METHOD FOR CALCULATING TSN CAPABILITY OF WIRELESS BRIDGE, COMMUNICATION SYSTEM AND WIRELESS BRIDGE

Abstract

The present disclosure relates to a method for calculating the capabilities of a wireless base station (WB) in a communication system in a TSN network. The WB includes at least a UPF, a BS, and a UE. The UPF includes at least one NW port that enables communication between the BS and a first sub-network. The UE is wirelessly connected to the BS in the TSN network and includes a DS port that enables communication with a second sub-network. The method is executed in the WB and includes estimating a performance capacity of a wireless connection between the BS and the UE based on information about a network usage purpose for time-sensitive communication in the WB, and calculating the TSN capabilities of the WB corresponding to the estimated performance capacity. The present disclosure further relates to a corresponding communication system.

Description

[Technical field]

[0001] The present disclosure relates generally to methods and systems for time-sensitive networking (TSN), and more particularly to TSN networks including wireless bridges, such as 5G systems (5GS). [Background technology]

[0002] In particular, a method for calculating the capacity of a wireless bridge in a time-sensitive network and a corresponding communication system and its wireless bridge are disclosed.

[0003] Fast and reliable communication and information networking, especially time-sensitive networking (TSN), is essential for a smart factory environment. Such networking enables factory-wide integration by tightly connecting individual production steps as well as production planning and logistics.

[0004] To provide the required connectivity, industrial networks need to support various types of traffic, including time-sensitive (TS) traffic used for real-time control of machines and services that require deterministic performance.

[0005] Time Sensitive Networking (TSN) is standardized by IEEE 802.1Q to provide industrial networks with deterministic latency to handle TS traffic. End-to-end communication deadlines and bounded jitter are guaranteed through multiple mechanisms such as time synchronization of individual frames and streams and TSN scheduling with respect to per-stream and per-port gate control lists generated for all components in the network.

[0006] Initially, TSN networking was targeted at wireline networks capable of supporting high-speed communication between sensors, actuators, controllers and IT centers. The transition from wired sensors and actuators to wireless sensors and actuators brings additional benefits such as mobility, scalability, and reduced maintenance costs. To connect wireless devices to TSN networks, wireless transmission technologies as defined in 3GPP are required.

[0007] The integration between wired and wireless technologies raises several challenges that have yet to be fully resolved, such as standards compatibility and the nature of data transmission over (limited) wired and (open) wireless media.

[0008] The particular compatibility issue between wired and wireless technologies that the present invention addresses relates to the variable throughput capacity (or bandwidth) of wireless bridges, which can alter the deterministic latency experienced by data streams passing through 5GS.

[0009] The following wireless bridge (WB) description focuses on 3GPP / 5G networks, but the description can be generalized to any wireless communication system.

[0010] Some TSN scheduling and routing use cases are computationally complex and require time-consuming computation of per-stream and / or per-port gate control lists for each component in the entire TSN network. To accommodate these types of use cases, it is beneficial to centralize the computation in a single centralized entity that can collect information about the entire network in order to find the best network configuration and schedule for each component in the network. Such a model is commonly referred to as centralized.

[0011] According to the centralized model, as defined by the IEEE 802.1Qcc standard, a Centralized User Configuration (CUC) entity 1 discovers end stations (ES) 2 (11), retrieves the TSN capabilities and user requirements of these end stations, in other words the traffic specifications at these end stations (12, 13), and configures TSN features at these end stations (14). The configuration information is then provided to a Centralized Network Configuration (CNC) entity 3 (15). The CNC entity discovers the physical topology (i.e. links and bridges) of the TSN network 4 (16), reads the TSN capabilities of all these bridges (17), then generates a schedule (18) and, if the schedule is successfully generated, configures all bridges for TS communication (19). To accomplish this task, the CNC needs a complete view of the physical topology of the network and the TSN capabilities of all bridges.

[0012] If the CNC cannot find a TSN schedule that satisfies the requirements of all stream requests obtained through the CUC, a new iteration of the TSN configuration loop is initiated through adjustment of the initial stream requests (20). Such a configuration loop proposed for wired TSN as defined by the IEEE 802.1Qcc standard is shown diagrammatically in Figure 1, followed by a detailed workflow description provided in Figure 2.

[0013] The embodiment in Fig. 1 assumes that the network physical topology and the TSN capabilities of all components (=bridges) are fixed and known before the TSN scheduling is determined. The static nature of the wired components, especially the fixed bandwidth, allows the CNC to read the network physical topology and wired TSN bridge capabilities only once and then use these data for the TSN scheduling procedure, no matter how many iterations are required to generate a successful TSN schedule corresponding to the initial (or adjusted) stream request.

[0014] In contrast, in the case of a wireless bridge such as 5GS, the throughput capacity of its internal wireless links is not fixed. Instead, this throughput capacity depends on the radio resource (RR) allocation scheme used by 5GS and the interdependent repartition of this RR among the wireless devices competing for this RR. That is, the more resources are provided to one link, the less resources remain available for the other resource(s).

[0015] On the one hand, the ability of Wireless Bridges (WBs) to adjust their link throughput capacity according to traffic load by choosing appropriate internal configurations gives them a great advantage compared to wired TSN bridges. On the other hand, the optimization problem of computing the appropriate repartition of limited RRs among WBs requires prior knowledge of the future traffic load, which is not available at the time when 5GS is requested to declare its TSN capabilities. This creates a chicken-egg problem that complicates 5GS / TSN integration.

[0016] In fact, in the case of WB, all internal wireless links pass through a shared wireless medium. Due to collisions in the wireless medium, in order to avoid crosstalk between different ports caused by the change of the wireless channel in space and time, different wireless resource allocation techniques and time / frequency multiplexing techniques are used, which results in different TSN capabilities of WB.

[0017] This makes the standard TSN network scheduling procedure depicted in FIG. 2 inefficient, at least in what concerns the selection of an appropriate internal configuration among different possible internal configurations of the WB that are available through different RR allocations.

[0018] Without prior knowledge of its intended use, the TSN capabilities of a WB can only be defined based on hypothetical worst-case scenarios, not actual traffic loads. Such "guesswork" may result in additional delays and jitter for streams going through 5GS, which could potentially be avoided by appropriate selection of the WB internal configuration.

[0019] The prior art does not teach how to solve this problem, which still exists in all TSN networks with the above-mentioned WB, characterized in that the wireless bridge has variable TSN capabilities associated with different resource allocations that can be used to support a given traffic load that is not yet defined at the time of the TSN capability declaration.

[0020] In particular, non-patent document 1 teaches that "the interface between the 5GS and the CNC enables the CNC to learn the characteristics of the 5G virtual bridge," but does not explain whether and how these characteristics are estimated and guaranteed for traffic loads that are not yet defined at the time the 5GS is requested by the CNC to declare its TSN capabilities.

[0021] In patent document 1, a solution enabling communication between 3GPP WB and IEEE TSN networks is proposed based on the use of a TSN network translator (TT). The TT is proposed to enable bidirectional communication between 5GS and TSN networks for both control and data information by translating TSN data (e.g. coded according to IEEE 802.1) with respect to internal specification parameters (e.g. defined by 3GPP) that the WB can understand, and vice versa. In particular, this allows the CNC to treat the 5GS as a virtual TSN bridge with the same properties as a standard wired TSN bridge. The use of the TT simplifies the integration between TSN and 5GS, but does not solve the problem of the variable TSN capabilities of 5GS.

[0022] In Patent Document 2, this problem is partially solved by proposing an iterative optimization procedure characterized by successive evaluation of different possible internal configurations of the 5GS and selection of the best one among the tested configurations based on a performance index function related at least to the success of the TSN scheduling procedure executed by the CNC. This procedure results in an improvement of the overall performance of the TSN network with 5GS, but does not teach how the 5GS can anticipate the traffic load that may make the optimization time longer. [Prior art documents] [Patent documents]

[0023] [Patent Document 1] International Patent Application Publication No. 2020 / 081062 [Patent Document 2] European Patent Application No. 20306136.1 (not yet published) [Non-patent literature]

[0024] [Non-Patent Document 1] “5G-TSN integration for industrial automation” by J. Farkas et al. (Ericsson Tech. Review, 2019) [Non-Patent Document 2] A. Boriskin and R. Sauleau, “Hybrid genetic algorithms for fast electromagnetic synthesis” (Book chapter in Real-world applications of genetic algorithms, O. Roeva (ed), IntechOpen, 2012, DOI: 10.5772 / 38099) [Non-Patent Document 3] A. Galan, R. Sauleau, and A. Boriskin, “Floating boundary particle swarm optimization algorithm” (Optim. Lett. 7, 1261-1280 (2013), DOI:10.1007 / s11590-012-0502-8) Summary of the Invention [Problem to be solved by the invention]

[0025] There is therefore a need to provide 5GS with a means to predict its intended usage (and in particular to forecast traffic loads per port). Such predictions can help to solve the chicken-egg problem and thus enable more efficient integration between wireless and wired networks.

[0026] The invention is defined by the appended independent claims. Further features and advantages of the concepts disclosed herein are set forth in the following description.

[0027] This disclosure aims to improve the situation. [Means for solving the problem]

[0028] To this end, the present disclosure provides a method for calculating a capacity of a wireless bridge (WB) in a communication system in a time-sensitive network comprising a first sub-network and a second sub-network, the communication system comprising the wireless bridge and further comprising at least two end stations (ES), -The wireless bridge comprises at least one user plane function entity (UPF: User Plane Function), at least one base station (gNB), and at least one user equipment (UE: user equipment); The at least one user plane functional entity comprises at least one network side port (NW) enabling communication between at least one base station and a first sub-network; - the at least one user equipment is wirelessly connected to the at least one base station; the at least one user equipment comprises at least one device-side port (DS) enabling communication with a second sub-network; At least one of the at least two end stations is connected to a corresponding device-side port of a corresponding user equipment; The method is carried out in a wireless bridge, estimating a performance capacity of a wireless connection between at least one base station and at least one user equipment based on information about a network usage purpose for time-sensitive communication in a wireless bridge; Calculating the TSN capabilities of the wireless bridge corresponding to the estimated performance capacity; A method is described, including:

[0029] In the context of this disclosure, a wireless bridge is to be interpreted as a network bridge, the internal links of which are wireless, connecting at least one base station to at least one user equipment.

[0030] In the context of this disclosure, performance capacity is interpreted to include throughput capacity and / or capacity with respect to at least one quality of service parameter, such as delay, latency, jitter, etc.

[0031] The proposed method allows for a more accurate and reliable calculation of WB TSN capacity compared to prior art methods, which do not forecast or estimate any form of actual future traffic load at the WB, but instead rely only on hypothetical worst-case scenarios.

[0032] As a result, with a more accurate and reliable calculated WB TSN capability, the quality of service can be improved in general, and the dependent delay per port pair of the WB can be reduced in particular. The delay induced by a network device to handle a data packet is the sum of the independent delay, which is the part of said delay that is not affected by the amount of data it handles, and the dependent delay, which is the part of said delay that is dependent on the amount of data it handles. Reducing these dependent delays is key to facilitating the reduction of the overall delay of communication between end stations (e2e delay), at least for streams passing through the WB. Additional advantages include: the time required by the CNC to find a suitable internal configuration of the WB that can best serve the needs of the ESs whose data flows can pass through it is reduced, taking into account that this scheduling task may be very time-consuming or even impossible if it is performed by brute force optimization; - More efficient integration between wired and wireless networks / bridges; and Improved interoperability and improved access to the market for off-the-shelf hardware products from various vendors, achieved by the proposed method of subnetwork discovery, which can be performed autonomously by the WB without (or with limited) assistance from the CUC and / or CNC.

[0033] Optionally, estimating a performance capacity of a wireless connection between the at least one base station and the at least one user equipment based on information about a network usage purpose for time-sensitive communication in the wireless bridge includes: - obtaining information regarding a network usage purpose for time-sensitive communication in a wireless bridge; generating a quality of service template representing a performance metric corresponding to said network usage objective for time-sensitive communication in the wireless bridge based on the obtained information; - estimating a performance capacity of a wireless connection between at least one base station and at least one user equipment based on the generated quality of service template; Includes.

[0034] In the context of the present invention, a quality of service template defines desired performance metrics, such as parameters related to quality-of-service (QoS) or application quality-of-service policy (AQP), per DS port, per stream, or any other unit that can be mapped to the port pair units used to define bridge TSN capabilities, e.g., per UE, per port pair, etc.

[0035] In a first example, the quality of service template is based on equal sharing, i.e. the same performance metric is set for each port, each stream or each user equipment. Equal sharing can be used as a default option when there is insufficient information about the intended network usage.

[0036] In a second example, the quality of service template is based on balanced sharing, specifically on a per-port, per-stream or per-user equipment basis. Balance sharing can be used as a refinement option in cases where information on network usage purpose is differentiated on a per-port, per-stream or per-user equipment basis. Such sharing can be balanced, for example, based on sub-network discovery. Sub-network discovery can be applied to a first sub-network (i.e., an upstream sub-network) or a second sub-network (i.e., a downstream sub-network). A downstream sub-network is a separate sub-network of a TSN network, which is associated with a device-side port(s) of a WB, comprises at least one end station (ES), and has no bypass connection (neither wired nor wireless) with an upstream sub-network associated with a network-side port(s) of the WB. Performing sub-network discovery means obtaining information on at least the number of ESs in the corresponding sub-network and using this information to forecast the traffic load per UE / DS port.

[0037] Optionally, the method further comprises determining, for at least one port, a performance metric associated with said port in at least one operation mode, and generating the quality of service template is further based on each determined performance metric, wherein the operation mode is taken to include a downlink mode and / or an uplink mode.

[0038] In a first example, two ESs are located in a downstream sub-network and communicate in a unidirectional downlink mode (as an exemplary mode of operation) through corresponding DS ports of corresponding UEs, while at least one ES is located in an upstream sub-network and accessible through a gNB. The WB thus supports two radio links between the gNB and each of the two UEs competing for radio resources. According to such an example, the QoS template can be based on the traffic load per UE, and the TSN capability per WB port pair can be defined as a function of the corresponding radio link bandwidth.

[0039] In a second example, an ES is located in a downstream sub-network and communicates with another ES located in a first sub-network in a bidirectional down / uplink mode (as another exemplary operation mode). The communication between these ESs is performed in the WB through a radio link between a single gNB and a single UE. These two modes can share resources in the time domain, for example. Indeed, in many cases, a UE cannot support downlink and uplink simultaneously. According to this second example, it is possible to define a traffic load distribution between downlink and uplink modes in the time domain, which is further mapped to the frequency domain in terms of bandwidth allocation, as a QoS template. The TSN capabilities of the downlink and uplink WBs can then be defined as a function of the fractional bandwidths allocated for the downlink and uplink.

[0040] In the following, different options are presented for obtaining information about the network usage intent, either in terms of generating a QoS template or directly in terms of estimating the performance capacity of a wireless connection between at least one base station and at least one user equipment, which options can of course be combined.

[0041] Optionally, the time-sensitive network further supports, for example includes, a centralized user configuration entity (CUC) configured to communicate with the end stations, and the method further includes obtaining at least a portion of the above information regarding the network usage purpose for the time-sensitive communication in the wireless bridge based on a communication between the centralized user configuration entity and at least one of the end stations. The information can be obtained, for example, from a communication between the wireless bridge entity and the centralized user configuration entity. Alternatively, the information can be obtained in the wireless bridge entity by intercepting a signaling message exchanged between the centralized user configuration entity and at least one of the end stations. Alternatively, such information regarding the network usage purpose for the time-sensitive communication in the wireless bridge can be obtained based on a communication between the wireless bridge entity acting as a centralized user configuration entity and at least one of the end stations. In other words, in such a case, it is not a requirement that the time-sensitive network further comprises a centralized user configuration entity. Instead, the wireless bridge entity can emulate a centralized user configuration entity for the end stations.

[0042] Communication between the CUC and the ES makes it possible to retrieve, for example, the identity of the ES (eg, MAC address), the type of ES, or traffic specifications.

[0043] Optionally, the time-sensitive network further supports, for example includes, a centralized network configuration entity (CNC) configured to perform network discovery according to a centralized TSN model, and the method further includes obtaining, based on said network discovery, at least a portion of said information regarding network usage purpose for time-sensitive communication in the wireless bridge. Information about network discovery can be obtained, for example, from a communication between the wireless bridge entity and the centralized network configuration entity. Alternatively, this information can be obtained in the wireless bridge entity by intercepting a signaling message exchanged between the centralized network configuration entity and at least one of the end stations to perform network discovery. Alternatively, the wireless bridge itself acting as the centralized network configuration entity can perform such network discovery. In other words, in such a case, it is not a requirement that the time-sensitive network actually comprises a centralized network configuration entity (CNC). Instead, the wireless bridge entity can emulate a centralized network configuration entity to the end stations.

[0044] CNC-like network discovery, i.e. network discovery according to the centralized TSN model, makes it possible to retrieve the physical topology of at least a subnetwork of a TSN network and / or the TSN capabilities of links and bridges in this subnetwork.

[0045] Optionally, the time-sensitive network further comprises a plurality of TSN bridges connecting end stations to the wireless bridge, at least one of said TSN bridges being configured to perform network discovery according to a distributed TSN model, and the method further comprises obtaining, based on the network discovery, at least a portion of said information relating to network usage intent for time-sensitive communication in the wireless bridge.

[0046] Bridged network discovery, i.e. network discovery according to a distributed TSN model, makes it possible, for example, to obtain the topology of downstream sub-networks and the TSN capabilities of any ES connected to a given bridge.

[0047] Optionally, obtaining at least a portion of said information is performed by receiving a communication sent by a configuration entity and destined for the wireless bridge.

[0048] Indeed, the WB can be configured to request, for example, from the CUC the physical or MAC addresses of the ESs in the TSN network, or to request from the CNC an indication of the topology of the upstream and / or downstream sub-networks. Direct communication with the CUC and / or CNC allows the WB to obtain at least partial information about the stream requests (from the CUC) and the network topology (from the CNC). It should be noted that at the time of such exchange between the CNC / CUC and the WB, the CNC is not yet able to calculate the TSN scheduling due to the lack of information about the TSN capabilities of the WB, and thus the path of the stream is not yet determined, hence the phrase "partial information" in the previous sentence.

[0049] Optionally, obtaining at least a portion of said information is performed by intercepting a communication sent by the configuration entity and destined for at least one end station.

[0050] Intercepting messages from the CUC or CNC passing through the WB and initially intended for a given ES in the downstream sub-network at the WB allows for covert and seamless acquisition of information about network usage. Such interception allows the WB to passively estimate the performance capacity of the radio connection between the base station and the user equipment associated with said ES without the need to actively send a previous request to the CUC / CNC. Various other applications are also possible, even outside the scope of the method for calculating the capacity of a wireless bridge. For example, several such intercepted messages can be time-stamped to form a time series and stored. Such a time series can allow for example to make inferences or predictions of future network usage at the WB, with the same advantage of passive, seamless and covert integration with respect to the normal operation of the CNC and / or CUC.

[0051] Of course, other discovery protocols may also be implemented, including standard known discovery protocols that retrieve an end station's identity, such as the end station's IP address or MAC address.

[0052] Optionally, the information regarding network usage intent for time-sensitive communication in the wireless bridge includes at least one element in a list of time-sensitive traffic load per port (device side and / or network side), packet loss rate, deterministic end-to-end latency per stream, and delay per TSN bridge.

[0053] The elements of this list are all types of metrics that can be used to determine the QoS template.

[0054] Optionally, the method further comprises balancing performance of the wireless bridge by adjusting radio resource allocation among a plurality of internal wireless links of the wireless bridge based on the quality of service template.

[0055] As a result, it is possible to first determine the desired QoS template and then balance the performance of the wireless bridges to match the desired QoS template.

[0056] Optionally, calculating a capability of the wireless bridge corresponding to the estimated performance capacity comprises calculating a delay based on a performance capacity of a wireless link between the base station and the user equipment.

[0057] Such a delay may be, for example, a dependent delay defined for a port pair consisting of a NW port of the UPF and a DS port of the UE connected by a radio link, or may be, for example, the sum of the dependent delay and an independent delay related exclusively to the bridge hardware.

[0058] Optionally, the method further comprises transmitting the calculated capabilities of the wireless bridge to another entity of the communication system, with consideration for managing the time sensitive network.

[0059] For example, the CNC can receive the calculated capabilities of the WB and can generate TSN scheduling based on these calculated capabilities.

[0060] Optionally, the method further comprises: obtaining time-sensitive network scheduling information determined based on the transmitted calculated capabilities; and - estimating an adjusted performance capacity of a wireless connection between the at least one base station and the at least one user equipment based on the obtained time-sensitive network scheduling information; - calculating an adjusted capacity of a wireless bridge corresponding to the estimated performance capacity; Further includes.

[0061] This corresponds to an interactive optimization loop that refines the QoS template based on the TSN scheduling information provided by the CNC after receiving the initial TSN capabilities of the WB. Of course, this interactive optimization loop can also take into account additional information, such as updated information on network usage intent, which can be used to estimate the adjusted performance capacity of said wireless connection.

[0062] In another aspect, the present disclosure provides a communication system in a time-sensitive network including a first sub-network and a second sub-network, the communication system including a wireless bridge and further including at least two end stations (ESs); The wireless bridge comprises at least one user plane function entity (UPF), at least one base station (gNB), and at least one user equipment (UE); The at least one user plane functional entity comprises at least one network side port (NW) enabling communication between at least one base station and a first sub-network; the at least one user equipment is wirelessly connected to the at least one base station in a time sensitive network; the at least one user equipment comprises at least one device-side port (DS) enabling communication with a second sub-network; At least one of the at least two end stations is connected to a corresponding device-side port of a corresponding user equipment; The wireless bridge is - estimating a performance capacity of a wireless connection between at least one base station and at least one user equipment based on information about a network usage purpose for time-sensitive communication in a wireless bridge; - calculating the capacity of the wireless bridge corresponding to the estimated performance capacity; A communication system configured to:

[0063] The present disclosure provides, in another aspect, a method for producing a method of a method for producing a pharmaceutical composition comprising: a wireless bridge for the above communication system; - a computer-readable storage medium comprising instructions that, when executed by a processing unit, cause the processing unit to perform the methods described herein; - a computer program accessible to a processing unit and comprising one or more stored sequences of instructions which, when executed by the processing unit, cause the processing unit to carry out a method as described herein; and - a processing circuit configured to perform the methods described herein, comprising a processing unit operably connected to a memory; Also described below. [Brief description of the drawings]

[0064] [Figure 1] FIG. 2 shows a TSN configuration loop corresponding to a fully centralized model as known from the prior art. [Diagram 2] FIG. 2 illustrates a TSN scheduling workflow associated with the TSN configuration loop of FIG. 1 as known from the prior art. [Diagram 3] FIG. 2 illustrates a general model of a wireless bridge according to an exemplary embodiment of the present invention. [Figure 4] A diagram showing a TSN network with a WB with three DS ports according to an exemplary embodiment of the present invention. [Diagram 5] FIG. 1 illustrates a TSN network having ESs located in a downstream sub-network that are identified as talkers and listeners based on profiles of the ESs according to an exemplary embodiment of the present invention. [Figure 6] FIG. 1 illustrates a TSN network comprising a WB having two base stations with overlapping coverage zones and three UEs, one of which is located within the overlapping coverage zone, according to an exemplary embodiment of the present invention. [Figure 7]FIG. 1 illustrates a TSN network with only one UE according to an exemplary embodiment of the present invention. [Figure 8] FIG. 1 illustrates a TSN network comprising one gNB, two UEs, and two ESs located in a downstream subnetwork according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0065] The present invention relates to a method for calculating the TSN capabilities of a wireless bridge in a TSN network. A wireless bridge (WB) is an entity of a TSN network whose function is to connect an upstream sub-network with a downstream sub-network through wireless means. An example of a wireless bridge is a 3GPP / 5G system. The general model of a WB is shown in Figure 3. The WB 31 comprises: - User Plane Function (UPF) 32, which represents a component of the 3GPP 5G core infrastructure system architecture; At least one base station (gNB) 33; and - at least one User Equipment (UE) 34;

[0066] At least one UE is wirelessly connected to the at least one gNB and comprises at least one device-side (DS) port 35. FIG. 3 illustrates K such DS ports, where K is any positive integer. The UPF has a wired connection to the at least one gNB and typically comprises at least one network-side (NW) port 36. FIG. 3 illustrates N such NW ports, where N is any positive integer. Finally, each port may comprise a TSN translator (TT) as defined in 3GPP TS 2013-01-13344, whose purpose is to hide the internal features of the WB from the CNC and other network entities. Note that some available NW and DS ports may not be used for TS communication and are therefore omitted from the WB TSN capability calculation.

[0067] Since the purpose of a TSN network is assumed to enable communication between multiple end stations, the TSN network further comprises at least two end stations (ESs) in the context of the present invention. In some embodiments, the TSN network may also comprise a centralized network configuration (CNC) entity 37 and / or a centralized user configuration (CUC) entity. The latter discovers the end stations (ESs), reads their user requirements and configures their TSN features. The former discovers the physical topology and TSN capabilities of the network infrastructure (e.g. links and bridges), calculates TSN scheduling, configures the TSN features of said bridges according to the TSN scheduling, thus enabling the desired connectivity of at least two ESs according to stream requests obtained directly from the ESs or through the CUC. The TSN network may further comprise at least one wired TSN bridge with a fixed configuration, in the sense that its TSN capabilities can be characterized by a single set of TSN parameters, e.g. minimum / maximum dependent delay and independent delay per port or per port pair according to the IEEE 802.1Qcc standard.

[0068] Unlike wired bridges, WBs are characterized by having variable TSN capabilities that can be adjusted on demand by adjusting the internal features of the WB for wireless communication. In particular, this may relate to the dependent delay per port pair on the throughput capacity of the corresponding wireless link. The dependent delay per port pair may also relate to the amount of resources periodically reserved for transmission and the wireless link quality associated with said transmission. The contention for the wireless resources used between UEs results in the interdependent nature of the wireless links: the more resources are provided to a given link, the less resources are left for other links. Such agility of WB throughput capacity per UE potentially makes it possible to adjust the WB performance according to the traffic load and in such a way to optimize its performance. However, the established TSN configuration workflow described in FIG. 2 does not provide a means for the bridge to obtain the relevant information by the time it is required to declare its TSN capabilities. This creates the chicken-egg problem that the present invention addresses.

[0069] The proposed method for calculating the TSN capability of a WB involves the following steps performed by the WB: obtaining information regarding a network usage purpose for time-sensitive communication in a wireless bridge; - optionally generating, based on the obtained information, a quality of service template representative of a performance metric corresponding to said network usage objective for time-sensitive communication in the wireless bridge; - estimating a performance capacity of a wireless connection between at least one base station and at least one user equipment, directly on the basis of the obtained information or on the basis of a quality of service template derived from this information; - calculating the TSN capabilities of the wireless bridge corresponding to the estimated performance capacity; It is characterized by:

[0070] By estimating the performance capacity of a wireless connection between at least one base station and at least one user equipment, the method according to the invention solves the chicken-egg problem.

[0071] The above-mentioned information on the network usage for time-sensitive communication in the wireless bridge can be predicted either in the WB or any other entity of the TSN based on one or more types of retrieved information. For example, the WB or any other entity of the TSN can first retrieve information on ESs in the downstream sub-network, such as by implementing any standard or known network discovery method, and then predict information on the network usage based on the retrieved information. The prediction quality on the network usage for time-sensitive (TS) communication in the WB can be improved by using additional types of information that may relate to, for example, ES types, traffic specifications, network topology, TSN capabilities of any other network entities (links and bridges) in the TSN network, and / or TSN scheduling. This additional type of information can be obtained by any available means, either autonomously (through direct communication with other network entities, such as ESs and bridges) or with the assistance of the CUC and / or CNC.

[0072] The obtained information can be directly used to estimate the performance capacity of the wireless connection between at least one base station and at least one user equipment, given the pre-set or pre-configured radio resource allocation in the WB. Alternatively, in some embodiments, the information on the network usage purpose for time-sensitive communication in the wireless bridge in the downstream sub-network, once obtained or predicted by the WB, is used by the WB to generate a Quality of Service (QoS) template. Such a QoS template can then be used to balance the WB performance through adjustment of the internal features of the WB for wireless communication. Such adjustment allows the WB to use the available radio resources (RR) more efficiently to provide improved performance, in other words improved throughput capacity, or improved QoS parameters (such as packet loss rate, latency, or jitter) or a set of parameters adapted to the expected traffic load. In these embodiments, the obtained information is used to balance the WB performance and optimize the performance capacity of the wireless connection between at least one base station and at least one user equipment. In such a case, the obtained information is used to estimate the optimized performance capacity of the wireless connection between at least one base station and at least one user equipment.

[0073] After estimating the (optimized or non-optimized) performance capacity of a wireless connection between at least one base station and at least one user equipment, based on such estimation (e.g. according to the IEEE 802.1Qcc standard), the TSN capabilities of the 5GS bridge can be calculated and provided to a CNC which calculates TSN scheduling across the TSN network.

[0074] In the following, first the steps of the method are described with reference to a simple exemplary TSN network shown in Figure 4. Then the method is generalized for the case where the WB comprises more than one base station and can perform advanced channel coding and time / frequency multiplexing. Finally, two specific minimalistic examples with a single UE are also provided.

[0075] In the following, a method for obtaining information about network usage intent for time-sensitive communication in a wireless bridge is described. This information may relate to, for example, traffic load per DS port.

[0076] In this regard, it is assumed that the WB divides the TSN network into two sub-networks, or sub-networks. It is further assumed that there is no side connection between these two sub-networks. In other words, the only connection between these two sub-networks is via the WB. These two sub-networks are arbitrarily referred to below as the "upstream" sub-network 41 and the "downstream" sub-network 42, respectively (FIG. 3), with the upstream domain being accessible through the NW port and the downstream domain being accessible through the DS port.

[0077] Under this assumption, all data flows between a first ES located in the upstream subnetwork and a second ES located in the downstream subnetwork will pass through the WB. Therefore, to predict the data flows through a DS port, it is sufficient to obtain information about the entities in the corresponding branch of the downstream subnetwork. Such entities comprise the ES 43 and, if present, the bridge 44. This is why in FIG. 4 the topology of the upstream domain 41 is represented in less detail, contrary to the topology of the downstream subnetwork 42, which in this example comprises K=3 branches supporting M=6 end stations 43. More precisely, in this example, in the downstream subnetwork: a first branch comprises a single bridge B1 linked to two end stations ES1 and ES2, the second branch comprises a single terminating station ES3, - the third branch comprises a first bridge B2 linked to a first end station ES5 and to a second bridge B3, which itself is further linked to a second end station ES4 and to a third end station ES6.

[0078] Conversely, to predict the data flow through a NW port, it is sufficient to obtain information about the ESs in the corresponding branch of the upstream subnetwork.

[0079] Throughout this specification, the focus is primarily on traffic loads passing through DS ports and ESs located in downstream subnetworks, but those skilled in the art will recognize that the same teachings and method steps apply equally to traffic loads passing through NW ports and ESs located in upstream subnetworks.

[0080] Information about the ESs can be obtained by any known method. For example, this can be done using a proprietary discovery protocol such as the Cisco Discovery Protocol (CDP) or a non-proprietary discovery protocol such as the Link Layer Discovery Protocol (LLDP). The latter is a link layer protocol used by network devices to advertise their identity, capabilities, and proximity on a local area network. Information that can be retrieved using such protocols includes the system and port names of the network devices, their MAC / PHY information, and system capabilities.

[0081] In the most basic embodiment, it can be assumed that all ESs behave as talkers and are characterized by the same traffic specification. In other words, all ESs generate data flows characterized by the same bandwidth, and all these flows pass through the WB towards the upstream sub-network that contains the listener. Under these assumptions, L k DS The traffic load per DS port, denoted by , can be defined as the sum of the streams generated by the ESs located on the corresponding branch of the downstream subnetwork 42 as follows:

number

[0082] Under the assumption that all talkers are identical, the traffic load per DS port is proportional to the number of ESs per branch of the downstream subnetwork, i.e.,

number

[0083] Finally, the distribution of traffic load among all DS ports can be defined in vector form, denoted by T, as follows:

number

[0084] As an example, for a TSN network as shown in FIG. 4 with three DS ports (K=3) and a WB with two ESs, one ES, and three ESs for each corresponding branch of the downstream subnetwork, respectively, the expected relative traffic load per DS port can be encoded as follows:

number

[0085] This relative traffic load per port, represented here in matrix form for simplicity, can be used as a QoS template to adjust the WB internal features to enable optimal allocation of RR among UE / DS ports and thus best handle non-identical traffic loads per DS port.

[0086] In the following, possible methods according to some embodiments are described to improve the prediction quality when predicting information about network usage intent for time-sensitive communication in wireless bridges.

[0087] The quality of the predictions for the traffic load per DS port can be improved by collecting additional information about the ESs located in the downstream subnetwork 42. In particular, it may be more useful to distinguish between ESs acting as talkers (T) and ESs acting as listeners (L). The former (i.e., talkers) aim to generate outgoing traffic flows towards listeners, while the latter (i.e., listeners) aim to receive incoming traffic flows from talkers. Such a distinction is illustrated in FIG. 5, which is completely identical to FIG. 4 except for whether an end station 43 acts as a talker 45 or as a listener 46. In particular, an end station 43 with labels ES1, ES3, ES4, and ES5 in FIG. 4 acts as a talker 45 and has labels T1, T2, T3, and T4 in FIG. 5, while an end station 43 with labels ES2 and ES6 in FIG. 4 acts as a listener 46 and has labels L1 and L2 in FIG. 5.

[0088] In some embodiments, information about the ES profile can be used to predict the traffic load per DS port in downlink (DL) and uplink (UL) modes, respectively, under the assumption that all traffic flows to and from the upstream subnetwork. DL and T UL The expected traffic loads for DL ​​and UL modes, denoted by , can be defined for the example of FIG.

number

[0089] This relative traffic load per DS port can be used as a target template to adjust the WB internal features to enable best handling of asymmetric traffic loads per DS port in DL and UL modes, respectively.

[0090] In yet another embodiment, it may be further beneficial to obtain information about the traffic specifications supported by each ES. Using this information, the parameters of the streams to each listener and from each talker located in the downstream subnetwork 42 can be estimated, so that the aggregate traffic flow through each DS port in DL and UL modes, respectively, can be estimated as follows:

number

[0091] In the case of the exemplary TSN network shown in FIG. 5, this corresponds to the following equation:

number

[0092] The allocation of traffic load (proportional to stream bandwidth) per DS port in equation (7) can be used as a refined template to adjust the internal features of 5GS to enable better handling of traffic load per DS port in DL mode and UL mode, respectively.

[0093] Having exhaustive information about all TS stream requests can also be used to identify the specific case where both the T and L end stations associated with the same stream are located in the downstream sub-network. If the stream path between a T and L pair does not pass through the WB (e.g., as observed in FIG. 5 for T1 and L1, which in some embodiments may constitute a pair of devices that communicate between them through a B1 bridge), then these two ESs can be excluded from the WB traffic load template.

[0094] On the other hand, if the stream path between a T and L pair passes through the WB (e.g., as observed in FIG. 5 for T2 and L2, which in one embodiment may constitute a pair of devices communicating through the WB), then this traffic flow can be represented as a sequence of UL and DL streams from T2 to the WB / UPF and from the WB / UPF to L2 with appropriate time multiplexing between the two events.

[0095] Finally, knowledge of all ESs that actually require TS communication can help to identify ESs operating in best effort (BE) mode whose traffic can be served on a lower priority basis and therefore excluded from the WB performance template used to calculate its TSN capabilities. The distinction between TS type ESs and BE type ESs can be made, for example, by matching a list of MAC addresses of TS type ESs obtained from the CUC with a list of MAC addresses of all devices located in the downstream sub-network that are discovered using the LLDP protocol or any other discovery protocol.

[0096] In the following, a Quality of Service (QoS) template is defined that can be generated in a WB based on information about a network usage objective for time-sensitive communication in a wireless bridge, the QoS template representing a performance metric corresponding to the network usage objective for time-sensitive communication in said wireless bridge.

[0097] Some examples of QoS templates have already been introduced in Equations 3-7, where the performance metric is the traffic load in these examples. -Packet loss rate, -end-to-end (e2e) delay or latency, and - any combination of the aforementioned metrics, usually referred to as Quality of Service (QoS), It can be further generalized for other performance metrics related to individual streams and / or aggregated data flows, such as:

[0098] Then, such a multi-criteria QoS template can be defined as follows:

number

[0099] Such QoS templates can be used for multi-objective optimization of WB performance aimed at improving the WB's ability to handle actual traffic load with guaranteed QoS (e.g., in terms of packet loss rate) and reducing bridge delay to meet the maximum tolerable e2e latency for each individual stream.

[0100] To achieve the latter goal (i.e. minimizing e2e latency), priority in radio resource allocation can be given to streams that have the lowest quality of radio link and have stricter requirements on e2e latency (e.g. derived from the stream request) and / or longer network paths (e.g. derived from information about TSN scheduling and network physical topology).

[0101] Finally, it is useful to note that QoS templates can be further defined and applied taking into account different traffic classes that can be used to prioritize some streams over others. This can be done by defining a template for each traffic class, which is a straightforward generalization of the proposed method.

[0102] In the following, different possibilities for obtaining information about the ES are explained.

[0103] For information on ES profile (talker / listener), traffic type (TS / BE) and traffic specification, - it can be obtained via direct communication between the WB and the ES, for example using one of the methods proposed in the IEEE 802.1Q and IEEE 802.1Qcc standards for the discovery and reading of the TSN capabilities of an ES by a CUC (in the case of a centralized TSN network) or a TSN Bridge (in the case of a distributed TSN network). In the former case, said information on the ES can be obtained in the form of a list of stream requests containing information on the MAC addresses of all ESs in the entire TSN network, their profiles and traffic specifications defined, for example, in terms of ES operation intervals, number of frames per interval, frame sizes and possible transmission offsets and jitters, or any other data allowing to estimate the stream bandwidth (usually defined in Mbps, the same as the traffic load) and to calculate the TSN scheduling.

[0104] Alternatively, the information may be obtained, at least in part, from another network entity. For example, the information may include: From a CUC used to collect and process information about ESs in a centralized TSN network to generate stream requests used as initial (or adjusted) requirements for TSN scheduling generated by a CNC; or -From a CNC receiving information about ES and stream requests from a CUC to calculate the above TSN scheduling of the entire TSN network, It can be obtained.

[0105] In yet another embodiment, this information may be obtained by intercepting communications between the ES and other network entities (eg, CUCs) that pass through the WB.

[0106] In the following, various scenarios are described of how to balance WB performance characteristics according to a QoS template.

[0107] For simplicity, we first consider a WB 31 with only one gNB 33 (M=1), as shown in FIG. 3. Each UE 34 has only one DS port 35. Under this assumption, the traffic load per DS port can be directly mapped to the payload of the radio link between the UE and the gNB. Moreover, the bridge TSN parameters, which are usually defined for a port pair, can be assigned for the DS port (or UE) instead. Thus, to best meet a given payload per DS port, the relative link capacity per radio link will correspond to a QoS template that defines the expected traffic flow per DS port and / or stream. For example, in the case of the most basic embodiment with all identical ESs acting as talkers, whose expected traffic load is described by equation (3), the relative capacity of the WB radio link can also be defined as a function of the number of ESs per branch of the downstream sub-network connected to the corresponding DS port.

[0108] The adjustment of the radio link capacity of the WB according to the QoS template is possible due to the interdependence between said radio links, which is an inherent feature of wireless communication systems. Unlike wired connections, all radio links pass through a shared wireless medium, whose overall capacity is usually characterized in terms of a radio channel, whose overall capacity depends on the available frequency bands and the resource allocation in time and frequency used by the UEs. To avoid collisions, the overall capacity is shared among all UEs. Appropriate time / frequency and spatial multiplexing techniques and / or interference mitigation methods are used to enable at least partial reuse of limited radio resources (RR), which allows the expansion of the overall capacity of the wireless channel.

[0109] In the following, the simplest case of time multiplexing is referred to unless specifically mentioned, however the generalization of this basic definition to obtain more advanced methods of RR allocation does not change the logic of the proposed method.

[0110] In the general case, the C of the radio link between the base station (gNB) and the kth UE, also denoted as UE(k), k The throughput capacity denoted by can be defined as follows:

number

[0111] In one embodiment featuring either equal expected traffic load per DS port or unknown traffic load per DS port, the RR allocation problem can be formulated and solved to obtain equal throughput capacity provided to all UEs. The corresponding optimization problem is maximizing the minimum link capacity, i.e.

number

[0112] The solution of equation (10) can also be used as a default configuration for applicable wireless bridges that have no information about the intended use of WB. Moreover, it can be seen as the first step of a WB balancing procedure that aims to find a reference radio resource allocation that results in a fair share of the available RR and equal throughput capacity for all UEs.

[0113] The same balancing procedure can be further applied to a given distribution of traffic loads per DS port. For example, if this traffic load distribution is defined by equation (3), which corresponds to the most basic embodiment, the above optimization problem can be rewritten as follows:

number

[0114] L k In another embodiment, where the traffic load per DS port, denoted by , is already known or predicted, the optimization problem can be rewritten as follows:

number

[0115] The optimization problem can be further generalized for multiple scenarios, which allows splitting the optimization problem into two (or more) independent optimization problems such as - UL and DL modes (as described by Equation 7), and / or Different traffic priorities, and / or -Dual frequency connectivity.

[0116] In all these cases, the optimization problem can be split into at least two independent optimization problems that are formulated and solved separately for each mode, e.g. DL mode, traffic of the same priority only, and traffic load per DS port of the RR corresponding to one selected frequency subband.

[0117] Finally, the same approach can be applied to the scenarios introduced below, e.g. - a multi-criteria QoS optimization template in the definition of a WB performance metric that includes at least two different criteria (e.g. traffic load and packet loss rate), and / or - a WB with two or more base stations (M>1), and / or - Advanced time / frequency / spatial multiplexing for RR allocations to enable frequency reuse, It can be used to address multi-objective optimization problems corresponding to any of the above.

[0118] In the following, multi-criteria optimization using QoS templates is further discussed. It should be noted that the QoS templates can take into account several other performance metrics in addition to the traffic load, such as packet loss rate and deterministic e2e latency. These additional criteria can then be included in the formulation of the optimization problem. For example, this can be done as follows:

number

number

[0119] Of course, these simple examples can be elaborated further to better explain realistic PHY / MAC level features and techniques that can be used to improve the QoS of wireless networks, such as changes in radio channel characteristics due to interference, shadowing and multipath propagation, as well as known methods for QoS control: beamforming, interference mitigation, channel coding and adaptive redundancy. These features will be reflected in the definition of the Figure of Merit (FoM) used to derive the optimization problem, but do not change the proposed method, which aims at the definition of a reference solution (defined in the form of a QoS template representing the expected traffic load and optionally additional QoS optimization criteria) that can be used with any single-parameter or multi-parameter FoM function.

[0120] In an exemplary embodiment, to limit the search space by restricting the pool of acceptable values ​​of the optimization parameters, or To formulate a multi-objective optimization problem of vector type aimed at the simultaneous improvement of at least two dependent or independent figures of merit (FoM) functions describing the sought metrics of WB performance, e.g., throughput capacity per port and packet loss rate per stream, Additional optimization criteria such as equations (13) and (14) can be considered along with the optimization problem formulated by equations (11) or (12).

[0121] The former case can be effectively solved by genetic algorithms (GA), which allow the search and discovery of global maxima of discrete-time piecewise multi-extremal FoM functions that are difficult or not possible using brute force and gradient type methods. Hybrid GA methods, such as those defined in [2], can be further used to enhance the success and convergence speed of the algorithm.

[0122] The latter can be effectively solved by multi-objective optimization methods, commonly referred to as Pareto optimization, that allow finding a set of solutions that offer an acceptable trade-off between competing FoM functions.

[0123] Finally, advanced optimization methods, such as those defined in non-patent document 3, can be used to overcome possible challenges with uncertainties in defining the initial variation ranges of the optimization parameters.

[0124] In the following, the general case is discussed where the WB as shown in Figure 3 comprises two or more base stations (M>1). Multiple base stations allow to expand the coverage area of ​​the wireless network without increasing the signal level. Moreover, different gNBs can reuse at least a portion of the available RR when communicating with a UE, provided that the gNBs or UEs are located far enough apart from each other and / or spatial multiplexing is possible (e.g., via beamforming).

[0125] In some embodiments, it may therefore be beneficial to first balance the traffic load among the gNBs and then balance the traffic load among the UEs in each group associated with each gNB. For the example TSN network in Figure 5, the procedure and its results may look as shown in Figure 6. Such a network topology may be derived using the following procedure.

[0126] First, the traffic load per UE / DS port is predicted based on the number of ESs, e.g., as in equation (3). Then, the predicted traffic load per base station is balanced by maximizing the minimum load per gNB, e.g., as in the following equation:

number

number

[0127] With continued reference to FIG. 6, in the most basic embodiment where the traffic load is defined solely based on the number of ESs 45, 46, the procedure described above results in the following distribution of load among the gNBs:

number

[0128] The above procedure can be generalized to take into account the actual traffic specifications of each ES, similar to what is done in equation (7). If necessary, handover of UE2 between gNB1 and gNB2 can be performed to balance the traffic load per gNB by the advantageous location of UE2 within an overlapping zone (the coverage area of ​​each gNB is represented diagrammatically by the shaded area in Figure 6) characterized by having a sufficient signal level to establish a reliable communication link with both gNBs.

[0129] The above procedure for balancing the load per gNB can be generalized to take into account DL / UL operating modes for different traffic types and different usage scenarios, as discussed above for UE / DS load.

[0130] Finally, it is useful to note that the above procedure should preferably be applied under conditions where the interference level between gNBs is negligible. Such a condition is met if at least one of the following is true: -Each gNB,Σ m=1 M R m =R Σ RR=0.05× ... - gNBs are located far enough apart to at least partially reuse the same RRs but ensure an acceptable signal to interference ratio (SIR) to each other and to the associated UEs, or Each gNB and associated UE is capable of interference mitigation and / or multiplexing, for example based on spatial beamforming, MIMO, OFDM or any other method that helps minimize mutual interference between devices sharing the same RR to an acceptable level.

[0131] On the other hand, if the reuse of RR by the gNB results in a non-negligible interference level, the optimization problem can be formulated in a most general way for all involved entities (i.e., gNB and UE) and the resource allocation (α k ) and the efficiency of the UE in using RR (η k ) and the full set of optimization parameters for overall RR can be scaled up or down depending on the gNB and UE capabilities for RR reuse and interference mitigation.

[0132] In order to reduce the computational complexity of the general formulation of the RR allocation optimization problem, it is highly recommended to follow the teachings of the present invention, which allows to obtain a quick and reasonable prediction of the intended use of WB in TSN networks without solving computationally heavy and often ill-posed multi-parameter multi-objective optimization problems, which makes the proposed method attractive for algorithms that can be executed on nodes with limited computational power.

[0133] In the following some exemplary implementations of the method according to the invention are presented. For simplicity's sake, these implementations are described in the context of a communication system comprising as few entities as possible.

[0134] Figure 7 shows A WB 31 having a single NW port 36, a single UE 34 having a single DS port 35, and a single UPF 32; a bridge 44 in the downstream sub-network 42 and two ESs, such as a talker 45 and a listener 46, whose pair are located in the upstream sub-network 41; 1 illustrates an example embodiment of a TSN network comprising:

[0135] Although in this embodiment, only one radio link is available between only one gNB and only one UE, the proposed method of calculating WB TSN capacity can still be applied. Indeed, the throughput capacity of this single radio link can still be shared, for example, between DL and UL modes, for example, via time multiplexing. Assuming that the downstream sub-network 42 comprises a talker and a listener (as shown in FIG. 7), and having no additional information except for the information about the existence and type of these two ESs, it can be decided to equally share the available RR between DL and UL modes. For example, this can be done via time domain multiplexing or frequency multiplexing. The corresponding QoS template and throughput capacity can be formulated as the following equation:

number

[0136] Similarly, as already described with reference to the exemplary embodiment in FIG. 4, this example can be further tailored for the case of known traffic specifications defined for each ES and / or a QoS template including additional performance metrics such as packet loss rate.

[0137] Finally, it can be assumed that the two ESs in Figure 7 represent only one ES located in a downstream sub-network operating sequentially in DL and UL modes, and the model remains valid under this assumption.

[0138] Figure 8 shows A WB 31 comprising a UPF 32, none of whose NW ports are used for TS communication, a single gNB 33, and two UEs 34, each of which has a DS port 35; two ESs 45, 46, both located in the downstream subnetwork 42 and communicating with each other through the WB 31; 1 illustrates an example embodiment of a TSN network comprising:

[0139] The special feature of this example is that it does not consider any ESs located in the upstream sub-network, however, the proposed method of calculating the TSN capability of the WB can still be applied to calculate and optimize the RR allocation between two UEs working in UL and DL modes, respectively.

[0140] In the following, it is explained how to calculate the TSN capabilities. The final step of the proposed method is to convert the internal settings of the WB regarding wireless communication into TSN capabilities, which represent a form in which these internal settings of the WB can be understood by other network entities. This step is not mandatory for the application of the proposed method, but may be required to enable interoperability of the WB (e.g. 3GPP / 5GS) with wired TSN networks operating under, for example, IEEE 802.1Q or IEEE 802.1Qcc.

[0141] According to the IEEE802.1Qcc standard, the TSN capability of a bridge is dependent on the delay D d The IEEE 802.1Qcc standard includes a parameter describing the frame size dependent bridge delay, referred to as the frame size delay. This parameter is usually defined per port pair and per traffic type as defined by the IEEE 802.1Qcc standard and is equal to the inverse of the throughput capacity of the links connecting two adjacent components, as follows:

number

[0142] In the case of WB, the minimum link capacity (i.e., the bottleneck) is usually associated with the wireless link. Therefore, the dependent delay (D d WB ) can be defined based on the wireless link capacity as follows:

number

[0143] Note that the right hand side of equation (20) does not include a reference to a NW port, because in this example the radio link responsible for the bottleneck is associated with the gNB / UE pair (and not with the UPF / NW pair).

[0144] Consistent with the definition provided in the IEEE 802.1Qcc standard, the overall delay per port pair can then be calculated as the sum of the independent delay (associated with the bridge hardware) and the dependent delay defined by equation (20).

[0145] In one embodiment, the step of calculating the TSN may be repeated after receiving the TSN scheduling from the CNC.

[0146] To do this, an optimization loop can be added to adjust the template based on additional information obtained after the initial TSN capability declaration. This can be done when the WB receives the TSN scheduling calculated by the CNC based on its initial declaration. The information about the scheduling typically includes per-port scheduling and per-stream information as defined by the IEEE 802.1Qbv and IEEE 802.1Qcc standards, which the WB can use to: - A list of streams that actually pass through the WB, -Identification (e.g. MAC address) of the relevant ES located in the downstream sub-network or internal per-stream routing information (e.g. ingress and egress ports of the WB for a given stream), - Actual traffic load per stream / per port / per traffic class / per ES, can be used to find

[0147] Based on this information, the QoS template can be adjusted according to the teachings of the present invention as described above and can be used to recalculate the WB TSN capabilities.

[0148] The new TSN capabilities can then be provided to the CNC, which adjusts TSN scheduling as needed.

Claims

1. 1. A method for calculating a capacity of a wireless bridge in a communication system in a time sensitive network (TSN) comprising a first sub-network and a second sub-network, the communication system comprising the wireless bridge and further comprising at least two end stations, The wireless bridge comprises at least one user plane functional entity, at least one base station, and at least one user equipment; the at least one user plane functional entity comprises at least one network side port enabling communication between the at least one base station and the first sub-network; the at least one user equipment is wirelessly connected to the at least one base station; the at least one user equipment comprises at least one device-side port enabling communication with the second sub-network; At least one of the at least two end stations is connected to a corresponding device-side port of a corresponding user equipment; The method is performed in the wireless bridge, the wireless bridge includes a port pair consisting of the at least one network-side port and the at least one device-side port, the port pair being associated with a delay that is a sum of an independent delay and a dependent delay, the independent delay being a delay that is not affected by an amount of data to be handled, and the dependent delay being a delay that depends on the amount of data to be handled; The method comprises: estimating a performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on information regarding a network usage purpose for time-sensitive communication in the wireless bridge; calculating the capability at the TSN of the wireless bridge corresponding to the estimated performance capacity, the calculating the capability at the TSN including determining the dependent delay for the port pair based on the performance capacity of the wireless connection; A method comprising:

2. estimating a performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on information regarding a network usage purpose for time-sensitive communication in the wireless bridge, comprising: obtaining the information regarding intended use of the network for time-sensitive communications in the wireless bridge; generating a quality of service template based on the obtained information, the quality of service template representing a performance metric corresponding to the network usage objective for time-sensitive communications in the wireless bridge; estimating the performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on the generated quality of service template; The method of claim 1 , comprising:

3. determining, for at least one port, a performance metric associated with the port in at least one mode of operation; The method of claim 2 , wherein generating the quality of service template is further based on the determined performance indicators.

4. the time sensitive network further supporting a centralized user configuration entity configured to communicate with the end stations; The method according to any one of claims 1 to 3, further comprising obtaining at least a portion of the information regarding the network usage intent for time-sensitive communications in the wireless bridge based on communication between the wireless bridge acting as the centralized user configuration entity and at least one of the end stations.

5. The time sensitive network further supports a centralized network configuration entity configured to perform network discovery according to a centralized TSN model; The method according to any one of claims 1 to 3, further comprising obtaining at least a portion of the information regarding the network usage purpose for time-sensitive communication in the wireless bridge based on a network discovery performed by the wireless bridge acting as the centralized network configuration entity.

6. The time sensitive network further comprises a plurality of TSN bridges connecting the end stations to the wireless bridge, at least one of the TSN bridges being configured to perform network discovery according to a distributed TSN model; The method according to any one of claims 1 to 3, further comprising obtaining at least a portion of the information regarding an intended use of the network for time-sensitive communications in the wireless bridge based on the network discovery.

7. 1. A method for calculating a capacity of a wireless bridge in a communication system in a time sensitive network (TSN) comprising a first sub-network and a second sub-network, the communication system comprising the wireless bridge and further comprising at least two end stations, The wireless bridge comprises at least one user plane functional entity, at least one base station, and at least one user equipment; the at least one user plane functional entity comprises at least one network side port enabling communication between the at least one base station and the first sub-network; the at least one user equipment is wirelessly connected to the at least one base station; the at least one user equipment comprises at least one device-side port enabling communication with the second sub-network; At least one of the at least two end stations is connected to a corresponding device-side port of a corresponding user equipment; The method is performed in the wireless bridge, the wireless bridge includes a port pair consisting of the at least one network-side port and the at least one device-side port, the port pair being associated with a delay that is a sum of an independent delay and a dependent delay, the independent delay being a delay that is not affected by an amount of data to be handled, and the dependent delay being a delay that depends on the amount of data to be handled; The method comprises: estimating a performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on information regarding a network usage purpose for time-sensitive communication in the wireless bridge; calculating the capability at the TSN of the wireless bridge corresponding to the estimated performance capacity, where the calculating the capability at the TSN includes determining the dependent delay for the port pair based on the performance capacity of the wireless connection; A method, wherein obtaining at least a portion of the information is performed by receiving a communication sent by a configuration entity and directed to the wireless bridge.

8. 1. A method for calculating a capacity of a wireless bridge in a communication system in a time sensitive network (TSN) comprising a first sub-network and a second sub-network, the communication system comprising the wireless bridge and further comprising at least two end stations, The wireless bridge comprises at least one user plane functional entity, at least one base station, and at least one user equipment; the at least one user plane functional entity comprises at least one network side port enabling communication between the at least one base station and the first sub-network; the at least one user equipment is wirelessly connected to the at least one base station; the at least one user equipment comprises at least one device-side port enabling communication with the second sub-network; At least one of the at least two end stations is connected to a corresponding device-side port of a corresponding user equipment; The method is performed in the wireless bridge, the wireless bridge includes a port pair consisting of the at least one network-side port and the at least one device-side port, the port pair being associated with a delay that is a sum of an independent delay and a dependent delay, the independent delay being a delay that is not affected by an amount of data to be handled, and the dependent delay being a delay that depends on the amount of data to be handled; estimating a performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on information regarding a network usage purpose for time-sensitive communication in the wireless bridge; calculating the capability at the TSN of the wireless bridge corresponding to the estimated performance capacity, where the calculating the capability at the TSN includes determining the dependent delay for the port pair based on the performance capacity of the wireless connection; A method, wherein obtaining at least a portion of the information is performed by intercepting a communication sent by a configuration entity and destined for at least one end station.

9. 9. A method according to claim 1, wherein the information on the network usage for time-sensitive communication in the wireless bridge comprises at least one element in the list of: time-sensitive traffic load per port, packet loss rate, deterministic end-to-end latency per stream, and delay per TSN bridge.

10. The method of claim 2 or 3, further comprising: balancing performance of the wireless bridge by adjusting radio resource allocation among a plurality of internal wireless links of the wireless bridge based on the quality of service template.

11. 11. The method of claim 1, wherein calculating the capability of the wireless bridge corresponding to the estimated performance capacity comprises calculating a delay based on a performance capacity of a wireless link between a base station and a user equipment.

12. The method according to any one of claims 1 to 11, further comprising transmitting the calculated capabilities of the wireless bridge to another entity of the communication system for consideration in managing the time sensitive network.

13. obtaining time-sensitive network scheduling information determined based on the transmitted calculated capabilities; and estimating an adjusted performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on the obtained time-sensitive network scheduling information; calculating an adjusted capacity of the wireless bridge corresponding to the estimated performance capacity; The method of claim 12 further comprising:

14. A communication system in a time-sensitive network comprising a first sub-network and a second sub-network, the communication system comprising a wireless bridge and further comprising at least two end stations; The wireless bridge comprises at least one user plane functional entity, at least one base station, and at least one user equipment; the at least one user plane functional entity comprises at least one network side port enabling communication between the at least one base station and the first sub-network; the at least one user equipment is wirelessly connected to the at least one base station within the time sensitive network; the at least one user equipment comprises at least one device-side port enabling communication with the second sub-network; At least one of the at least two end stations is connected to a corresponding device-side port of a corresponding user equipment; the wireless bridge includes a port pair consisting of the at least one network-side port and the at least one device-side port, the port pair being associated with a delay that is a sum of an independent delay and a dependent delay, the independent delay being a delay that is not affected by an amount of data to be handled, and the dependent delay being a delay that depends on the amount of data to be handled; The wireless bridge comprises: estimating a performance capacity of the wireless connection between the at least one base station and the at least one user equipment based on information regarding a network usage purpose for time-sensitive communication in the wireless bridge; calculating a capability of the wireless bridge corresponding to the estimated performance capacity, the capability calculation including determining, for the port pair, the dependent delay based on the performance capacity of the wireless connection; A communication system configured to:

15. A wireless bridge of the communication system according to claim 14.

Images