least one frame (20) is communicated via at least one network switch (12) of at least one computing unit (10) in a frame-based computer network (1), comprising the steps a. Receiving input data from the at least one network switch (12) of the at least one computing unit (10)(SI); wherein b. the input data comprises at least one information associated with the at least one frame (20) and / or at least one network switch (12); c.deriving at least one current frame delay of the at least one frame (20) from the at least one information (S2); d. adapting the at least one pre-defined frame delay depending on the at least one current frame delay (S3); and e.Providing the at least one adapted frame delay (S4). Further, the invention relates to a corresponding computer program product and monitoring unit.

Description

[0001] Description

[0002] Computer-implemented method for adapting at least one pre defined frame delay

[0003] 1. Technical field

[0004] The present invention relates to a computer-implemented meth od for adapting at least one pre-defined frame delay, wherein the corresponding at least one frame is communicated via at least one network switch of at least one computing unit in a frame-based computer network. Further, the invention relates to a corresponding computer program product and monitoring unit .

[0005] 2. Prior art

[0006] Computer networks are known from the prior art, according to which computing devices or units as nodes exchange data with each other using data links as connections. Local Area Net works (LANs) are computer networks which are confined to a small geographic area. LAN technologies include Ethernet, Arcnet und Token Ring. Ethernet is the most popular LAN tech nology in use today. The Ethernet standard is defined by the Institute for Electrical and Electronic Engineers (IEEE) as IEEE Standard 802.3. In Ethernet, a stream of data is divided into shorter pieces called frames for data transmission. Ac cordingly, Ethernet is a frame-based or frame-based transmis sion technology for the transmission of data.

[0007] Efforts are being made to provide real-time capability for standard Ethernet. One attempt is Time-Sensitive Network (TSN) . TSN extends the Ethernet bridging standard according to IEEE 802. IQ. In contrast to standard Ethernet according to IEEE 802.3 and Ethernet bridging according to IEEE 802. IQ, time plays an important role in TSN to enable the real-time communication . TSN provides several new features for real-time Ethernet, in cluding time synchronization and deterministic network commu nication. As depicted in Figure 2 in a simplified form, in such computer networks 1, usually, the computing units 10 do not all share one channel through a shared cable or a repeat er hub; instead, each computing unit 10 communicates with a switch 12, which in turn forwards the incoming data traffic to the target computing unit 10 or destination station. Each computing unit 10 communicates only with its own switch 10, without competition for the cable with others. The computing units 10 exchange the frames 20. Regarding TSN, the switches can be referred to as TSN switches.

[0008] Usually, the planning of the data traffic in a TSN is based on a worst-case analysis of the data transmission time of the transmitted frames. This analysis is used to determine wheth er enough bandwidth is available for the intended frame transmission .

[0009] Thereby, the worst-case analysis typically uses the upper boundary of random frame delays, such as Bridge-Delays and / or the a priori unknown interference delays caused by different frame usages of the same communication links, i.e. same out put ports in a switch. However, most of the upper boundaries and interference delays are very conservative, resulting in unnecessary reservation or resources, including the bandwidth reservation .

[0010] It is therefore an objective of the invention to provide a computer-implemented method for adapting at least one pre defined frame delay, wherein the corresponding at least one frame is communicated via at least one network switch of at least one computing unit in a frame-based computer network, in an efficient and reliable manner to improve the reserva tion of resources and thus data traffic capacity.

[0011] 3. Summary of the invention This problem is according to one aspect of the invention solved by a computer-implemented method for adapting at least one pre-defined frame delay, wherein the corresponding at least one frame is communicated via at least one network switch of at least one computing unit in a frame-based com puter network, comprising the steps: a. Receiving input data from the at least one network

[0012] switch of the at least one computing unit; wherein b. the input data comprises at least one information asso ciated with the at least one frame and / or at least one network switch; c. deriving at least one current frame delay of the at

[0013] least one frame from the at least one information; d. adapting the at least one pre-defined frame delay de

[0014] pending on the at least one current frame delay; and e. Providing the at least one adapted frame delay.

[0015] Accordingly, the method is directed to a computer-implemented method for adapting at least one pre-defined frame delay. The delay can be defined as delay of a frame, which is communi cated via a network switch of a computing unit in a frame- based computer network.

[0016] In other words, the frame delay is the time required to transmit a frame from one computing unit to another computing unit within the computer network. For example, the frame de lay can be the sum of transmission time and other delays e.g. Bridge delay and interference delay. Additionally or alterna tively to the pre-defined frame delay, at least one a priori estimate of the maximum queue-occupancy at the output ports of switches can be adapted.

[0017] Thereby, the frame is a digital data transmission unit. In packet switched systems, a frame is a simple container for a single network packet. Preferably, the frame-based computer network is an Ethernet network, such as TSN, as explained in more detail further above. Alternatively, any other frame- based computer network can be used.

[0018] First, the input data is received from one or more network switches of computing units of the frame-based computer net work. Preferably, a plurality of input data sets from differ ent network switches of the frame-based computer network are collected and received. The input data comprises information regarding the frame and / or the respective network switch.

[0019] In other words, the network switches are extended with a mon itoring or collecting capability. The network switches are able to monitor the traffic of data or data transmission. In particular, the traffic of the frames is monitored by the switches since the frames are communicated by them. The frames are received and forwarded by the network switches. Thus, for example, the time as time information as e.g. in the form of time stamps of the frames entering and / or leaving the switches can be monitored and collected. Additionally or alternatively, the entry or sending times of frames can be monitored and collected. Additionally or alternatively, the queue occupancy or occupancy of queues at e.g. the output ports can be checked, stored and / or forwarded. This infor mation can be collected and stored in any kind of storage unit, e.g. database or cloud. The input data can be transmit ted from the storage unit to a receiving unit.

[0020] The information is necessary to derive the current frame de lay in a next step. The adapted frame delay determined by ad justing or correcting the pre-defined frame delay using the current frame delay. According to which, the pre-defined un derlying or associated resource reservation e.g. bandwidth is adjusted, respectively.

[0021] This adjustment can be based on one current frame delay as single value for the frame. Alternatively, the aforementioned steps can be repeated to collect a plurality of current fame delays e.g. over time for one frame for determining a normal ized, mean or other statistical value as current frame delay. This way, the reliability of the current frame delay can be increased since more data is considered.

[0022] This way, the conservative or worst-case pre-defined or spec ified frame delay can be improved and thus the reservation of resources as well. Contrary to prior art, the method accord ing to the invention is based on real data traffic.

[0023] The method according to the invention results in an improved determination of the resource reservation, such as bandwidth in comparison to the prior art estimation performed in the network planning phase.

[0024] Moreover, the method according to the invention ensures an improved efficiency and accuracy since the method steps are performed by one central monitoring unit. The resulting adapted frame delay is more reliable compared to prior art since the method does not rely on mere estimations.

[0025] In one aspect the input data is raw or processed. According ly, the input data can be raw data or processed input data ensuring an increased flexibility with regard to the input format .

[0026] In another aspect the information is time information of the at least one frame and / or information about a queue occupancy of a storage unit of the at least one switch, wherein the storage unit stores the at least one frame. Accordingly, any kind of information associated with the frame, network switch and / or other data communicated by the switch can be used to derive the current frame delay and check possible frame- losses. The time information can comprise time stamps, see further above. Thus, the method can be applied in a flexible manner according to the specific application case, underlying technical system and user requirements. In another aspect the time information comprises at least two time stamps and the at least one current frame delay is de rived as the time difference between the at least two time stamps associated the at least one frame. Accordingly, the change in time or time difference can be determined based on the time stamps.

[0027] In another aspect the at least one current frame delay is a change of the queue occupancy. Accordingly, the change of the queue occupancy can be determined.

[0028] In another aspect the adaptation is an increase or decrease of the pre-defined frame delay.

[0029] In another aspect the pre-defined frame delay is a predefined conservative frame delay, reflecting the worst-case scenario of the at least one frame delay.

[0030] In another aspect the adaption is a decrease of the pre defined frame delay, if the current frame delay of the frame is below the pre-defined frame delay. Accordingly, the pre defined frame delay or estimation of the resource reservation can be corrected in a reliable and efficient manner.

[0031] In another aspect the method further comprises at least one step selected from the group, comprising:

[0032] - outputting the adapted frame delay to a user for further evaluation;

[0033] - displaying the adapted frame delay to a user for further evaluation; and

[0034] - transmitting the adapted frame delay to a computing unit for further evaluation, and - applying the adapted frame delay on the at least one compu ting unit, frame-based computer network or other units of the frame-based computer network.

[0035] Accordingly, the adapted frame delay as output data, data of intermediate method steps and / or any other related infor mation can be further handled or processed. One or more ac tions can be performed. The action can be equally referred to as measure or further method step. The actions can be trig gered depending on the user knowledge, evaluation by the user or any other criteria. The actions can be performed sequen tially or simultaneously. Actions include e.g. storing and processing steps. The advantage is that appropriate actions can be performed in a timely manner. For example, the deter mined adapted frame delay can be applied on the computer units of the network to set or implement the according re source reservation.

[0036] For example, in one exemplary use case, the method according to the invention can be used to check whether additional frames can be sent in the same cycle. Maybe with the previous very conservative values it looked not possible, but the method results in reduced estimates and hence additional frames .

[0037] A further aspect of the invention is a computer program prod uct directly loadable into an internal memory of a computer, comprising software code portions for performing the steps according to any one of the preceding claims when said com puter program product is running on a computer.

[0038] A further aspect of the invention is a monitoring unit.

[0039] The units may be realized as any devices, or any means, for computing, in particular for executing a software, an app, or an algorithm. For example, the monitoring unit may comprise a central processing unit (CPU) and a memory operatively con nected to the CPU. The monitoring unit may also comprise an array of CPUs, an array of graphical processing units (GPUs) , at least one application-specific integrated circuit (ASIC) , at least one field-programmable gate array, or any combina tion of the foregoing. The units may comprise at least one module which in turn may comprise software and / or hardware. Some, or even all, modules of the units may be implemented by a cloud computing platform.

[0040] Preferably the monitoring unit is a network controller.

[0041] 4. Short description of the drawings

[0042] In the following detailed description, presently preferred embodiments of the invention are further described with ref erence to the following figures:

[0043] Fig. 1 illustrates a flowchart of the method according to the invention.

[0044] Fig. 2 illustrates an exemplary frame-based computer net work according to prior art.

[0045] 5. Detailed description of preferred embodiments

[0046] Fig. 1 illustrates a flowchart of the method according to the invention with the method steps SI to S4. The method steps are performed by a monitoring unit (not shown) . According to an embodiment of the invention, the monitoring unit is a net work controller.

[0047] The network controller receives the input data with the in formation associated with the frame 20 and / or network switch 12 of the computing unit 10, SI.

[0048] The information can be time based, if the time information is available. Additionally or alternatively, the information can be based on the correlation between the traffic of the frames 20 at the input ports and the changes in the occupancy of the appropriate queues of the frames at the output ports. This information is essential for the further method steps S2 to S4, and thus to determine whether the pre-defined frame delay is correctly specified or rather in other words, whether the worst-case frame interference really takes place or not. The pre-defined frame delay and thus estimation of the resource reservation is corrected in step d. depending on said infor mation .

[0049] This way, the estimation of the resource reservation can be improved allowing for increased or additional data traffic.

[0050] In other words, the traffic capacity is significantly im proved within the frame-based computer network.

[0051] Reference signs

[0052] 1 network

[0053] 10 computing unit or computing device

[0054] 12 network switch

[0055] 20 frame

[0056] SI to S4 method steps SI to S4

Claims

Claims1. Computer-implemented method for adapting at least one pre-defined frame delay, wherein the corresponding at least one frame (20) is communicated via at least one network switch (12) of at least one computing unit (10) in a frame-based computer network (1), comprising the steps : a. Receiving input data from the at least one network switch (12) of the at least one computing unit (10) (SI) ; wherein b. the input data comprises at least one information associated with the at least one frame (20) and / or at least one network switch (12); c. deriving at least one current frame delay of the at least one frame (20) from the at least one infor mation ( S2 ) ; d. adapting the at least one pre-defined frame delay depending on the at least one current frame delay (S3) ; and e. Providing the at least one adapted frame delay (S4 ) .

2. Method according to claim 1, wherein the input data is raw or processed.

3. Method according to claim 1 or claim 2, wherein theinformation istime information of the at least one frame and / or information about a queue occupancy of a storage unit of the at least one switch, wherein the storage unit stores the at least one frame.

4. Method according to clam 3, wherein the time information comprises at least two time stamps and the at least one current frame delay is derived as the time difference between the at least two time stamps associated the at least one frame.

5. Method according to claim 3, wherein the at least one current frame delay is a change of the queue occupancy.

6. Method according to any of the proceeding claims, where in the adaptation is an increase or decrease of the pre defined frame delay.

7. Method according to any of the proceeding claims, where in the pre-defined frame delay is a predefined conserva tive frame delay, reflecting the worst-case scenario of the at least one frame delay.

8. Method according to claim 6 or claim 7, wherein theadaption is a decrease of the pre-defined frame delay, if the current frame delay of the frame is below the pre-defined frame delay.

9. Method according to any of the proceeding claims, where in the method further comprises at least one step se lected from the group, comprising:- outputting the adapted frame delay to a user for fur ther evaluation;- displaying the adapted frame delay to a user for fur ther evaluation; and- transmitting the adapted frame delay to a computing unit for further evaluation, and- applying the adapted frame delay on the at least one computing unit, frame-based computer network or other units of the frame-based computer network.

10. A computer program product directly loadable into an in ternal memory of a computer, comprising software code portions for performing the steps according to any one of the preceding claims when said computer program prod uct is running on a computer.

11. Monitoring unit for performing the method according to claims 1 - 9.

12. Monitoring unit according to clam 11, wherein the moni toring unit is a network controller of the frame-based computer network.