AI-ML 5G TSN flow modeling

Abstract

A communication network configured to update a control law of a device for transmitting a control signal based on an artificial intelligence (AI) function. A control signal is transmitted that is configured to support time-sensitive deterministic communications based on a time-sensitive network (TSN) mechanism. The processor transmits quality of service requirement information for transmitting the control signal to the end device, polls a component of the communication network to obtain operating parameter information, transmits the control signal to the end device based on the quality of service requirement information, receives the operating parameter information from the component of the communication network in response to the polling of the operating parameter, and transmits the received operating parameter information to the end device. Feedback information is received from the end equipment; and updating a control law of the device for transmitting the control signal based on an artificial intelligence (AI) function to compensate for a delay in the transmission of the control signal.

Description

Cross-references to related applications

[0001] This disclosure claims the benefit of U.S. Provisional Application Serial No. 63 / 583,811 entitled “AI-ML 5G TSNFlow Modeling”, filed on September 19, 2023, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0002] This specification generally relates to a device and / or system for interacting with a wireless communication network system that integrates a time-sensitive networking (TSN) mechanism, and more specifically, for example, to configuring the device to simultaneously optimize the control law of the controller based on an artificial intelligence (AI) model to control the operation of a target industrial system and TSN scheduling. Background Technology

[0003] Applications such as industrial automation and manufacturing require ubiquitous and seamless connectivity, where communication between various devices or components of the application (e.g., industrial controllers, sensors, actuators, etc.) has stringent deterministic timing requirements. To meet these requirements, TSN systems that provide deterministic communication with relatively stringent Quality of Service (QoS) parameters (e.g., latency, jitter, and reliability requirements for data traffic) can be integrated with wireless communication networks conforming to 3GPP specifications (e.g., 5G, 6G, etc.) that provide highly reliable services, such as Ultra-Reliable Low-Latency Communication (URLLC). Therefore, the controller generates and transmits control signals via the wireless communication network to control the operation of end devices (e.g., actuators, sensors) in the target industrial system (e.g., manufacturing plant, factory, etc.). Summary of the Invention

[0004] Therefore, there is a need for apparatus and methods for transmitting control signals via a communication network configured to support time-sensitive deterministic communication. In one aspect, some embodiments include an apparatus for transmitting control signals via a communication network configured to support time-sensitive deterministic communication based on a Time-Sensitive Networking (TSN) mechanism. The apparatus includes a processor and a memory communicatively coupled to the processor. The memory stores one or more instructions that, when executed by the processor, cause the processor to: send quality of service requirement information to an end device for transmitting a control signal configured to control the operation of the end device in a system associated with the communication network; poll components of the communication network to obtain operating parameter information related to operating parameters of the communication network; send the control signal to the end device based on the quality of service requirement information, wherein the control signal, based on a control law, receives the operating parameter information from components of the communication network in response to polling of the operating parameters and receives feedback information from the end device; and, based on artificial intelligence (AI) functionality, update the control law of the apparatus for transmitting the control signal based on the operating parameter information and the feedback information to compensate for delays in the transmission of the control signal.

[0005] In another aspect, some implementations include a method for transmitting a control signal via an associated communication network configured to support time-sensitive deterministic communication based on a time-sensitive networking (TSN) mechanism. The method includes: sending quality of service (QoS) requirement information to an end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; polling components of the communication network to obtain operating parameter information related to operating parameters of the communication network; sending the control signal to the end device based on the QoS requirement information, wherein the control signal is based on a control law, receives the operating parameter information from components of the communication network in response to polling of the operating parameters, receives feedback information from the end device, and updates the control law of the device for transmitting the control signal based on the operating parameter information and the feedback information using artificial intelligence (AI) functionality to compensate for delays in the transmission of the control signal.

[0006] On the other hand, according to some implementations, a system configured to perform any of the above methods is provided.

[0007] In another aspect, a non-transitory computer-readable storage medium has one or more processors and a memory storing one or more programs executable by the one or more processors. The one or more programs include instructions for performing any of the methods described above. Attached Figure Description

[0008] Some features of this subject matter are set forth in the appended claims. However, for purposes of explanation, several aspects of this subject matter are illustrated in the following figures.

[0009] Figure 1 An example of a conventional integrated TSN-5G system according to one or more implementation methods is illustrated.

[0010] Figure 2 The diagram illustrates a block diagram of an example integrated TSN-5G system architecture according to one or more implementation methods.

[0011] Figure 3 An example of an integrated TSN-5G system according to one or more implementations is illustrated.

[0012] Figure 4 A block diagram of an example TSN block according to one or more embodiments is shown.

[0013] Figure 5 The illustration shows an example set of parameters for an example TSN block according to one or more embodiments.

[0014] Figure 6A , Figure 6B The diagram illustrates a block diagram of an example integrated TSN-5G system architecture including MEC according to one or more implementations.

[0015] Figure 7 The illustration shows an electronic system that can implement one or more embodiments of the technology of this subject.

[0016] Figure 8 The diagram illustrates a block diagram of an example integrated TSN-5G system architecture according to one or more implementation methods.

[0017] Figure 9 An example of a control system according to one or more embodiments is illustrated.

[0018] Figure 10 The diagram illustrates a flowchart of an example operation of a control system according to one or more embodiments.

[0019] Figure 11 An example of the operation of an AI / ML system according to one or more implementations is illustrated.

[0020] Figure 12 An example of a computational graph of a neural network (NN) according to one or more embodiments is illustrated.

[0021] Figure 13 The diagram illustrates a flowchart of an example operation of a control system according to one or more embodiments.

[0022] Figure 14A block diagram of an example control system according to one or more embodiments is shown. Detailed Implementation

[0023] The specific embodiments described below are intended as descriptions of various configurations of the subject matter and are not intended to represent the only configuration in which the subject matter can be practiced. The accompanying drawings are incorporated herein and form part of the specific embodiments. The specific embodiments include detailed descriptions to provide a thorough understanding of the subject matter. However, the subject matter is not limited to the specific details set forth herein and can be practiced using one or more other embodiments. In one or more embodiments, structures and components are shown in block diagram form to avoid obscuring the concepts of the subject matter.

[0024] As mentioned above, for some applications (such as, but not limited to, industrial automation, manufacturing, and airborne communications in aerospace and automotive), a TSN system providing deterministic communication can be integrated with communication networks (such as wireless networks or networks compatible with 3GPP standards, such as 5G networks or sixth-generation (6G) wireless communication systems, or any communication network or system defined according to 3GPP standards or IEEE 802.1 standards). However, typically in such integrated TSN-5G systems, the entire 5G system is configured to operate as a single TSN component (e.g., a TSN bridge), and the integrated system is set up as a fully centralized configuration model, for example, using a centralized TSN configuration controller. Therefore, such integrated systems may not support the flexibility of deploying 5G systems that include various components from different vendors, nor do they support distributed TSN configuration of the integrated system. Additionally, 5G systems can support reliable communication by using redundant paths for data transmission. However, since the 5G system is set up as a single TSN component in a typical TSN-5G integrated system, redundant paths are configured across the entire 5G system through all its components (e.g., from User Equipment (UE) to User Plane Function (UPF)). This does not allow for the flexibility to set up redundant data transmission paths for only a portion of the 5G system (e.g., the portion of the 5G system that is more prone to data latency and / or errors, such as the air interface between the UE and the radio access network (RAN) / gNodeB)).

[0025] To address the aforementioned issues in integrated TSN-5G systems, this subject matter provides a novel architecture in which the 5G system is configured as a set of discrete 5G components (each 5G component being configured as a discrete TSN block), rather than as a single TSN component or block. In other words, this disclosure provides an architecture in which the 5G system integrated into a TSN-5G system is divided into multiple TSN blocks, each TSN block being configured according to TSN specifications (e.g., according to IEEE 802.1Q and related standards) as, for example, a TSN bridge, a TSN end device, or a combination of both. Furthermore, this subject matter provides multiple distributed configuration modules in the TSN-5G system, rather than having a centralized configuration controller control the TSN-5G system. These configuration modules can be interconnected in one or more topologies (including mesh, star, tree, or random), and each configuration module can be responsible for communicating with and configuring one or more TSN blocks.

[0026] As discussed in detail below, each of the plurality of TSN blocks in a 5G system includes a set of parameters describing the capabilities to support and execute data flows (e.g., carrying URLLC data traffic) through that TSN block. Additionally, each TSN block may include at least one configuration interface configured to support interaction between the TSN block and its corresponding configuration module. The configuration interface can be used to provide the parameter set of the TSN block to the corresponding configuration module and receive configuration data (e.g., transport scheduling, data flow identifiers, regulatory rules, etc.) from the configuration module to support one or more data flows through the TSN block. Each TSN block can be configured to perform or operate according to the configuration data (received via the configuration interface) and to transmit data for each data flow according to the specifications provided in the configuration data. Finally, each TSN block may also include a monitoring and diagnostic module configured to monitor the runtime behavior of the TSN block and report this behavior via the configuration interface or a separate interface of the TSN block.

[0027] Each of the plurality of distributed configuration modules may be an external utility responsible for configuring one or more TSN blocks, or may be configured as a software module within a TSN block that configures only that TSN block. Each configuration module may be configured to identify the TSN blocks controlled by the configuration module and provide them with configuration data, including TSN scheduling for one or more data streams to be transmitted through these TSN blocks. The configuration modules may exchange information with each other using a standardized application programming interface (API). The exchanged information may include: information about the cycle time of the TSN system (e.g., supported management cycle times (ACT) including discrete-level ACT buckets, each corresponding to a specific data stream), maximum / minimum cycle time), configuration data including transmission scheduling for one or more TSN blocks (including time offsets / durations / resources for transmission), and information for requesting or responding to resource allocation requests. In some implementations, one or more of the plurality of configuration modules may be adapted as a "controller" or "controller module," which may include components configured or adapted to provide instructions, control, operation, or any form of communication to operational components to influence the operation of those operational components. The controller module may include any known processor, microcontroller, or logic device, including but not limited to: field-programmable gate array (FPGA), application-specific integrated circuit (ASIC), full authority digital engine control (FADEC), aerospace systems, proportional controller (P), proportional-integral controller (PI), proportional-derivative controller (PD), proportional-integral-derivative controller (PID controller), hardware-accelerated logic controller (e.g., for encoding, decoding, transcoding, etc.), or combinations thereof.

[0028] 3GPP Release 8 categorizes self-organizing networks (SONs) into three main types: self-configuration, self-optimization, and self-healing. Self-organization is considered a mechanism or process that enables a system or network to change its organization during its execution time without explicit command. Self-configuration is defined as the process of incorporating new network elements (NEs) into a service, requiring minimal human intervention. A network element is a manageable logical entity that combines one or more physical devices. In some implementations, TSN blocks (discussed below) can be self-configured and incorporated into the 5G system as NEs with minimal human intervention. This can be achieved through TSN applications that automatically share their TSN stream (streaming) characteristics and latency requirements with the TSN block. TSN blocks can cooperate to share stream characteristics and determine feasible TSN scheduling.

[0029] In some embodiments, this disclosure provides a communication network, such as a wireless network, like a fifth-generation (5G) or sixth-generation (6G) wireless communication system or network (or any communication network or system defined according to 3GPP standards or IEEE 802.1 standards), configured to support time-sensitive deterministic communication based on a Time-Sensitive Networking (TSN) mechanism. In the context of this disclosure, any references to “communication network,” “wireless network,” and “communication / wireless network” refer to a network having various components designed, arranged, and / or configured for transmitting data, signals, or information across the network from a source node or device to a destination node or device, wherein at least a portion of the communication is performed wirelessly. Thus, the various components of such a network can be communicatively interconnected via wireless links or channels and / or via wired links or channels. The various components of the network include modules and channels / links implemented in hardware, software, and combinations of hardware and software.

[0030] A communication network may include multiple discrete communication network (CN) components. Each CN component may be configured to provide discrete functionality for data communication across the network from a source device to a destination device. The communication / wireless network may further include a processor arranged to configure at least one (or each) of the multiple CN components using a TSN parameter set of a TSN mechanism. After configuration using the TSN parameter set, the at least one (or each) of the multiple CN components can support time-sensitive deterministic communication as a TSN block in a TSN network, based on the TSN parameter set. In other words, the at least one (or each) of the multiple CN components can operate as a typical TSN block in a TSN network.

[0031] The plurality of discrete CN components in a 5G network may include user equipment (UE), radio access network (RAN), user plane functions (UPF), control plane functions, and transport network channels connecting the CN components, including transport network channels connecting the RAN and UPF. In some embodiments, at least one CN component configured with TSN parameters is a 5G transport network channel connecting the RAN and UPF.

[0032] The communication network may include a configuration module configured to determine a set of TSN parameters for at least one of the plurality of CN components. The TSN parameter set includes one or more of the following: TSN stream identifier, transmission scheduling, deadline or delay budget, filtering configuration, and redundancy scheme. In some embodiments, the configuration module is configured within the Session Management Function (SMF) of the control plane of a 3GPP-compliant communication network.

[0033] In some embodiments, this disclosure provides a communication network, such as a wireless network (e.g., a 5G network), comprising a plurality of configuration modules interconnected in some arrangement, and configured to provide each of the plurality of CN components with a unique set of TSN parameters, such that each of the plurality of CN components supports time-sensitive deterministic communication as a corresponding TSN block in the TSN network based on the corresponding unique TSN parameter set. The plurality of configuration modules may be interconnected in a hierarchical, mesh, star, tree, or combination thereof arrangement.

[0034] In some embodiments, this disclosure provides a system including a first communication (e.g., 5G / 6G) network, a second communication (e.g., 5G / 6G) network, and a TSN transport channel connected to the first and second communication networks. The first network may include a first plurality of discrete communication network (CN) components, each CN component configured to provide discrete functionality for data communication via the first network, wherein at least one of the first plurality of CN components is configured to provide time-sensitive deterministic communication according to a first TSN parameter set. The second network may include a second plurality of discrete CN components, each CN component configured to provide discrete functionality for data communication via the second network, wherein at least one of the second plurality of CN components is configured to provide time-sensitive deterministic communication according to a second TSN parameter set. The TSN transport channel may be configured to facilitate time-sensitive deterministic communication of data exchanged between the first and second networks according to a third TSN parameter set. In some embodiments, the TSN transport channel is communicatively connected to a first TSN converter of the first wireless network and a second TSN converter of the second wireless network.

[0035] In some embodiments, this disclosure provides a system including a first communication (e.g., 5G / 6G) network, a second communication (e.g., 5G / 6G) network, and a TSN transport channel connected to the first and second communication networks. The first network may include a first plurality of discrete communication network (CN) components, each CN component configured to provide discrete functionality for data communication via the first network, wherein at least one of the first plurality of CN components is configured to provide time-sensitive deterministic communication according to a first TSN parameter set. The second network may include a second plurality of discrete CN components, each CN component configured to provide discrete functionality for data communication via the second network, wherein at least one of the second plurality of CN components is configured to provide time-sensitive deterministic communication according to a second TSN parameter set. The TSN transport channel may be configured to facilitate time-sensitive deterministic communication of data exchanged between the first and second networks according to a third TSN parameter set. In some embodiments, the TSN transport channel is communicatively connected to a first TSN converter of the first wireless network and a second TSN converter of the second wireless network.

[0036] Industrial systems can be controlled by control signals provided by the control system and transmitted via wireless communication networks (such as 5G, 6G, etc.) according to standards defined by 3GPP. While wireless communication networks provide the communication needed to interconnect control system components (typically assumed to be Ultra-Reliable Low-Latency Communication (URLLC)), they lack the inherent control system knowledge or capabilities of user applications. Most control systems assume negligible communication latency. As discussed above, wireless communication networks can be integrated with Time-Sensitive Network (TSN) systems to provide bounded, predetermined delays (scheduled flow of traffic) to the control system, preventing it from becoming unstable. The stability of a control system is defined as the ability of any system to provide bounded outputs when bounded inputs are applied to it. If messages experience variable delays or latency in communication, the controller's response will occur too early or too late, causing the system to become unstable (i.e., loss of control over the system with no hope of regaining control). For a wireless communication network (TSN) to succeed, it is crucial that it maintains the stability of the control system in wireless environments susceptible to noise and the greater challenges that lead to variable delays or latency. However, the controller's control law does not account for this time delay or latency (which is a mathematical formula used by the controller to determine the control signal), which may impair the controller's performance. Therefore, a solution is needed to compensate for variable time delay or latency.

[0037] To address the aforementioned issues, this subject matter provides a control system that utilizes an artificial intelligence / machine learning (AI / ML) system associated with a wireless communication network. Therefore, in some embodiments, the control system may include a device for transmitting control signals via a communication network configured to support time-sensitive deterministic communication based on a time-sensitive networking (TSN) mechanism. The device includes a processor and a memory communicatively coupled to the processor, and one or more instructions stored in the memory. When executed by the processor, the one or more instructions cause the processor to: send quality of service (QoS) requirement information to an end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; poll components of the communication network to obtain operating parameter information related to operating parameters of the communication network; send the control signal to the end device based on the QoS requirement information, wherein the control signal, based on a control law, receives the operating parameter information from components of the communication network in response to polling of the operating parameters and receives feedback information from the end device; and, based on artificial intelligence (AI) functionality, update the control law of the device for transmitting the control signal based on the operating parameter information and the feedback information to compensate for delays in the transmission of the control signal.

[0038] Figure 1The illustration depicts a non-limiting example of a conventional integrated TSN-5G system 100 in which the 5G network or system 106 is configured to emulate a single TSN component (e.g., a TSN bridge). Generally, system 100 is configured as a deterministic TSN system to transmit data between end devices (e.g., input / output (I / O) devices 102 and controller 104) via the 5G system 106 (emulated as a TSN bridge) and one or more (conventional) TSN bridges 108 and using a TSN controller 110. System 100 is configured based on standard methods for time synchronization and traffic management, thereby allowing deterministic communication between end devices (e.g., I / O devices 102 and controller 104) over a standard Ethernet network. For example, system 100 can operate according to the IEEE 802.1Q TSN specification suite, which standardizes Layer 2 communication for networking protocols that provide deterministic communication while sharing the same infrastructure. For example, many standards have established various technical paradigms for TSN systems—clock synchronization (802.1AS, General Precision Time Protocol (gPTP)), frame preemption (802.3br and 802.1Qbu), traffic scheduling (802.1Qbv), and redundancy management (Frame Replication and Deletion for Reliability (FRER) IEEE 802.1CB). These standards work together at Ethernet Layer 2 to ensure that control and security functions simultaneously meet their respective deadlines and constraints. As another implementation, similar integrated systems can be configured to implement TSN technology over wireless local area networks (wireless LANs), such as Wi-Fi networks (e.g., based on Wi-Fi 6 and other common wireless LANs).

[0039] For example, the 802.1Qbv TSN standard provides scheduled delivery of security-critical data frames in a predetermined manner, and is incorporated herein in its entirety. As used herein, “TSN mode” can be used without limitation to mean a network, component, element, unit, node, hub, switch, controller, module, path, data, data frame, traffic, protocol, operation, transport, or combination thereof that conforms to, is configured for, or is compliant with one or more standards in the IEEE 802.1 TSN standard. The 802.1Qbv TSN standard addresses the delivery of critical and non-critical data traffic within a TSN. Critical data traffic is guaranteed to be delivered at the scheduled time, while non-critical data traffic is typically assigned lower priority. Various traffic categories have been established according to IEEE 802.1Q for prioritizing different types of data traffic.

[0040] Ethernet frame preemption, defined by the IEEE 802.3br and IEEE 802.1Qbu standards, can suspend the transmission of non-critical Ethernet frames and also helps reduce latency and latency variations in critical traffic. Resource management is based on the TSN configuration model (IEEE 802.1Qcc). For example, as specified in IEEE 802.1Qdj [D1.0, Final Opinion Processing (v1)], centralized network configuration (CNC) 112 can be applied to network devices (bridges, such as 5G system bridges 106 and 108), while centralized user configuration (CUC) 114 can be applied to user equipment (end stations, such as I / O devices 102). The fully centralized configuration model follows a software-defined networking (SDN) approach. In other words, CNC 112 and CUC 114 in controller 110 provide the control plane rather than a distributed protocol. In contrast, distributed control protocols are applied in a fully distributed model, where there may be no CNC or CUC.

[0041] As a result of ultra-reliability, high availability can be provided by Frame Replication and Elimination for Reliability (FRER) (IEEE 802.1CB) for data flows through a per-packet-level reliability mechanism. This provides reliability by transmitting multiple copies of the same data packets on disjoint paths in the network. Flow filtering and policing (802.1Qci) enhances reliability by preventing bandwidth violations, failures, and malicious behavior. Additionally, time synchronization in a TSN system can be defined by the Universal Precision Time Protocol (gPTP) (802.1AS), which is the configuration file for the Precision Time Protocol standard (IEEE 1588). gPTP provides reliable time synchronization and can be used by other TSN tools, such as scheduled traffic (802.1Qbv).

[0042] To achieve the desired level of reliability, TSN employs time synchronization and time-aware data traffic shaping. Data traffic shaping uses scheduling to control gating of transmissions on network switches and bridges (e.g., nodes). In some respects, this data traffic scheduling in the TSN can be determined before network operation. In other respects, data traffic scheduling can be determined based on system requirements during the initial design phase and updated as needed. For example, in addition to defining the TSN topology (including communication paths, bandwidth reservations, and various other parameters), the network-wide synchronization time for data transmission can be predefined. This data transmission plan for communication paths on the network is commonly referred to as "communication scheduling" or simply "scheduling." The scheduling of data traffic on the TSN can be determined for specific packets at specific times, on specific paths, and for specific durations. Non-limiting examples of techniques for generating TSN data traffic scheduling are discussed in U.S. Application No. 17 / 100,356, which is incorporated herein by reference in its entirety.

[0043] Time-critical communication between end devices or nodes in a TSN (e.g., I / O device 102 and controller 104) comprises a “TSN flow,” also known as a “data flow” or simply a “flow.” For example, a data flow can include datagrams, such as packets or data frames. Each data flow is unidirectional, from a first initiating or source end device in the system (e.g., I / O device 102) to a second destination end device (e.g., controller 104), and has a unique identifier and timing requirements. These source and destination devices are typically referred to as the “talker” and the “listener.” Specifically, the “talker” and the “listener” are the source and destination of the data flow, respectively, and each data flow is uniquely identified by the end device operating in the system. It should be understood that for a given network topology comprising multiple interconnected devices, a set of data flows can be defined between the interconnected devices or nodes. For example, this set of data flows can be between interconnected devices. Various subsets or permutations of the data flows can be additionally defined for this set of data flows. Additionally, time-critical communication between end devices or nodes in a TSN includes "TSN streams" or "streams," where each TSN stream can originate from a specific speaking node intended to transmit to one or more listening nodes. Therefore, each TSN stream can include one or more data streams, each between the speaking node (the node from which the TSN stream originates) and the listening node.

[0044] End devices (e.g., 102, 104) and switches (often referred to as "bridges" or "switching nodes") (e.g., 106, 108) transmit and receive data (Ethernet frames in a non-limiting example) in a data stream based on a predetermined time schedule. Switching nodes and end devices must be time-synchronized to ensure that the predetermined time schedule of the data stream is correctly followed throughout the network. For example, in... Figure 1 In this context, clock 116 indicates that the various switching nodes and end devices in the TSN system 100 (including in the 5G system 106) are time-synchronized with reference to a global clock (master clock timing). In some other respects, only the switches can transmit data based on a predetermined schedule, while end devices (such as legacy devices) can transmit data in an unscheduled manner.

[0045] A single device (e.g., controller 110) can be used to schedule data flows within a TSN, taking a fixed, invariant path across the network between the speaking / listening device and the switching nodes in the network. Alternatively, a set of devices or modules can be used to schedule data flows. The scheduling device (whether a single device or a set of devices) can be arranged to define a centralized scheduler. In other aspects, the scheduler device can include a distributed arrangement. The TSN can also receive non-time-sensitive communication, such as rate-limited communication. In a non-limiting example, the scheduling device can include an offline scheduling system or module.

[0046] Various mechanisms can be used to tag TSN traffic, including VLAN tags, Ethernet addresses, IP header information, and combinations of VLAN tags, Ethernet addresses, and IP header information. Traffic can be identified and tagged anywhere in the system before requiring Protocol Data Unit (PDU) identification. TSN speakers can create multiple TSN streams with different TSN latency and deterministic requirements, and these streams can be assigned different paths to meet those requirements. In some embodiments of the invention, latency and deterministic values ​​can be specified and provided to the TSN application as a finite set of static discrete values, rather than providing an infinite set of continuous values.

[0047] In some embodiments, the I / O end device 102 can be a complex mechanical entity, such as a production line in a factory, a gas-fired power plant, an avionics data bus on an aircraft, a jet engine on an aircraft in a fleet (e.g., two or more aircraft), a digital backbone in an aircraft, an avionics system, a mission or flight network, a wind farm, a locomotive, etc. In various embodiments, the I / O end device 102 can include any number of end devices, such as sensors, actuators, motors, and software applications. Sensors can include any conventional sensor or transducer, such as a camera that generates video or image data, an X-ray detector, an acoustic pickup device, a tachometer, a GPS receiver, a wireless device that transmits wireless signals and detects reflections of those signals to generate image data, or other devices.

[0048] Additionally, actuators (e.g., devices, apparatuses, or machines that move to perform one or more operations of I / O device 102) can communicate using TSN system 100. Non-limiting examples of actuators may include brakes, throttle valves, robotic devices, medical imaging equipment, lamps, turbines, etc. The actuator can transmit its status data to one or more other devices (e.g., via TSN system 100 to other I / O devices 102, controller 104). The status data may represent the position, status, health, etc., of the actuator sending the status data. The actuator can receive command data from one or more other devices of TSN system 100 (e.g., other I / O devices 102, controller 104). The command data may represent instructions instructing how or when the actuator should move, operate, etc.

[0049] In some implementations, controller 104 can transmit various data between or among I / O devices 102 via TSN 100. For example, control system 104 can transmit command data to or from one or more devices 102, such as status data or sensor data. Therefore, controller 104 can be configured to control the operation of I / O devices 102 based on data acquired or generated by or transmitted between I / O devices 102, to achieve, for example, automatic control of I / O devices 102 and to provide information to operators or users of I / O devices 102. Controller 104 can define or determine data flows and data flow characteristics within TSN system 100.

[0050] The 5G system 106 within reference system 100, or 5G network or system 106, is a wireless communication network or system for carrying TSN traffic between various TSN end devices (e.g., I / O device 102 and controller 104). In some implementations, 5G system 106 is configured to emulate a TSN bridge (similar to TSN bridge 108 according to the TSN standard discussed above) for each user plane function (UPF). 5G system 106 may be a new radio (NR) network implemented according to the 3GPP 23 and 38 series of specifications (which are incorporated herein in their entirety) and integrated into system 100 according to 3GPP Release 17 23.501 standards (e.g., V17.1.1 and V17.2.0) (which are incorporated herein in their entirety). As shown in the figure, the 5G system 106 may include various communication network (CN) components, such as User Equipment (UE) 118, RAN (gNB) 120, and User Plane Functions (UPF) 122 in the 5G user plane, and Application Functions (AF) 124, Session Management Functions (SMF), and Policy Control Functions (PCF) 126 in the 5G control plane, as defined in the 3GPP 23.501 standard. In some implementations, the 5G system 106 may be configured to provide Ultra Reliable Low Latency Communication (URLLC) services. The 5G system 106 based on the New Radio (NR) interface includes several functions for achieving low latency for selected data streams. NR enables shorter time slots in radio subframes, which is beneficial for low-latency applications. NR also introduces micro-slots, where priority transmission can begin without waiting for slot boundaries, further reducing latency. As part of prioritizing and faster radio access URLLC traffic, NR introduces preemption—where URLLC data transmission can preempt ongoing non-URLLC transmissions. In addition, NR employs very fast processing, enabling retransmissions even within short latency limits.

[0051] In some implementations, 5G defines an ultra-robust transmission mode to improve the reliability of both data channels and control radio channels. Reliability is further enhanced through various techniques, such as multi-antenna transmission based on multiple-input multiple-output (MIMO), the use of multiple carriers, and packet duplication on independent radio links.

[0052] Embedding time synchronization as an essential part of operation in 5G cellular radio systems is a common practice in earlier generations of cellular networks. The radio network components themselves are also time-synchronized, for example, through precise time protocol telecommunications profiles, such as based on the 5G internal system clock 190. This provides a good foundation for providing synchronization for time-critical applications. For URLLC services, the 5G system 106 uses time synchronization for its own operation, as well as for multiple antennas and radio channels to provide reliability. In addition to 5G RAN features, the 5G system 106 can also provide solutions for Ethernet networking and URLLC in the core network. The 5G core network supports native Ethernet Protocol Data Unit (PDU) sessions. 5G assisted establishment of redundant user plane paths via 5GS (including RAN, core network, and transport network). 5GS also allows for redundant user planes between RAN and core network nodes and between UE and RAN nodes, respectively.

[0053] As described above, in the integrated system 100, the 5G system 106 includes a TSN (virtual) bridge for each UPF. The 5G system 106 includes a TSN converter (TT) function to adapt the 5G system 106 to the TSN domain (both user plane and control plane), thereby hiding the internal processes of the 5G system 106 from the TSN bridging network. The 5G system 106 provides TSN bridge inbound and outbound port operation through the TT function. For example, TT supports hold and forwarding functions for jitter reduction. Figure 1 The illustration shows a scenario where the 5G system 106 connects the terminal station 102 to the bridging network 108; however, the 5G system 106 can also interconnect the bridges 108.

[0054] For the 5G system 106 to be integrated into the TSN system 100, the TSN streaming requirements can only be met if resource management allocates network resources for each hop along the entire path. This is achieved through the 5G system 106 in conjunction with a configuration controller (e.g., a centralized configuration controller 110 (including CUC 114 and CNC 112)) and / or a set of distributed controller modules (e.g., as described below regarding...). Figure 2The interaction between the discussed components is used to achieve this. The interface between the 5G system 106 and the CNC allows the CNC 112 to learn the characteristics of the 5G virtual bridge and allows the 5G system 106 to establish a connection with specific parameters based on the information received from the CNC 112. Bounded latency requires deterministic latency from 5G and QoS alignment between the TSN domain and the 5G domain. For example, if the 5G virtual bridge acts as a TSN bridge, the 5G system 106 simulates time-controlled packet transmission conforming to, for example, 802.1Qbv scheduled traffic. For the 5G control plane, the TT in AF 124 receives transmission time information for the TSN traffic category from the CNC 112. In the 5G user plane, the TT at UE 118 and the TT at UPF 122 can be adjusted accordingly for time-based packet transmission. As part of the QoS alignment between the TSN domain and the 5G domain, different TSN traffic categories can be mapped to different 5G QoS indicators (5QIs) in AF 124 and PCF 126, and the different 5QIs are processed according to their QoS requirements.

[0055] Regarding time synchronization, the 5G system 106 can implement gPTP on the connected TSN network. The 5G system 106 can act as a virtual gPTP time-aware system and supports forwarding gPTP time synchronization information between the terminal station 102 and the bridge 108 via the 5G user plane TT. All various 3GPP and TSN standards mentioned in this disclosure are incorporated herein by reference in their entirety.

[0056] Now for reference Figure 2The diagram illustrates a block diagram of the architecture of system 200 according to some embodiments of the subject matter. In a broader sense, system 200 depicts a block diagram of an example embodiment of an integrated TSN-5G system similar to system 100 described above, and also includes physical components similar to those in system 100 discussed above. However, unlike system 100, system 200 provides a novel architecture for an integrated TSN-5G system in which 5G system 106 is configured as a set of discrete 5G components, each of which is configured to emulate a discrete TSN block or element 202. In other words, in system 200, 5G system 106 is configured as a de-aggregated structure comprising multiple TSN blocks 202-1 to 202-N, wherein each TSN block 202 is configured according to TSN specifications (e.g., according to IEEE 802.1 and the relevant standards discussed above) as, for example, a TSN bridge, a TSN end device (i.e., as a TSN speaker and / or TSN listener), or a combination of both. Furthermore, this subject matter provides multiple distributed configuration modules 215 in the distributed controller 210 of the TSN-5G system 200, instead of having a centralized configuration controller 110 control the TSN-5G system. Configuration modules 215-a to 215-f can be interconnected in one or more topologies (mesh, star, tree), and each configuration module 215 can be responsible for communicating with and configuring one or more TSN blocks 202. In some implementations, one or more of the configuration modules 215-a to 215-f can be implemented within or as part of the functions of the control plane of the 5G system 106 (e.g., SMF 126). As used herein, "topology" can refer to one or more arrangements of a network, which may include multiple nodes in the network (e.g., transmitting devices, receiving devices, switches, or bridges) and connections between nodes (e.g., communication links or "hops," including wired or wireless communication links). Each link can communicatively couple a corresponding pair of nodes. A set of links can be sequentially coupled via their respective nodes to define, for example, a link path between an originating node and a destination node. The topology may include, but is not limited to, one or more of the following: mesh topology, star topology, bus topology, ring topology and tree topology.

[0057] In some implementations, system 200 is configured to support and manage deterministic TSN data flows between data source 204 (“source device”) and data destination device 206 (“destination device”) via 5G system 106, based on a TSN configuration including TSN scheduling determined by one or more of the configuration modules 215. Data source 204 and data destination 206 may include one or more of I / O device 102 and controller 104. Although not shown, system 200 may also include TSN bridge 108 and other TSN components.

[0058] In some implementations, in the de-aggregation structure, each of the plurality of TSN blocks 202-1 to 202-N may correspond to a 5G system (e.g., Figure 1 A specific component of the 5G network or system 106 shown and discussed above. For example, as Figure 3 As shown, UE 118 can be configured to emulate TSN block 202-1, RAN 120 can be configured to emulate TSN block 202-2, the 5G transport network link 140 between RAN 120 and UPF 122 can be configured to emulate TSN block 202-3, UPF 122 can be configured to emulate TSN block 202-4, and the core network and / or other typical components of the 5G system (e.g., fronthaul, backhaul, multi-access edge computing (MEC) modules) can be configured to emulate one or more TSN blocks 202-N. Each TSN block 202-1 to 202-N is configured according to TSN specifications (e.g., according to IEEE 802.1 and related standards discussed above) as, for example, a TSN bridge, a TSN end device (TSN speaker and / or TSN listener), or a combination of both.

[0059] In some implementations, such as Figure 4 As shown, each TSN block 202 includes a processor 402, a memory device 404, an internal configuration interface (ICI) 406, a transmission module 408, a reporting module 410, and TSN converters (TT)-1 412-1, TT-2 412-2, and TT-3 412-3. The processor 402 may be a microprocessor or multi-core processor, an integrated circuit, a field-programmable gate array, etc., which processes TSN configuration data and executes instructions (e.g., instructions stored in memory device 404) to process and transmit TSN data traffic from one or more TSN data streams according to the TSN configuration data.

[0060] Memory device 404 may store a set of parameters describing the ability to support and execute data flows (e.g., carrying URLLC data traffic) through the corresponding TSN block 202. In some implementations, this set of parameters includes, but is not limited to, identification, link quality, and link bandwidth. Identification parameters may include device type (i.e., whether TSN block 202 is a TSN bridge or a TSN end station). Latency parameters may include at least port-to-port (from the beginning to the end of the TSN block) latency and latency variation (often referred to as "jitter"). Link quality parameters may include at least packet error rate. Link bandwidth parameters may include at least available bandwidth in bits per second.

[0061] In some implementations, the parameter set for TSN block 202 may include a subset of parameters specific to the 5G RAN, including short transmission time intervals, TSC auxiliary information (TSCAI), configuration authorization (CG) information, semi-persistent scheduling (SPS) allocation, and / or other parameters as specified, for example, in 3GPP TS 28.540. Additionally, in some implementations, the parameter set for TSN block 202 may include a subset of parameters specific to TSN, including time synchronization attributes, scheduled transmission (Qbv) attributes, and redundancy attributes (including the number of connected RANs, the number of paths to the UPF, path diversity, the number of available frequencies, propagation characteristics, available radios, and different physical media (e.g., free-space optics)). As an example, Figure 5 The illustration shows the example parameter set 502 used for example TSN block 202.

[0062] In some implementations, the parameter set for TSN block 202 may define worst-case time synchronization error, worst-case gate operation error, maximum gating list size, maximum period time, maximum gate interval duration, transmission start delay, etc., or combinations thereof. This parameter set may include a discrete deterministic set of parameters that varies depending on the type of traffic processed by TSN block 202. This parameter set can be used at least in part to generate TSN scheduling, configuration, etc., that are hardware-implementable for TSN block 202. In the absence of such a parameter set, the TSN scheduling module must employ a least common denominator approach, where all devices in system 200 (including TSN block 202) are assumed to have the most restrictive characteristics, leading to suboptimal solutions. In this sense, the parameter set enables better scheduling solutions to be implemented in TSN system 200, resulting in improved performance metrics, including latency, jitter, packet delay variation, and bandwidth utilization.

[0063] In some implementations, the parameter set for TSN block 202 may further describe or relate to devices created, programmed, or otherwise operated by different operators or manufacturers. In this sense, the parameter set may define a collection or subset of different devices (e.g., heterogeneous devices from multiple vendors) rather than homogeneous or all-similar devices. For example, the parameter set may include, but is not limited to, definitions of specific configuration models, error tolerances, hardware limitations, software limitations, and firmware options for each end node and switching node in system 200. In this sense, the parameter set enables the scheduling and configuration of heterogeneous networks comprising devices from multiple vendors and with different characteristics.

[0064] In some implementations, the parameter set for TSN block 202 can further define specific TSN features supported by each end node and switching node in TSN system 200. For example, the parameter set can define whether a node or TSN block 202 supports one or more of the following: time synchronization, time-aware shaping, asynchronous shaping, frame duplication and elimination for reliability, frame preemption, incoming policing, and other TSN features. The parameter set can further define specific versions or variants of features or standards supported by the end nodes and switching nodes in TSN system 200. These parameter sets enable the scheduling and configuration of hybrid capability networks where end nodes and switching nodes have varying degrees of support (including no support) for desired TSN features and versions.

[0065] Further non-limiting examples of the parameter set for the TSN block 202 device may include, but are not limited to, additional parameters for programming the functionality of the corresponding TSN block 202. For example, the additional parameter set may define or enable programming of the corresponding TSN block 202 using a generated or scheduled TSN configuration. Non-limiting examples of parameter sets that can define or enable programming of the corresponding TSN block 202 may include, but are not limited to, programming methods, communication protocols, device login names, device login passwords, device programming ports, device programming file structures or file paths for programming data, device programming file formats, device configuration file formats, device scheduling file formats, and combinations thereof. In this sense, the parameter set or subset of parameters defines or enables programming of the corresponding TSN block 202 using a generated or scheduled TSN configuration (collectively, “programming parameters”), and enables or allows system 200 to update, install, program, configure, or otherwise modify a set of TSN blocks 202 in response to or according to the scheduling or configuration of TSN system 200.

[0066] Additionally, each TSN block 202 may include at least one Internal Configuration Interface (ICI) 406 configured to support interaction between the TSN block 202 and, for example, its corresponding configuration module 215. The ICI 406 may be used to provide some or all of the parameter set of the TSN block 202 to the corresponding configuration module 215 and to receive configuration data (e.g., transport scheduling, data stream identifiers, policing rules, etc.) from the configuration module 215 to support one or more data streams passing through the TSN block 202. Each TSN block 202 may be configured to perform or operate according to the configuration data (received via the ICI 406) and to transmit data for each data stream according to the specifications provided in the configuration data.

[0067] In some implementations, the configuration data received at TSN block 202 from its corresponding configuration module 215 via ICI 406 may include stream identification information, transmission scheduling or deadline or delay budget or (for rate-limited traffic) data rate, filtering and policing configuration information, redundancy schemes and / or other TSN configuration information.

[0068] TSN speaker information can be divided into different frequency components with varying TSN stream delay and deterministic requirements. Conversely, TSN streams from different TSN speakers can be aggregated into a single TSN stream to achieve greater capacity and higher channel utilization. Discrete cycle times in integrated TSN-5G systems help ensure the ease of TSN stream aggregation.

[0069] In some implementations, each TSN block 202 may include a transmission module 408 configured to transmit a specific data stream based on a schedule, deadline, delay budget, or data rate received from configuration data received via ICI 406 from configuration module 215. In one example where TSN block 202-1 corresponds to UE 118 of a 5G system, transmission module 408 is configured to transmit data based on resource scheduling of the 5G air interface (between UE 118 and RAN 120). The resource scheduling of the 5G air interface may be phase-aligned with the cycle time of the integrated TSN-5G system. Transmission module 408 may assign resource elements of the 5G component corresponding to TSN block 202 (of which 408 is) in a manner that allows scheduled transmission of each Qbv stream to be satisfied. For example, UE 118 corresponding to TSN block 202-1 may use a phase offset (reference cycle time) to align the transmission of TSN data streams with the assigned schedule, rather than using classic 802.1 Qbv-style gating. In this example, UE 118 can obtain this phase offset from RAN 120 as a special command, instead of UE 118 obtaining the phase offset from CNC.

[0070] In some implementations, each TSN block 202 may also include a reporting module 410 configured to monitor the runtime behavior of the TSN block 202 and report that behavior via ICI 406 or a separate interface of the TSN block 202. This behavior monitoring includes metrics, including but not limited to packet loss and missed transmission windows.

[0071] Return to reference Figure 2This technical subject provides multiple distributed configuration modules 215 in a distributed controller 210 of a TSN-5G system 200. Configuration modules 215-a to 215-f can be interconnected in one or more topologies (mesh, star, tree), and each configuration module (CM) 215 can be responsible for communicating with and configuring one or more TSN blocks 202. For example, CM 215-a can be responsible for TSN blocks 202-1 and 202-2 and is operatively and communicatively connected to these TSN blocks. CM 215-b can be responsible for TSN block 202-3 and is operatively and communicatively connected to this TSN block. CM 215-c can be responsible for TSN blocks 202-3 and 202-N and is operatively and communicatively connected to these TSN blocks. Therefore, TSN block 202-3 can be configured and controlled by both CM 215-b and CM 215-c. For example, a portion of the functionality at TSN block 202-3 (e.g., regarding TSN applications of the first type) can be configured and controlled by CM 215-b, while another portion of the functionality at TSN block 202-3 (e.g., regarding TSN applications of the second type) can be configured and controlled by CM 215-c.

[0072] exist Figure 2 In the example shown, configuration modules 215 are arranged in a tree structure, wherein CM 215-f forms the highest level of the tree structure, CM 215-a, 215-b, and 215-c form the lowest level of the tree structure, and CM 215-d and 215-e are located between the highest and lowest levels of the tree structure. However, configuration modules 215 are operatively and communicatively connected to each other via API 230. In some embodiments, in the tree structure (e.g., as...), Figure 2 As shown), configuration module 215 can communicate with other configuration modules 215 at adjacent tree levels (one tree level higher or lower). However, in other topologies (e.g., in mesh or peer-to-peer structures), any two configuration modules 215 of the distributed controller 210 can be directly and operatively connected to and communicate with each other.

[0073] In some implementations, each configuration module 215 may be an external utility responsible for configuring one or more corresponding TSN blocks 202, or it may be configured as a software module within the TSN block 202. Each configuration module 215 may be configured to identify the TSN block 202 controlled by the configuration module 215 and provide it with configuration data, including TSN scheduling for one or more data streams to be transmitted through the TSN block 202. The configuration modules 215 may exchange information with each other using a standardized API 230. The exchanged information may include: information about the cycle time of the TSN system (e.g., supported management cycle times (ACT) including discrete-level ACT buckets, each corresponding to a specific data stream), maximum / minimum cycle time), configuration data including transmission scheduling for one or more TSN blocks 202 (including time offsets / durations / resources for transmission), and information for requesting resource allocation or responding to resource allocation requests. In some implementations, one or more of the configuration modules 215 may be configured as the CNC 112 and / or CUC 114 of the TSN controller 110 discussed above. In some implementations, one or more of configuration modules 215-a to 215-f may be implemented within or as part of a control plane function (e.g., SMF 126) and / or user plane element or function (e.g., 5G transport network 202-3) of the 5G system 106. For example, SMF 126 may be configured to support the functions of CUC 114 and / or CNC 112. In some implementations, 5G transport network 202-3 may be configured to support the functions of CNC 112.

[0074] In some implementations, each configuration module (CM) 215 receives some or all of the parameter set of TSN block 202 (discussed above) from ICI 406 of TSN block 202 controlled by CM 215. CM 215 also receives information about the data flow to be configured via TSN block 202 from another entity in system 200 via API 230. In a non-limiting example, data source 204 and / or data destination 206 provide CM 215 with requirements for the data flow between data source 204 and data destination 206, directly or via an intermediary such as CUC 114. In some implementations, the CM itself may allow the user to define a set of data flows to be configured via a user interface. As used herein, information about the data flows may include or define a set of data flows, data streams, transport paths (pre-defined or otherwise adapted), etc., to define the desired TSN communication path between data source 202 and data destination 206. A set of non-limiting examples of information about a data stream may include maximum permissible latency, data rate, data frame size (“payload”), data frame destination, bandwidth allocation gaps, etc., or combinations thereof.

[0075] Based at least on the received data stream information and the parameter set for TSN block 202, CM 215 determines a "solution" (or configuration data) that indicates how to process each data stream passing through TSN block 202. This solution may include time-aware scheduling, regulatory rules, etc., as discussed in U.S. Application No. 17 / 100,356, which is incorporated herein by reference in its entirety. CM 215 can then send the solution or configuration data to TSN block 202 via ICI 406. TSN block 202 can then execute the solution and transmit data for each stream according to its configuration. In some implementations, as an example process for enabling the distributed CM 215 to compute solutions, the same System Modulo Theory (SMT) solver can be used at each level of the tree structure of CM 215, where data streams and their requirements are expressed as constraints, and a linear programming method is used to solve for feasible solutions. Solutions from lower levels of the CM tree are input as used resources (again represented by constraints) to higher levels of the CM tree. Repeat this process until the top level of the CM tree is reached, where the global solution is determined.

[0076] In some implementations, different data sources (e.g., data source 204) and their applications operate with different periodic times or intervals. Correspondingly, different data sources and destinations (and their applications) require different levels of time determinism for the numerous data streams between them. In conventional TSN systems, a convergence periodic time (often referred to as the “administrative periodic time”) is determined for all data streams in the network. However, in some implementations disclosed in this subject matter, integrated TSN-5G systems can use a set of discrete / quantized periodic times in the network. Each data stream selects one of the available quantized periodic times to operate. Scheduling of scheduled transmissions in TSN-5G system 100 can be based on a set of quantized / discrete periodic times. As an example, integrated TSN-5G system 100 can limit the available stream intervals and therefore the corresponding periodic times to a discrete set of values, including but not limited to substantially 1, 10, 100, 1000 milliseconds. Similarly, stream or data stream requirements can be limited; for example, jitter (packet delay variation) requirements can be limited to a predetermined discrete set of values, including but not limited to substantially 1, 10, 100, 1000 microseconds. In some implementations, different discrete value sets may be used depending on the applications and use cases supported by the integrated TSN-5G system. For example, geographically dispersed systems may use discrete cycle times on the order of milliseconds. In yet another example, systems confined to a local facility may use discrete cycle times on the order of microseconds. Members of this discrete set may be regularly or irregularly spaced or follow other statistical distributions (including but not limited to logarithmic, linear, and Gaussian), rather than being assumed to be a continuous set of cycle times. In another implementation, a set of cycle times is normalized such that all TSN blocks have a cycle time as the product of elements selected from a smaller common set of prime numbers. This ensures that the TSN scheduler will easily compute all composite cycle times and produce a common network cycle time.

[0077] Each TSN block in an integrated TSN-5G system can support a set of period times (where a set may include one or more). CM 215 configures TSN-5G system 106 or 200 to enable scheduled transmission of data streams across TSN blocks 202 operating with different period times. In some implementations, TSN blocks 202 may need to operate with compatible period times, where compatibility means that the period times are integer harmonics of each other. When an application requests an interval that is not directly mapped to a set of discrete period times available on a set of TSN blocks 202 traversed by the stream, CM 215 can fit to the nearest available period time. The nearest available period time will be an integer multiple or integer divisor of the available period times. CM 215 can exchange information about the supported quantized / discrete set of period times with each other during the configuration process. In this case, this subject matter disclosure allows de-aggregated TSN-5G systems to create feasible configurations for large data streams / streams. In the absence of quantized period / intervals, configuration typically requires significant computation time and may even prevent the discovery of feasible solutions.

[0078] In some instances, each integrated TSN-5G network slice in 5G system 106 may have a predefined set of supported period times and jitter limits. In some instances, network slicing may be more granular than a typical 5G network slice according to 3GPP specification 23.501, which is incorporated herein by reference. In some implementations, TSN-5G systems 106 or 200 may be sliced ​​based on TSN period times. For example, a 5G network supporting multiple critical services may have periodic URLLC slices dedicated to applications and their streams. For example, a service with an application operating at a period of approximately 1 millisecond may have a dedicated slice operating at a period time of 1 millisecond. Similarly, services and applications operating at a period (or interval) of 100 milliseconds may have dedicated slices operating at a period time of 100 milliseconds in the integrated TSN-5G system. Such period time slicing improves configuration speed and overall network performance. In some implementations, TSN block 202 exposes the supported period times of a given slice to its corresponding CM 215 via its ICI 406. CM 215 can then exchange the supported cycle times of the TSN blocks under its management with each other via the configuration module inter-module API 230 in order to create a configuration solution for the sliced ​​TSN-5G system.

[0079] In some implementations, the solution or configuration data determined by CM 215 may include, but is not limited to, a set or group of configurations, timings, commands, controls, instructions, etc., or combinations thereof, for operating the corresponding TSN block 202 according to its characteristics (e.g., characteristics defined by a parameter set). In some aspects, the configuration data may include specific transmission information for individual or collective (e.g., “global”) data frame transmissions of one or more corresponding TSN blocks 202. The transmission information may include timing information for transmitting the data frames. In one or more aspects, the configuration data for the data frames may include a transmission start time. For example, the transmission start time may be the time from which data frame transmission of the corresponding TSN block 202 is initiated. In one aspect, data frame transmission can be initiated by selectively opening a gate in the corresponding TSN block 202 for transmitting the data frame as a data stream to a destination node (e.g., another TSN block 202). Conversely, data frame transmission can be stopped or blocked by selectively closing a gate in the corresponding TSN block 202 for transmitting the data frame. The configuration data can also define or assign specific paths or links that communicatively couple the corresponding TSN block 202 and another node for transmitting data streams. Additionally, the configuration data can define the duration for transmitting a corresponding data stream from the corresponding TSN block 202. In one aspect, the duration of data stream transmission can be defined by the time interval between selectively opening the door of the corresponding node (i.e., to transmit data frames) and selectively closing the door (i.e., to stop transmitting data frames to the destination node).

[0080] In traditional TSN systems, TSN scheduling is represented as an absolute time offset within a periodic period, under which a TSN block is instructed to transmit the data. However, this can be too restrictive for a 5G system 106 composed of components from multiple vendors. In some implementations, deadline-based scheduling is determined by CM 215 and provided along with configuration data. Deadline-based scheduling can instruct TSN block 202 to transmit the configured data stream no later than the deadline (represented by absolute time within a periodic period). In some implementations, a delay budget-based approach instructs TSN blocks to transmit the configured data stream's data frames within the delay budget. Therefore, under the delay budget-based approach, TSN block 202 is required to send data frames arriving at its inbound port to its outbound port within a certain duration. This scheme does not require each TSN block 202 to be time-synchronized. In some implementations where TSN block 202 is configured under rate constraints, TSN block 202 is configured to transmit data frames of a given data stream at an average or peak transmission rate (in bits per second) not exceeding a configured value (based on configuration data from CM215).

[0081] In some 5G systems, a TSN block can be a set of shared resources available through network slicing based on service profiles that define network latency and periodicity (including but not limited to latency / budget). In such implementations, two levels of scheduling can exist, where, in addition to slice-level TSN scheduling, 5GS TSN-AF can also enable the configuration of TSN block 202 as a shared resource. In either case, configuration attributes such as resource identifiers can identify the TSN block used for appropriate configuration. As an example, a service provider can have multiple service profiles with specific periodicity, and multiple tenants of the service provider can utilize the same TSN blocks as specified by the 5GS TSN-AF configuration and can perform TSN flow aggregation. In some implementations, a service provider can provide a set of non-shared TSN blocks, where only a single-layer service profile may exist. Device-specific operational / required resource sharing modes can be made available to CNC 112 via TSN-AF.

[0082] As an example implementation of the deadline / delay budget method for TSN block 202, when a data frame arrives at the incoming port of TSN block 202, TSN block 202 records the arrival time of the data frame using its local clock. TSN block 202 can then identify the frame as belonging to the configured data stream and can then start a countdown timer equal to the configured delay budget for that data stream. Using transmission module 408, TSN block 202 can prioritize transmitting data with the least remaining time. If a packet's timer expires before it is transmitted, the event is recorded as a missed transmission and included in the monitoring metric by recording module 410.

[0083] In some implementations, scheduling transfers between TSN block 202-1 (corresponding to UE 118) and TSN block 202-2 (corresponding to RAN 120) may involve "enhanced" allocation (assignment and transfer) of uplink and downlink transfers between UE 118 and RAN 120 to satisfy the scheduling assigned by CM 215-a to TSN block 202-1 corresponding to UE 118. In some implementations, CM 215-a may consider the buffer state and radio conditions of UE 118, as reported by RAN 120, when instantiating the TSN schedule, and may adjust or report desired changes to the requested schedule. In some other implementations, CM 215-a may send real-time feedback to the master CM (e.g., CM 215-d) regarding radio conditions received from RAN 120. This feedback loop may support recalculation of the TSN schedule to satisfy the packet delay budget at that particular TSN block or between TSN blocks on a given end-to-end path.

[0084] In this implementation, link quality can be monitored, and CM 215-a can continuously adjust radio resources in the configuration data to meet transmission scheduling. Radio resources may include, but are not limited to, logical channels, transmit power, and UE-specific time slot durations. In some implementations, fixed / deterministic uplink and downlink time slots for a given UE 118 can be statically assigned, such that, for example, all UEs connected to a given RAN slice are assigned scheduled transmission time slots. 5G native over-the-air scheduling can be used to determine whether a transmission from UE 118 will meet a transmission deadline. If not, UE 118 can request an upgrade to RAN 120 to achieve the scheduled transmission. In some implementations, scheduling of transmissions in 5G system 106 can be based on a quantized / discrete set of periodic times, wherein the set includes at least a 100-millisecond management periodic time. According to the subject matter, the radio link between the UE (e.g., represented by TSN block 202-1) and the RAN (e.g., represented by TSN block 202-2) can be sliced ​​based on periodic times. In some implementations, the uplink and downlink between TSN block 202-1 and TSN block 202-2 may have radio resources allocated to each network slice based on slice period time. For example, a 1-millisecond period time slice would require radio resources (channel, air interface time, etc.) capable of transmitting data at a rate of 1 millisecond.

[0085] In some implementations, 5G resource elements (e.g., frequencies and time slots) can be scheduled such that they satisfy TSN flow latency requirements in addition to meeting "standard" 5G scheduling traffic priority requirements. More specifically, 5G time slots can be allocated for TSN flows such that these time slots transmit TSN flow messages at appropriate periodic time offsets (phase) and within the time constraints (TSN window time) required for TSN scheduling. In this case, the 5G scheduler differs from a "traditional" TSN Ethernet port in that multiple messages can be transmitted simultaneously if transmitted on different frequencies. In some aspects, in the presence of poor RF channel conditions, the 5G system 106 can send multiple copies of messages on different frequencies to increase the probability of meeting transmission scheduling and / or deadlines.

[0086] To achieve reliable data transmission in system 200, redundant flow paths can be implemented. The de-aggregation of TSN blocks 202 in 5G system 106 enables better handling of errors (delayed, dropped, or corrupted frames). In some implementations, UE 118 can initiate two redundant disjoint PDU sessions to UPF 122 for redundancy; in this case, 5GC can configure NG-RAN for dual connectivity according to 3GPP 38.300. In some other implementations, FRERs can be used between some TSN blocks 202 but not between others. For example, redundant streams can be implemented on the air interface between UE 118 (TSN block 202-1) and RAN 120 (TSN block 202-2) and then combined at RAN 120, and can be further split on the core network (TSN blocks 202-3, 202-4) if necessary. Figure 1 As shown, the current redundancy requirement according to 3GPP 23.501 is the maximum number of disjoint paths between UE 118 and UPF 122. However, according to the techniques of this subject matter, it is not necessary to establish redundant disjoint paths across the entire 5G system; instead, it can be implemented only for a portion of the 5G system. For example, in the case of the air interface between UE 118 and RAN 120, the redundancy requirement can specify that the two paths should be on different frequencies, different MIMO channels, or different time slots. In some instances, this subject matter discloses that allows for flexible use of redundancy, including but not limited to more than two data paths, merging and splitting data streams between TSN blocks, and supporting TSN blocks with different levels of redundancy capabilities.

[0087] The concept of dividing 5GS into multiple TSN blocks (e.g., as mentioned above) Figure 2As discussed, if strict scheduling can prevent malicious traffic from flowing between the blocks, security can potentially be enhanced. However, enabling 5GS to operate as multiple TSN blocks can introduce security vulnerabilities, primarily through configuration. Specifically, TSN users may learn internal 5GS details and connectivity affecting other users sharing the same physical and logical infrastructure. This can occur during the required TSN network discovery phase (e.g., via information returned by the Link Layer Discovery Protocol (LLDP)). TSN users may also misconfigure their own or other users' configurations. This can be partially addressed using the NETCONF concept of a secure subtree, where users have a limited view of their own data model subtree. 5G systems can inherently have the concept of isolated network slices, depending on whether, for example, the TSN is implemented virtually (via software) or physically (via hardware). Infrastructure providers may have to set restrictions on the physical and logical functions open to each TSN user. This can be done via 5G Network Open Functions (NEF). This topic discloses a solution to the security problem by using "virtual TSN blocks." A virtual TSN block is an internal 5G TSN block that contains only the capabilities provided by the user's 5G network slice. In other words, users can only see and configure the TSN block information exposed via 5G network slices, but cannot see or configure anything more. In this sense, the internal 5G TSN block is the intersection of a set of information contained in the internal TSN block and the 5G network slices provided to the user.

[0088] In some implementations, the IETF DETNET standard (as provided at IETF: “Deterministic Networking Working Group” (https: / / datatracker.ietf.org / wg / detnet / about / )) can be implemented or integrated into 5GS to interconnect smaller Ethernet islands compliant with TSN. 5GS can leverage DETNET to enable the transmission of TSN messages over IP Layer 3 (instead of Layer 2). Envisioning such a system, the technologies discussed in this disclosure include 5GS supporting DETNET edge nodes, relay nodes, and hop nodes that interconnect spatially separated TSN networks to create a larger composite TSN over 5G. All aspects of the integrated TSN-5G system provided in this disclosure are applicable to (but not limited to) 5G DETNET TSN islands, and particularly to TSN blocks that de-aggregate TSNs.

[0089] DETNET consists of the following components: (1) TSN end system: an IEEE-compliant end system that communicates with DETNET edge nodes; (2) DETNET edge node: processes TSN frames into DETNET; (3) DETNET relay node; (4) DETNET relay node: provides congestion avoidance for time-sensitive messages. DETNET is routed rather than bridged, thus enabling routable TSN messages between TSN LANs. During transmissions between LANs, critical TSN Ethernet frame information is either transmitted or reconstructed. DETNET achieves its higher-level functionality by adding sublayers: (1) DetNet service sublayer: provides DetNet services (e.g., service protection) to higher layers in the protocol stack and applications, and (2) DetNet transport sublayer: provides DetNet service support for DetNet flows in the underlying network (e.g., by providing explicit routing and congestion protection) and encapsulates TSN Ethernet frames. DETNET routing incorporates IP headers modified according to standard router behavior, such as Time to Live (TTL) handling, where TTL specifies the maximum time a routable IP message is allowed to live, which is clearly related to the maximum latency requirement of TSN.

[0090] DETNET components can reside within any computing element of the 5GS, specifically within the 5G MEC or core. Therefore, in one implementation, the 5GS can be a fully DETNET that interconnects TSN LANs. To provide determinism, DETNET can reserve data plane resources for DetNet flows in some or all intermediate nodes along the flow path. DETNET can provide explicit routing for DetNet flows. DETNET can distribute data from DetNet flow packets temporally and / or spatially to ensure that data for each packet is delivered even if a path is lost. Therefore, as mentioned above, the TSN CNC / DNC and scheduler may require interaction with DETNET, specifically to configure flow paths (including redundant flow paths), TTL, and the ability to obtain routing latency and jitter. TSN traffic shapers, time-aware shaping, and network calculus can be used to achieve the required level of determinism in 5G DETNET.

[0091] It should also be noted that a hybrid 5GS, consisting of part TSN (Layer 2) and part DETNET (Layer 3), can coexist and interoperate. In this case, the DETNET portion of the network can itself be considered as TSN block 202, as discussed above.

[0092] Furthermore, regarding caching within 5GS to minimize latency and increase the determinism of TSN applications, the storage, location, and naming of information within 5GS can be managed to minimize jitter in a way that enables fast and shorter proximity access. Each message can be cryptographically signed to enhance its security, and access can be provided through the hash of the cryptographic signature and information stored in 5GS components such as MEC. The cache forwarding unit will track each data request to allow for optimal forwarding, placement, and service of cached data. The TSN CNC / DNC and scheduler can calculate where to locate cached information within the network (specifically 5GS) to maximize access and minimize jitter (packet latency variance) between 5G applications.

[0093] Real-time MEC applications typically have well-defined characteristics of their behavior. Cloud (cloud computing) and fog (fog computing) 5G MEC applications can be partitioned into microservices with hard real-time constraints, which are provided to the TSN scheduler. Constraints can be the longest (worst-case) time to complete a call to a service, or a well-defined statistical description of network computation requirements based on arrival or service curves. In some implementations, microservices can be chained together to create a complete MEC application. By decomposing computation into a series of smaller microservices, each service can be better controlled and managed, and each service can provide more determinism. Each microservice can be abstracted as an internal TSN block, consisting of deterministic inputs, outputs, schedulable operations, and coordination with 5G-TSN AF. Microservices can reside on the same or spatially distinct processing systems interconnected via communications scheduled by the TSN.

[0094] This hard real-time processing can include 5G MEC and 5G core functions. Messages entering and leaving the MEC TSN block follow a deterministic schedule that can be computed by the TSN scheduler. Note that this TSN block will be referred to as a "computed TSN block". Given that the messages generated by the MEC and their corresponding sizes and transmission times can vary depending on the computational complexity, processing load, and application state of the MEC processing tasks, the TSN scheduling component can utilize an internal model of the MEC processor to achieve TSN scheduling objectives. This model can be a simulation, emulation, pure analysis, or a hybrid of each (aka digital twin). The TSN scheduling component can then generate a complete, end-to-end TSN schedule for all messages flowing between the 5G UE, MEC, and 5G core, as well as any cloud processing required by the real-time application, where the 5G core and cloud processing are similarly modeled by the TSN scheduler. As the effective processing rate and therefore the output message transmission time change, the TSN schedule can be dynamically recomputed as needed to maintain the determinism required by 5G TSN real-time MEC / cloud applications. Taking into account link speed, variations, processing power, available memory, etc., the TSN scheduler can provide application developers with feedback (and for management and deployment) on the optimal location of each processing component (UE, MEC, cloud) for a real-time 5G application. The TSN system can employ gate-based methods or rate control mechanisms (such as leaky buckets) and can use any number of optimization techniques or network calculus to determine feasible scheduling. Therefore, the complete flow of a TSN application will include not only the end-to-end path of a specific message across the network, but also the complete processing path through all computed TSN blocks (microservices) acting on the information in the message. IEEE 802.1CB redundant paths can be configured through redundant or parallel computed TSN blocks (microservices). In this invention, it should be possible to visualize the complete real-time processing activity encoded in TSN scheduling, where computed TSN blocks are represented as Ethernet bridges, the difference being that these computed TSN blocks either process messages or can create new messages to be sent out under deterministic scheduling.

[0095] MEC applications are configured to use NETCONF, RESTCONF, or a RESTful API to utilize TSN flows (for participation as a TSN speaker or listener). The messages defined in the aforementioned protocols contain IEEE 802.1Qcc information required to configure the TSN flows used by the MEC application. Additionally, the MEC application is designed to migrate from one MEC platform to another, and the RESTful API is defined to query the current platform's TSN configuration to verify that it meets the required deterministic communication requirements, specifically latency and jitter. It also has a RESTful API containing the aforementioned information required to notify the TSN CNC of its new location, as well as the information required to dynamically reschedule TSN traffic to its new location. As mentioned, IEEE 802.1CB can be used to establish redundant TSN flows to the intended location to which the MEC application can migrate. The TSN scheduler (i.e., a centralized network configurator (CNC) or a distributed network configurator (DNC)) can be a MEC application that provides 5G TSN scheduling-as-a-service and configuration-as-a-service for dynamic rescheduling requiring low latency.

[0096] TSN applications can transmit timing performance profile query requests to microservices, aiming to determine statistics regarding processing time. When responding to such a request, the microservice executes the request and returns both the results and the timing performance profile. The timing performance profile can at least include the network start and end times of microservice calls, and optionally, the start and end times of all called sub-functions. This information can also include the average MEC processor load during the timing performance profile. TSN applications can use timing performance profiles to refine TSN scheduling, where microservices are required to handle TSN traffic flows. Timing performance profiles can be obtained during live operation (where the output from the profile is used normally) or as a special test sample prior to live operation. Timing performance profile information can be used to determine which MEC hardware to use for current and future operations and can be used as constraint information for the TSN scheduler.

[0097] This topic disclosure also provides additional new aspects, including: (1) 5G core functions of TSN scheduling; (2) ensuring direct time synchronization between the CPU and the Ethernet hardware timestamp mechanism; (3) implementing TSN scheduling for specific 5G network functions and processes; (4) adding new time-aware programming features, such as time-based event processing; and (5) integrating time-based conditional processing, such as real-time programming implemented via YANG configuration for microservice scheduling to implement service chains (see...). https: / / www.rfc-editor.org / rfc / pdfrfc / rfc7758.txt.pdf(For example of how to operate the YANG tone).

[0098] Additionally, a LinkDelayStatistic (implemented as a YANG model) cumulative distribution function data model can be implemented within the TSN-5G system 106 or 200. In this implementation, the radio links can collect information and feed it to the CNC's scheduler, which then uses this information to schedule variable-speed links. The LinkDelayStatistic YANG model can also provide information on whether the link delay is stationary and ergodic. The scheduler can use this information to predict the future based on a given sample and create accurate scheduling. The CNC can then utilize this knowledge for each TSN block to deploy the best possible TSN traffic shaping or gate scheduling, including using network calculus when determining the outcome.

[0099] This disclosure also envisions Virtualized Network Functions (VNF-TSNs) that can reside as software within a 5G system. These may require dedicated processor hardware to support real-time operation. In some implementations, the TSN can be provided as a software-defined TSN (SDN-TSN) and a 5G TSN-as-a-Service (TaaS). In some implementations, VNF-TSNs can be configured within each 5G sub-component (TSN block): UE, radio head, CU / DU, RAN, MEC, core, etc. As discussed above, microservices can be chained together as part of the TSN schedule. Each microservice can expose its service time characteristics to the TSN schedule. This service is part of the latency, but it can have a larger variance than the communication link. In this implementation, the service is the link between the microservice and the TSN schedule in this integration. Network calculus can be used to incorporate processing latency. TSN inputs provide a clear arrival curve. Processor execution time provides the service curve (we assume that 5G devices have well-defined processing times).

[0100] In another aspect of this disclosure, a time-aware MEC platform is defined as a platform that acts as a PTP client (802.1AS compliant end station) and (if needed) as a PTP bridge (802.1AS compliant bridge). Virtualized MEC platforms running multiple slices (OS, VMs, containers) in parallel require PTP bridges and bridging functionality. Typically, virtual bridges / switches are used in virtualized computing platforms. In this disclosure, a TSN-enabled virtual switch is defined, which includes time-aware capabilities. The time-aware PTP client in the MEC platform runs a PTP state machine and servo mechanisms to synchronize the locally available clock with the master clock in the network. Furthermore, the MEC platform synchronizes the system clock to the PTP clock, where the system includes the operating system, network stack, application stack, or any other software and hardware components that utilize the clock. This enables each MEC application to run at synchronized PTP time. In one example, the MEC host runs the PTP client and provides synchronized time (e.g., as system / host time) to all MEC applications via the virtualization infrastructure. In another instance, each MEC application can run a separate instance of a PTP client, which is connected to the host via a time-aware bridge.

[0101] As a configuration interface, 3GPP 23.501 specifies a centralized configuration model where the CNC configures the 5GS as a time-aware bridge. Similarly, the MEC platform should be able to utilize the CNC to be configured as a time-aware end station or a time-aware bridge. This MEC can support configuring TSN features using the 802.1Qcc model, including time-aware shaping, forwarding, and frame duplication and elimination for redundancy. Furthermore, the MEC host can have a CUC component that uses the 802.1Qcc interface to provide the CNC with data flow requirements from resident applications. The MEC can also provide the CNC with information about its TSN capabilities, enabling the CNC to accurately model the MEC. For example, the MEC can present itself as a TSN end station initiating a set of data flows. The CNC will then appropriately model the MEC within the network and generate the correct configuration. The MEC should support collaboration with the OS and applications to identify data streams. Specific functionality depends on the TSN awareness capabilities of the MEC components. Non-TSN-aware applications on the MEC host require an IP stream identifier in the MEC bridge.

[0102] TSN fronthaul / backhaul (e.g., TSN block 202) can directly connect to the MEC, providing deterministic input to the MEC. However, MEC applications are destined to continuously migrate, or perhaps more intuitively, “float” to the edge of the 5G network to remain closest to the clients they may move to. Therefore, deterministic MEC applications will specify a minimum acceptable packet delay variance (MPDV) threshold. MPDV can be incorporated into the specifications of all MEC applications, which limits the viable application locations to those MEC platforms that exist as TSN blocks within the 5GS. Applications must ensure that the correct TSN stream identification and translation rules are implemented on the new MEC platform (which could be a single processor or a processor subnet), enabling the correct labeling and processing of MEC speaker / listener message frames. MEC applications may wish to carry their own configuration instructions, where possible, for “self-installation” when migrating to a new MEC platform. It should be noted that load balancing mechanisms can be employed to ensure that users do not overload any set of MEC applications and to ensure that MEC applications are distributed optimally. In other scenarios, redundant MEC platforms and applications can be instantiated, and / or MEC applications can migrate due to noise rather than strictly due to mobility. Multiple MECs can also be connected as a TSN redundant system.

[0103] Figure 6A and Figure 6B The illustration depicts an integrated TSN-5G system 600 (similar to system 200) that includes a TSN block 605 (similar to TSN block 202-N) representing or corresponding to a multi-access edge computing (MEC) module. The TSN block 605 is configured as a data source / data destination (i.e., as a TSN terminal), but is located within the 5G system 106, rather than at the edge of the 5G system 106. The TSN block 605 can be configured as a bridging terminal.

[0104] refer to Figure 8 Application data streams can span multiple integrated TSN-5G systems 805 and 810. In this case, TSN blocks can be geographically distributed, interconnecting more than one 5G system using TSN-compatible transports 820 (including but not limited to 5G backbones, private network tunnels, or other wide area networks). In this invention, TSN blocks are configured by their respective CMs 215. In some embodiments, the configuration across two TSN-5G systems 805, 810 can be coordinated via a centralized configuration utility (CNC) 112. In some other embodiments, the CMs 215 between the two TSN-5G systems 805, 810 can communicate directly. In some embodiments, multiple CUCs and CNCs can be used to capture user requirements and generate the TSN solutions and technologies (e.g., TSN scheduling, forwarding instructions, etc.) provided in this disclosure.

[0105] Figure 7 An electronic system 700 is illustrated that can be used to implement one or more embodiments of the present subject matter. The electronic system 700 may be a TSN block 202 and / or a configuration module 215, and / or a part thereof. The electronic system 700 may include various types of computer-readable media and interfaces for various other types of computer-readable media. The electronic system 700 includes a bus 708, one or more processing units 712, a system memory 704 (and / or a buffer), a ROM 710, a permanent storage device 702, an input device interface 714, an output device interface 706, and one or more network interfaces 716, or subsets and variations thereof.

[0106] Bus 708 collectively represents all system buses, peripheral buses, and chipset buses that communicatively connect the numerous internal devices of electronic system 700. In one or more embodiments, bus 708 communicatively connects one or more processing units 712 to ROM 710, system memory 704, and permanent storage device 702. From these various memory units, one or more processing units 712 obtain instructions to be executed and data to be processed in order to perform the processes disclosed in this subject matter. In different embodiments, the one or more processing units 712 may be a single processor or a multi-core processor.

[0107] ROM 710 stores static data and instructions required by one or more processing units 712 and other modules of electronic system 700. On the other hand, permanent storage device 702 can be a read-write memory device. Permanent storage device 702 can be a non-volatile memory cell that stores instructions and data even when electronic system 700 is powered off. In one or more embodiments, a mass storage device (such as a magnetic disk or optical disk and its corresponding disk drive) can be used as permanent storage device 702.

[0108] In one or more embodiments, a removable storage device (such as a floppy disk, flash drive, and its corresponding disk drive) can be used as permanent storage device 702. Like permanent storage device 702, system memory 704 can be a read-write memory device. However, unlike permanent storage device 702, system memory 704 can be volatile read-write memory, such as random access memory. System memory 704 can store any instructions and data that one or more processing units 712 may need during operation. In one or more embodiments, the processes disclosed in this subject matter are stored in system memory 704, permanent storage device 702, and / or ROM 710 (each implemented as a non-transitory computer-readable medium). From these various memory units, one or more processing units 712 retrieve instructions to be executed and data to be processed in order to perform the processes of one or more embodiments.

[0109] Bus 708 is also connected to input device interface 714 and output device interface 706. Input device interface 714 enables a user to transmit information and select commands to electronic system 700. Input devices that can be used with input device interface 714 may include, for example, an alphanumeric keypad and a pointing device (also known as a "cursor control device"). Output device interface 706 enables, for example, the display of images generated by electronic system 700. Output devices that can be used with output device interface 706 may include, for example, printers and display devices such as liquid crystal displays (LCDs), light-emitting diode (LED) displays, organic light-emitting diode (OLED) displays, flexible displays, flat panel displays, solid-state displays, projectors, or any other device for outputting information. One or more embodiments may include a device that acts as both an input device and an output device, such as a touchscreen. In these embodiments, the feedback provided to the user can be any form of sensory feedback, such as visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, voice, or tactile input.

[0110] Finally, as Figure 7 As shown, bus 708 also couples electronic system 700 to one or more networks and / or to one or more network nodes via one or more network interfaces 716. In this way, electronic system 700 can be part of a computer network (such as a LAN, wide area network (“WAN”), or intranet, or a network of networks (such as the Internet)). Any or all components of electronic system 700 can be used in conjunction with the disclosure of this subject matter.

[0111] These functionalities can be implemented in computer software, firmware, or hardware. These technologies can be implemented using one or more computer program products. Programmable processors and computers can be included in or packaged as mobile devices. Processes and logical flows can be executed by one or more programmable processors and one or more programmable logic circuits. General-purpose computing devices, special-purpose computing devices, and storage devices can be interconnected via communication networks. The functionalities disclosed herein can be implemented using quantum computing, pulse-coupled oscillation (PCO) / Ising computing.

[0112] Some implementations include electronic components, such as microprocessors, storage devices, and memories, that store computer program instructions in a machine-readable or computer-readable medium (also known as a computer-readable storage medium, machine-readable medium, or machine-readable storage medium). Examples of such computer-readable media include RAM, ROM, read-only optical discs (CD-ROM), recordable optical discs (CD-R), rewritable optical discs (CD-RW), read-only digital universal discs (e.g., DVD-ROM, dual-layer DVD-ROM), various rewritable / rewritable DVDs (e.g., DVD-RAM, DVD-RW, DVD+RW, etc.), flash memory (e.g., SD cards, mini-SD cards, micro-SD cards, etc.), magnetic and / or solid-state drives, read-only and recordable Blu-ray® discs, ultra-density optical discs, any other optical or magnetic media, and floppy disks. The computer-readable medium can store a computer program that can be executed by at least one processing unit and includes a set of instructions for performing various operations. Examples of computer programs or computer code include, for example, machine code generated by a compiler and files containing high-level code executed by a computer, electronic components, or microprocessor using an interpreter.

[0113] While the above discussion primarily concerns microprocessors or multi-core processors that execute software, some implementations are performed by one or more integrated circuits, such as application-specific integrated circuits (ASICs) or field-programmable gate arrays (FPGAs). In some implementations, such integrated circuits execute instructions stored on the circuit itself.

[0114] As used in this specification and any claim of this application, the terms "computer," "server," "processor," and "memory" all refer to electronic devices or other technical devices. These terms do not include people or groups of people. For the purposes of this specification, the term "display" means display on an electronic device. As used in this specification and any claim of this application, the term "computer-readable medium" is entirely limited to tangible physical objects that store information in a computer-readable form. These terms do not include any wireless signals, wired download signals, or any other transient signals.

[0115] To provide user interaction, embodiments of the subject matter described in this specification can be implemented on a computer having a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide user interaction; for example, feedback provided to the user can be any form of sensory feedback, such as visual, auditory, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input. Additionally, the computer can interact with the user by sending and receiving documents from the device used by the user; for example, by sending a web page to a web browser in response to a request received from a web browser on the user's client device.

[0116] The aspects of the subject matter described in this specification can be implemented in a computing system that includes back-end components (e.g., as a data server), or middleware components (e.g., an application server), or front-end components (e.g., a client computer with a graphical user interface or a web browser through which a user can interact with embodiments of the subject matter described in this specification), or any combination of one or more such back-end, middleware, or front-end components. The components of the system can be interconnected via digital data communication (e.g., a communication network) of any form or medium. Examples of communication networks include local area networks (“LANs”) and wide area networks (“WANs”), interconnected networks (e.g., the Internet) and peer-to-peer networks (e.g., self-organizing peer-to-peer networks).

[0117] According to various aspects of this disclosure, a wireless network (e.g., a 5G network 106) is provided, configured to support time-sensitive deterministic communication based on a Time-Sensitive Networking (TSN) mechanism. The wireless network may include multiple discrete communication network (CN) components, such as components 110, 118, 120, 122, 124, 126, 140, 190, etc., discussed above. Each CN component may be configured to provide discrete functionality for data communication across the wireless network (e.g., 106) from a source device (e.g., 102) to a destination device (e.g., 104). A processor (e.g., 402) may be arranged to configure at least one of the multiple CN components (e.g., 140) with a TSN parameter set of the TSN mechanism, such that at least one of the multiple CN components supports time-sensitive deterministic communication as a TSN block (e.g., 202-3) in the TSN network according to the TSN parameter set. At least one CN component configured with the TSN parameter set is a 5G transport network channel 140 connecting RAN 120 and UPF 122. In some implementations, the processor is arranged to configure each of a plurality of CN components with a corresponding TSN parameter set of the TSN mechanism, such that each of the plurality of CN components supports time-sensitive deterministic communication as a corresponding TSN block in the TSN network according to the corresponding TSN parameter set.

[0118] The wireless network may further include a configuration module (e.g., 110 or 210) configured to determine a set of TSN parameters for at least one of a plurality of CN components. The TSN parameter set may include one or more of the following: TSN stream identifier, transmission scheduling, deadline or delay budget, filtering configuration, and redundancy scheme. In some embodiments, the processor is configured to assign one or more CN resources for data transmission in at least one of the plurality of CN components according to the transmission scheduling or deadline or delay budget. In some embodiments, the configuration module is configured within a control plane component (e.g., SMF 126) of the plurality of CN components.

[0119] In some implementations, the configuration module is configured to determine a set of TSN parameters based on one or more CN attributes assigned to at least one of the plurality of CN components, wherein the one or more CN attributes are associated with discrete functions provided by at least one of the plurality of CN components.

[0120] The processor can be configured to monitor data transfers performed by at least one of a plurality of CN components and report any changes to the TSN parameter set or one or more CN attributes to the configuration module, and the configuration module is configured to update the TSN parameter set based on the reported changes.

[0121] According to various aspects of this disclosure, a wireless network (e.g., a 5G network 106) is provided, configured to support time-sensitive deterministic communication based on a Time-Sensitive Networking (TSN) mechanism. The wireless network may include multiple discrete communication network (CN) components, such as components 110, 118, 120, 122, 124, 126, 140, 190, etc., discussed above. Each CN component may be configured to provide discrete functionality for data communication across the wireless network (e.g., 106) from a source device (e.g., 102, 204) to a destination device (e.g., 104, 206). The wireless network may include multiple configuration modules (e.g., 210, 215) interconnected in some arrangement (e.g., via 230) and configured to provide each of the multiple CN components with a unique set of TSN parameters, such that each of the multiple CN components supports time-sensitive deterministic communication as a corresponding TSN block in the TSN network according to the corresponding unique TSN parameter set. The multiple configuration modules may be interconnected in a hierarchical arrangement, a mesh arrangement, a star arrangement, a tree arrangement, or a combination thereof.

[0122] As mentioned above Figure 2 As illustrated and discussed, the corresponding TSN block (e.g., 202-3) can be configured to receive its unique TSN parameter set from the configuration module (e.g., 215-b and / or 215-c) assigned to the corresponding TSN block from among multiple configuration modules. This unique TSN parameter set may include one or more of the following: TSN stream identifier, transmission schedule, deadline or delay budget, filtering configuration, and redundancy scheme.

[0123] In some implementations, at least one of the multiple configuration modules is configured as CUC 114 and / or CNC 112, and implemented within the SMF 126 of the control plane and / or within TSN block 202-3 (which represents the TSN configuration of the 5G transport channel 140). In implementations according to various aspects of this disclosure, wherein the 5G transport channel 140 is configured with TSN parameters, RAN 120 and UPF 122 may each be configured to support TSN speaker and / or listener functions as defined in IEEE P802.Qdj [xx]. SMF 126 / CUC 114 may be responsible for converting 5GS parameters to and from parameters in IEEE P802.1Qdj [xx]. SMF 126 / CUC 114 may be configured to communicate with the TSN speaker and / or listener in RAN 120 and UPF 122 to exchange datasets defined in IEEE P802.1Qdj [xx]. In some implementations, during the establishment or modification of a QoS flow, the SMF 126 / CUC 114 creates a speaker group and a listener group for each QoS flow as defined in Appendix Ix and sends them to the TSN CNC 112. The TSN CNC 112 uses the speaker group and listener group as input to select a path and calculate the schedule in the TSN. The TSN CNC 112 provides a status group to the SMF 126 / CUC 114. The SMF 126 / CUC 114 provides the status group to the TSN speakers and / or listeners in RAN 120 and UPF 122, respectively. Upon receiving the status group, the TSN speakers and / or listeners send data to the N3 interface according to the configuration in the status group.

[0124] In some implementations, the processor (e.g., 402) is configured to assign one or more CN resources for data transmission of at least one of a plurality of CN components according to a transmission schedule or deadline or delay budget.

[0125] According to various aspects of this disclosure, a system is provided (e.g., as per [reference to...]). Figure 8The illustrated and described system 100 includes a first wireless network (e.g., 805) and a second wireless network (e.g., 810). The first wireless network may include a first plurality of discrete communication network (CN) components, each CN component being configured to provide discrete functionality for data communication via the first wireless network, wherein at least one of the first plurality of CN components is configured to provide time-sensitive deterministic communication according to a first TSN parameter set. The second wireless network may include a second plurality of CN components, each CN component being configured to provide discrete functionality for data communication via the second wireless network, wherein at least one of the second plurality of CN components is configured to provide time-sensitive deterministic communication according to a second TSN parameter set.

[0126] In some implementations, a TSN transport channel (e.g., 820) is provided, which is configured to facilitate time-sensitive deterministic communication of data exchanged between a first wireless network and a second wireless network according to a third TSN parameter set. The TSN transport channel may be communicatively connected to a first TSN converter of the first wireless network (e.g., NW-TT in 122 of 805) and a second TSN converter of the second wireless network (e.g., NW-TT in 122 of 810).

[0127] According to various aspects of this disclosure, a method is provided that includes configuring at least one (or each) of a plurality of CN components of a wireless network (e.g., a 5G network) using a TSN parameter set of a TSN mechanism, such that at least one (or each) of the plurality of CN components supports or provides time-sensitive deterministic communication as a TSN block in the TSN network, based on the TSN parameter set. At least one of the plurality of CN components may be a 5G transport network channel connecting the RAN and UPF.

[0128] The method may further include, for example, determining a TSN parameter set for at least one of a plurality of CN components at one or more configuration modules. The TSN parameter set may include one or more of the following: TSN stream identifier, transmission schedule, deadline or delay budget, filtering configuration, and redundancy scheme. The method may further include assigning one or more CN resources for data transmission by at least one of the plurality of CN components according to the transmission schedule or deadline or delay budget. The method may also include monitoring data transmission performed by at least one of the plurality of CN components and reporting any changes to the TSN parameter set or one or more CN attributes to the configuration module, wherein the TSN parameter set is updated based on the reported changes.

[0129] Wireless communication networks based on standards defined by 3GPP (e.g., 5G, 6G, etc.) can transmit information quickly with low latency. Industrial systems can be controlled by control signals, which are provided by the control system and transmitted via such wireless communication networks. Control systems typically require... r (Reference Input - Target Output) e (Output error) u (Control parameter status) and y (Actual Output). A control system may include users (entities that control the reference inputs), controllers (systems that decide what actions to take), the target industrial system (e.g., a factory, manufacturing plant, etc.), and sensors, which may be located remotely to each other. Additionally, models, typically in the form of differential equations, correlate the rate of change of the target industrial system with changes in control parameters. While wireless communication networks provide the communication needed to interconnect control system components (often assumed to be ultra-reliable low-latency communication (URLLC)), they do not possess the inherent control system knowledge or capabilities of the user application. Most control systems assume negligible communication latency. However, the performance of a control system depends heavily on providing accurate, real-time control information about the control system's state. For example, if the components of a control system are distributed across a wireless communication network, messages between components may arrive with variable delays or latency. As discussed above, wireless communication networks can be integrated with Time-Sensitive Network (TSN) systems to provide bounded, predetermined delays (scheduled flow of traffic) to the control system, thus preventing the control system from becoming unstable. The stability of a control system is defined as the ability of any system to provide bounded outputs when bounded inputs are applied to it. If messages experience variable delays or latency during communication, the controller's response will occur too early or too late, leading to system instability (i.e., loss of control with no hope of regaining it). For a wireless communication network (TSN) to succeed, it is crucial that the control system remain stable in wireless environments susceptible to noise and the greater challenges of variable delays or latency. However, the controller's control law (the mathematical formula used to determine the control signal) does not account for such delays or latency, potentially impairing its performance. Therefore, solutions that compensate for variable delays or latency are needed.

[0130] To address the aforementioned issues, various solutions have been provided, such as enabling communication networks to provide low deterministic latency, increasing the complexity of control laws to compensate for variable communication latency, or a combination of both. One technique for solving this problem is to enable wireless communication networks—TSNs—to become more proactive participants in the control system by co-developing control laws and TSN scheduling rather than developing them separately. Therefore, this subject matter provides a control system that utilizes an artificial intelligence / machine learning (AI / ML) system associated with a wireless communication network. The AI / ML system can minimize operational complexity by leveraging efficient algorithms. The AI / ML system can dynamically update the controller's control law to increase its complexity. The control law is used by the controller to determine the output to be sent to a target industrial system (e.g., a factory). u The mathematical formula. In feedback control scenarios, the output... u It can generate robustness to uncertainty and can be used to design system dynamics. A control law is typically part of a controller that takes the output of the process to be controlled as its input. This input is then used by the algorithm employed by the controller to determine u, which is the input to the target industrial system. For example, in a cruise control system, the control law compares a reference speed with the current speed and uses this error in a PID control algorithm. v_r-v ) to determine the signal u Then the signal u Send to the end device (e.g., an actuator) to reduce the error between the desired speed and the actual speed.

[0131] In addition, the AI / ML system can simultaneously learn the corresponding Time-Sensitive Network (TSN) flow scheduling that supports the system through the communication network, and create the required TSN flow under network constraints based on the updated control law. Therefore, in some embodiments, the control system may include a device for transmitting control signals via a communication network configured to support time-sensitive deterministic communication based on a time-sensitive networking (TSN) mechanism. The device includes a processor and a memory communicatively coupled to the processor, and one or more instructions stored in the memory. When executed by the processor, the one or more instructions cause the processor to: send quality of service (QoS) requirement information to an end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; poll components of the communication network to obtain operating parameter information related to operating parameters of the communication network; send the control signal to the end device based on the QoS requirement information, wherein the control signal, based on a control law, receives the operating parameter information from components of the communication network in response to polling of the operating parameters and receives feedback information from the end device; and, based on artificial intelligence (AI) functionality, update the control law of the device for transmitting the control signal based on the operating parameter information and the feedback information to compensate for delays in the transmission of the control signal.

[0132] Figure 9 An example of a control system according to one or more embodiments is illustrated.

[0133] Figure 9 The control system 900 depicts a block diagram of an example implementation of an integrated TSN-wireless communication network system similar to systems 100 and 200 described above, and also includes physical components similar to those discussed in systems 100 and 200. Additionally, the control system 900 includes an artificial intelligence / machine learning (AI / ML) control application 910 configured to interact with the wireless communication network 106 and end devices 102 (e.g., actuators, sensors, etc.). Figure 9 As shown, the AI / ML control application 910 includes a control law optimizer 920, which is configured to provide AI / ML functionality for optimizing the control laws of a controller (e.g., controller 104). In some embodiments, the control law optimizer 920 may be referred to as an AI / ML system.

[0134] As discussed above, the control system 900 typically requires r (Reference Input - Target Output) e (Output error) u (Control parameter status) and y(Actual output) The control system 900 includes a controller that determines what action to take. The control law is used by the controller to determine the output sent to the target system (or end device). u The mathematical formula. A control law uses the output of the process to be controlled as input to determine... u This is to generate robustness to uncertainties, which can be used to design system dynamics. The determined... u It can be the input to the target system (factory). For example, in a cruise control system, the control law compares the reference speed with the current speed, and uses this error in the PID control algorithm. v_r-v ) to determine the signal u Then the signal u The control law is sent to the target system (e.g., an actuator) to reduce the error between the desired and actual speeds. For example, the control law can be enhanced to account for sensing delays to mitigate any dangerous oscillating behavior (e.g., obstacles in front of a car), provided the control law is intelligent enough to learn and incorporate these delays. The type of control law can be varied, including, for example: 1. Dynamic programming, Bellman equations, optimal value functions, value and policy iteration, shortest path, Markov decision processes; Hamilton-Jakibi-Bellman equations and approximation methods; Pontryagin maximum principle, ODE and gradient descent methods, and their relationship to classical mechanics; linear quadratic Gaussian control, Riccati equations, iterative linear approximations of nonlinear problems; optimal recursive estimation, Kalman filters, Zakai equations; duality of optimal control and optimal estimation (including new results). Furthermore, the configuration and parameters of the control law can also be changed.

[0135] In some implementations, closing control loops on wireless channels can introduce noise and lead to message loss in control systems. Proper allocation of limited resources (e.g., transmission power, MIMO configuration, beam pointing, resource blocks, bandwidth, and scheduling) is crucial for maintaining reliable operation. Many AI / ML techniques, including deep reinforcement learning, can be used to find resource allocation strategies based on neural networks. Model-free policy gradient methods can be used to directly learn continuous communication allocation strategies without knowing the dynamics or communication model of the controlled object (plant). In some implementations, control system 900 may have numerous distributed controllers that need to be allocated 5GS resources (including TSN scheduling). In some implementations, AI / ML system 920 can learn how to allocate such distributed controllers.

[0136] When interacting with a wireless communication network, the control system 900 can open its control law requirements to the wireless communication network 106, and allow the communication network 106 to perform trade-offs and allocate resources as needed. For example, the control system sensors can receive low but variable time delays, while the reference...r and actual output y Signals can receive higher but more deterministic time delays, or any combination thereof. In some implementations, AI / ML can be used to control all types of control systems. Many types of control systems exist. A common type of controller is the proportional-integral-derivative controller (PID controller or three-term controller) and employs a feedback control loop mechanism, widely used in industrial control systems and various other applications requiring continuous modulation control. The PID controller continuously calculates the error value as the difference between the desired setpoint (SP) and the measured process variable (PV), and applies corrections based on the proportional, integral, and derivative terms (denoted as P, I, and D, respectively), hence its name. In practice, the PID controller automatically applies accurate and responsive corrections to the control function. A common example is cruise control in a car, where applying constant engine power will cause the speed to decrease when going uphill. The controller's PID algorithm restores the measured speed to the desired speed with minimal delay and overshoot by increasing the engine power output in a controlled manner. In some implementations, the AI / ML system 920 can learn to adjust the PID configuration parameters. For example, the AI / ML system 920 learns to optimize the K_p, Ki, and K_d coefficients of the PID controller.

[0137] More generally, the stability of a control system 900 is defined as the system's ability to provide a bounded output when a bounded input is applied. Stability allows the system to reach a steady state and remain in that state for a given input after all system parameters have changed. The relative stability of a feedback loop is described in terms of gain and phase margin. Gain and phase margin measure the tolerance of the control loop to changes in the open-loop system response. One of the primary goals of the AI / ML system 920 is to provide sufficient wireless communication network capability to maintain the required gain and phase margin. Other aspects of control system performance include damping ratio, settling time, bandwidth, and system sensitivity. Damping ratio is defined as the number of oscillations in the system that may decay or be suppressed after an interruption, and it is a dimensionless measurement. When the control system 900 is interrupted or disturbed from its initial position, it can exhibit oscillating patterns. Settling time is the time required for the response to become stable. Steady-state error is the difference between the actual output and the desired output over an infinite time range. Response time is the time it takes for the control system 900 to reach its steady-state value while maintaining its initial rate of change (after a delay time). The delay time is the time required for the response to reach half of the final (target) value (the time required to reach half of the final value for the first time when oscillating around the target value). In most systems, this becomes the dominant time constant. The bandwidth of a closed-loop control system is defined as the frequency range in which the magnitude of the closed-loop gain does not drop below -3 dB. Sensitivity characterizes the effect of changes in process dynamics. Controller parameters are typically matched to process characteristics, and because the process can change, it is important that the controller parameters be chosen so that the closed-loop system is insensitive to changes in process dynamics.

[0138] The AI / ML system 920 associated with the wireless communication network 106 should absorb information about the control system 900 (e.g., control laws) and information about the instantiation of the wireless communication network, and return an optimal configuration that satisfies both the wireless communication network 106 (consuming the minimum amount of network resources) and the control system 900 (meeting the previously discussed control system performance characteristics). The AI / ML system 920 should sample both network resources and control system parameters during its learning process.

[0139] The framework of the AI / ML system 920 is defined by 3GPP specifications 22, 23, and 28, such as 3GPP TS 22.261 (2022-12, Release 19), 3GPP TS 23.288 (2022-12, Release 18), 3GPP TR 22.874 (2021-12, Release 18), 3GPP TR 22.876 (2023-03, Release 19), 3GPP TR 23.700-18 (2022-12, Release 18), 3GPP TR 23.700-36 (2022-12, Release 18), 3GPP TR 23.700-80 (2022-12, Release 18), and 3GPP TS 22.261 (2022-12, Release 19), 3GPP TR 23.700-18 ...18 (2022-12, Release 18), 3GPP TR 23.700-18 (2022-12, Release 18), 3GPP TR 23.700-18 (2022-12, Release 18), 3GPP TR 23.700-18 (2022-12, Release 18), 3GPP TR 23.700-18 (2022-12, 28.104 (2022-12, Release 17), 3GPP TS 28.105 (2022-12, Release 17), 3GPP TR 28.838 (2023-01, Release 18), and 3GPP TR 28.908 (2023-03, Release 18), which are incorporated herein by reference in their entirety.

[0140] The wireless communication network 106 is similar to the 5G system 106 discussed above. The wireless communication network 106 includes multiple communication network (CN) components, such as components 110, 118, 120, 122, 124, 126, 140, 190, etc. The wireless communication network 106 may be referred to as a communication network. Figure 9As shown, the AI / ML control application 910 can send control input 912 to the wireless communication network 106. As described above, the control input 912 includes control signals generated based on the controller's control law for the operation of control end devices (e.g., sensors, actuators, etc.). These control signals are transmitted to the end devices. Each CN component is configured to provide discrete functions for transmitting control signals to end devices across the communication network. In some embodiments, at least one of the multiple CN components is configured to provide network function 126 according to 3GPP standards. Network function 126 includes control plane functions and user plane functions. In some embodiments, the control plane functions may be referred to as 5G core control functions. Network function 126 includes, but is not limited to, Artificial Intelligence Application Function (AI-AF), Access and Mobility Management Function (AMF), Session Management Function (SMF), Network Data Analysis Function (NWDAF), Network Slice Selection Assist Information (NSSAI), and User Plane Function (UPF). AI-AF specifies how the wireless communication network 106 interacts with the control system 900. AMF is a component of the 3GPP core network architecture that manages user equipment registration, authentication, identification, and mobility. SMF is a fundamental component of 5G Service-Based Architecture (SBA) and is primarily responsible for decoupling data plane interaction, creating, updating, and removing Protocol Data Unit (PDU) sessions, and managing session context with UPF. NWDAF is a 3GPP standard approach for collecting data from user equipment, network functions, and Operations, Administration, and Maintenance (OAM) systems, which can be used for analysis, from 5G core, cloud, and edge networks. NSSAI is an interface used to determine resource and availability or create network slices. UPF connects data transmitted over the Radio Local Area Network (RAN) to the desired endpoints. QoS is Quality of Service, which quantifies the latency and determinism of message delivery time relative to the requirements of a real-time control system.

[0141] In some implementations, the AI / ML control application 910 can interact with the AMF in the communication network 106. For example, as Figure 9 As shown, the AI / ML control application 910 can send a request for the status of the communication network to the communication network 106 via the AMF. The request for the communication network status includes a request to poll components of the communication network to obtain operational parameter information related to the operating parameters of the communication network. Therefore, the AI / ML control application 910 can poll components of the communication network to obtain operational parameter information related to the operational parameters of the communication network regarding the provisioning of data paths used for transmitting control signals. In some embodiments, the AI / ML control application 910 polls components of the communication network to obtain operational parameter information related to the operational parameters of the communication network regarding bandwidth, latency, and predicted load along the path to the end device 102.

[0142] In some implementations, in response to a request for communication network status, at least one of the plurality of CN components can be configured to provide at least one network function of network function 126 to collect data and send a response including communication network status 930 to AI / ML control application 910. In some implementations, the response including communication network status 930 includes data from NWDAF 126 and a YANG model associated with network functions such as AMF, SMF, and UPF. The response 930 from communication network 106 can be provided to AI / ML system 920 as input data (or features). In some implementations, data from NWDAF 126 is provided as primary input data to understand the status of wireless communication network 106 and predict load, while the YANG model associated with network functions such as AMF, SMF, and UPF is provided as secondary input data. It is assumed that sufficient network time synchronization is implemented to support TSN, and that all sampled data used for AI / ML features is accurately timestamped. In machine learning and pattern recognition, each AI / ML feature is a separate measurable characteristic or feature of a phenomenon.

[0143] like Figure 9 As shown, in some embodiments, the wireless communication network 106 transmits control input 912 to end device 102 and receives feedback information 935 from end device 102. In some embodiments, the feedback information 935 may include sensor values ​​(e.g., speed in a cruise control system). In some embodiments, the feedback information 935 may include flight control information in an aircraft control system, laser control for additive manufacturing, video quality in a video codec, channel state estimation or beam pointing in a wireless system, or any other desired real-time control parameters. The wireless communication network 106 transmits the feedback information 935 to the AI / ML control application 910. In some embodiments, the feedback information 935 is also provided as input data to the AI / ML system 920.

[0144] AI / ML control application 910 can receive responses 930 from wireless communication network 106, including communication network status information and feedback information 935, and provide them as input data to AI / ML system 920. AI / ML system 920 learns the controller's control law and optimizes it to compensate for latency in the data path used to transmit control signals. The output of AI / ML system 920 includes one or more policies for configuring network slicing and MEC applications. One or more policies can define TSN scheduling, network slicing information, and network access priorities. For example, one or more policies can define additional constraints such as traffic load, energy usage, RF interference, compute load, or priority for higher-level urgent traffic flows. In some implementations, AI / ML control application 910 is designed to issue control signals that should be received at sensors every predetermined time interval (e.g., 20 ms). Based on feedback information 935 from wireless communication network 106, AI / ML system 920 learns that wireless communication network 106 can support latency from 15 ms to 40 ms with jitter of 10 ms. Using this learned information, one of the AI / ML models in the AI / ML system 920 can output a strategy that optimizes the control law, thereby transmitting control signals at intervals determined by the AI / ML model, ensuring that the end device always receives the control signal at 20 ms. The joint optimization of the control law and network strategy will attempt to compensate for network jitter to ensure that the sensor receives the control signal at 20 ms.

[0145] In some implementations, the control system 900 may include a TSN scheduler (e.g., configuration module 215) that determines TSN scheduling. At least one of the plurality of CN components is configured to provide TSN application functionality (TSN-AF) 124 to manage the configuration of TSN based on TSN scheduling. As discussed above, TSN is used to provide a bounded, predetermined delay (scheduled flow) for the control system 900 to prevent the control system 900 from becoming unstable. TSN scheduling determines how long a message may take to travel between the controller, sensor 102, and actuator 102. Therefore, the AI / ML system 920 may reposition or regroup the microservices of the AI / ML control application 910 based on the updated control law and the received responses 930 and feedback information 935. Each microservice provides each function of the AI / ML control application 910 and can be implemented using the resources of the wireless communication network 106. The AI / ML system 920 may split at least one of the microservices or combine at least one of the microservices to reposition or regroup the microservices on the resources of the wireless communication network 106. Therefore, the TSN scheduling and configuration must be changed accordingly. In some implementations, the TSN scheduler updates the TSN schedule based on the updated control law, such that each of the CN components is provided with the TSN-AF function to change the configuration of the TSN block based on the updated TSN schedule. Therefore, the control system 900 can autonomously and simultaneously update both the control law and the TSN schedule.

[0146] Figure 10 The diagram illustrates a flowchart of an example operation of a control system according to one or more embodiments.

[0147] Flowchart 1000 illustrates the operation of the control system 900 discussed above. The control system 900 includes a control application host that provides a virtualized environment for an AI / ML control application 910. In operation 1010, the control application host can transmit data path requirement information as a network slice SLA to the wireless communication network 106, and instantiate a provisioning using network functions 126 such as AMF and SMF, and TSN-AF 124. In some embodiments, the data path requirement information includes computational delay information and communication delay information required for transmitting control signals. In some embodiments, the wireless communication network 106 provides network functions 126 to configure data path provisioning for transmitting control signals generated by the controller based on control laws, based on the data path requirements. Figure 9As discussed herein, the AI / ML control application 910 may send a request for a communication network status 915 to the communication network 106 via the AMF. The request for the communication network status 915 includes a request for polling components of the communication network to obtain operational parameter information related to the operating parameters of the communication network. In some embodiments, the operational parameter information related to the operating parameters of the communication network includes network status information regarding bandwidth, latency, and predicted load along the data path to the end device. Therefore, in operation 1020, the AI / ML control application 910 polls for the wireless communication network status provided for the data path. In operation 1030, the AI / ML control application 910 polls for the wireless communication network status regarding bandwidth, latency, and predicted load along the data path to the end device 102 (e.g., a sensor or actuator). In operation 1040, the AI / ML control application 910 sends a control signal (or multiple control signals) to the end device 102 via the wireless communication network 106. In operation 1050, the AI / ML control application 910 collects feedback information 935 from the end device 102 and collects responses to polling 930 from the wireless communication network 106. For example... Figure 9 As discussed herein, the AI / ML control application 910 provides feedback information 935 and response 930 as input data to the AI / ML system 920. Therefore, as shown in operation 1060, the AI / ML control application 910 optimizes the controller's control law to compensate for latency in the data path. In operation 1070, the AI / ML control application 910 may request new data path requirements from the control host. New data path requirements include, but are not limited to, network slicing and network access priority (network slicing SLA). Operations 1010 to 1070 can be cycled periodically to adapt to changes.

[0148] Figure 11 An example of the operation of an AI / ML system according to one or more implementations is illustrated.

[0149] In some implementations... Figure 10 The AI / ML system 1100 in the middle can be implemented as Figure 9 The AI / ML system 920 is used to update the control laws of the controllers discussed above. For example... Figure 11 As shown, the AI / ML system 1100 may include a data collection module 1110, a model training module 1120, a model inference module 1130, and an actuator module 1140. The operation of the AI / ML system 1100 can be implemented in a distributed manner across the wireless communication network 106.

[0150] In some examples, the data collection module 1110 provides input data to the model training module 1120 and the model inference module 1130. For example... Figure 9 and Figure 10 As discussed herein, the input data may include data from NWDAF 126 and YANG models related to network functions (such as AMF, SMF, and UPF). In some embodiments, data from NWDAF 126 is provided as primary input data to understand the state of wireless communication network 106 and predict load, while YANG models related to network functions (such as AMF, SMF, and UPF) are provided as secondary input data. In some embodiments, AI / ML algorithm-specific data preparation (e.g., data preprocessing and cleaning, formatting, and transformation) is not performed in data collection module 1110. In some embodiments, input data may include measurements from various CN components (such as UE) or different network entities, feedback 1145 from actuator module 1140, and / or data outputs from AI / ML models.

[0151] In some implementations, based on input data, the data collection module 1110 may provide training data 1112 to the model training module 1120 during the training phase of the AI / ML system 1100; and inference data 1114 to the model inference module 1130 during the inference phase of the AI / ML system 1100. In some implementations, the training data 1112 is the data required at the model training module 1120 for training, validating, and testing the AI / ML model as input; and the inference data 1114 is the data required at the model inference module 1130 for performing inference functions using the trained AI / ML model as input.

[0152] In some implementations, the AI / ML system 1100 may be in a training phase, an inference phase, or both in real time. One purpose of the training phase is to train (and validate and test) a function or model, for example, by learning some optimal parameters or weights in the function or model based on some optimization metric, given known inputs and known outputs from the training data. One purpose of the inference phase is to apply the trained function or model to some known inputs from the inference data to infer the output of the trained function or model based on the learned optimal parameters or weights.

[0153] During the training phase of the AI / ML system 1100, the model training module 1120 performs training, validation, and testing of the AI / ML model, which may generate model performance metrics as part of the model testing procedure. In some embodiments, if needed or required, the model training module 1120 may also be responsible for data preparation (e.g., data preprocessing, cleaning, formatting, and transformation) based on the training data 1112 transmitted by the data collection module 1110. At the end of the training phase, the model training module 1120 may generate a model deployment / update 1122 to initially deploy the trained, validated, and tested AI / ML model to the model inference module 1130, or to transmit an updated AI / ML model to the model inference module 1130.

[0154] During the inference phase of the AI / ML system 1100, the model inference module 1130 applies a trained or retrained (updated) AI / ML model to inference data 1114 to generate output 1135, which may include, for example, a prediction or decision. The model inference module 1130 may provide model performance feedback 1132 based on the prediction or decision to the model training module 1120, where applicable. In some embodiments, if desired or required, the model inference module 1130 may also be responsible for data preparation (e.g., data preprocessing and cleaning, formatting, and transformation) based on the inference data 1114 transmitted by the data collection module 1110. The model performance feedback 1132 may be used by the model training module 1120 to monitor the performance of the AI / ML model (when available). Output 1135 may include the inference output of the AI / ML model after it has been applied by the model inference module 1130. In some implementations, the operation of the model inference module 1130 can be implemented in a distributed manner across the wireless communication network 106, which includes both internal processors (UE, RAN, MEC, core) and external processors (client processors connected via the wireless communication network 106) for wireless communication network instances according to 3GPP standard specifications (e.g., 3GPP TR 37.817 V17.0.0 (2022-04) which are incorporated herein by reference in their entirety).

[0155] In some implementations, the actuator module 1140 receives output 1135 from the model inference module 1130 and triggers or executes a corresponding action. For example, the corresponding action may include updating the control law of the controller as discussed above.

[0156] The actuator module 1140 can trigger actions against other entities or itself. Feedback 1045 may include information that may be needed to obtain additional training data 1112, additional inference data 1114, or to monitor the performance of the AI / ML model and its impact on the network by updating key performance indicators (KPIs) and performance counters.

[0157] In some implementations, trained AI / ML models can simulate cognitive functions associated with humans and other human minds. In particular, by training on sufficient training data, trained AI / ML models can adapt to new situations and detect and extrapolate patterns.

[0158] In some implementations, the parameters of a trained AI / ML model can be tuned through training. Specifically, combinations of supervised training, semi-supervised training, unsupervised training, reinforcement learning, and / or active learning can be used. Furthermore, representation learning (an alternative term is "feature learning") can be used. In particular, the parameters of a trained AI / ML model can be iteratively tuned through several training steps.

[0159] In some implementations, the trained AI / ML model may include neural networks, support vector machines, decision trees, and / or Bayesian networks. The trained AI / ML model may be based on k-means clustering, Q-learning, genetic algorithms, and / or association rules. Specifically, the neural network may be a deep neural network, a convolutional neural network, or a convolutional deep neural network. Furthermore, the neural network may be an adversarial network, a deep adversarial network, and / or a generative adversarial network.

[0160] Figure 12 An example of a computational graph of a neural network (NN) according to one or more embodiments is illustrated.

[0161] In some implementations, artificial intelligence / machine learning (AI / ML) systems (e.g., AI / ML system 902, AI / ML system 1100) can be implemented as neural networks (NNs), but are not limited to this. A neural network is a family of algorithms that attempts to identify underlying relationships in a dataset by mimicking the process of the human brain. In some implementations, AI / ML systems can use the Neural Network Exchange Format (NNEF) to exchange information about trained neural networks. The goal of NNEF is to provide a standard platform for connecting accelerated neural network execution engines and available deep learning tools. Ideally, neural networks trained in deep learning frameworks will be exported to NNEF, and neural network accelerator libraries can use these neural networks without concern for compatibility with all deep learning frameworks. For application programmers, NNEF is a standardized way to store and transfer neural networks. Given a neural network in NNEF and a driver or library capable of importing that neural network, application programmers do not need to worry about where the network comes from or what type of underlying hardware will execute it, as long as they have the capability to do so. Application programmers can query the driver of the underlying hardware to see if it can execute a given network.

[0162] like Figure 12As shown, computation graph 1200 describes a neural network (NN). A computation graph is a directed graph with two types of components: data nodes (nodes corresponding to variables) and operation nodes (nodes corresponding to operations). Data nodes can feed their values, representing multidimensional arrays (such as vectors, matrices, or higher-dimensional arrays), to operation nodes, and operation nodes can feed their outputs to other operation nodes. Thus, each node in computation graph 1200 defines a function of a variable. Figure 12 In this diagram, data nodes are represented as squares 1210, 1212, 1214, 1214-1, 1214-5, and 1216, and operation modes are represented as ellipses 1220, 1220-1, and 1220-2. In some implementations, squares 1210, 1212, 1214, 1214-1, 1214-5, and 1216 represent values ​​fed into and / or output from operation nodes (E: external 1210, C: constant 1212, T: regular tensors 1214, 1214-1, and 1214-5, V: variable 1216). These values ​​are called tensors or tensor data. Computation graph 1200 begins with a data node (e.g., square 1210) representing externally provided data or constant or variable tensors within computation graph 1200. For consistency, such data nodes are the result of special tensor introduction operations (e.g., ellipse 1220). Therefore, the entire computation graph 1200 is described by a set of operations (e.g., ellipse 1220) interconnected by data nodes (e.g., square 1210).

[0163] Figure 12 The computational graph 1200 depicts a linear graph starting from the input data node (e.g., square 1210), which describes a simple multilayer feedforward network as an example of a neural network, but is not limited thereto. Each layer corresponds to an operation node (e.g., ellipse 1220-1) that produces new intermediate data nodes (e.g., square 1214-1). An operation node (e.g., ellipse 1220-2) can take the intermediate data represented by the new intermediate data node (e.g., square 1214-1) as input and eventually produce some output data (e.g., square 1214-5). A directed edge (e.g., arrow 1230) from a data node (e.g., square 1210) to an operation node (e.g., ellipse 1220) indicates that the operation represented by that operation node can take the corresponding data represented by that data node as its input. A directed edge (e.g., arrow 1240) from an operation node (e.g., ellipse 1220) to a data node (e.g., square 1214) signifies that the operation represented by that operation node can produce the corresponding data represented by that data node as its output. Edges from data node to data node or from operation node to operation node are not allowed.

[0164] The components of computation diagram 1200 can reside within remotely distributed processing elements of wireless communication network 106. For example, the components of computation diagram 1200 can be distributed across specific instances of wireless communication network 106 and within UEs, RANs, multi-access edge computing (MECs), cores, and cloud processors, enabling input data and operations to be located and interconnected within wireless communication network 106 and over Time-Sensitive Networking (TSN) streams. Converged Remote Direct Memory Access over Ethernet (RDMA) (RoCE) is a network protocol that leverages RDMA capabilities to accelerate communication between applications hosted on server clusters and storage arrays, and can include interconnects. TSN over RDMA represents an anticipated possibility for better coordination of message arrival times and operation occurrences. Therefore, computation graph 1200 requires properties representing the network location of each component and the communication channels interconnecting the components, as described in the following reference, which is incorporated herein by reference in its entirety: Neural Network Exchange Format, the Khronos NNEF Working Group, Version 1.0.5, Revision 1, 2022-02-16. In some implementations, computation graph 1200 has an associated graph execution time, which is analogous to the runtime of a programming language. When many such computation graphs are integrated, efficiency will require data to arrive at the operators simultaneously so that data input to the operators is not blocked while waiting for other input data to arrive. This synchronization problem becomes severe over longer channels, such as wireless communication networks (e.g., 5G, 5G+, etc.) interconnecting between operations. As discussed above, a TSN system is provided to ensure that all data arrives at the operator nodes at the optimal time for processing. In some implementations, the computation graph may be provided as an additional input to the TSN scheduler (e.g., configuration module 215).

[0165] Figure 13 The diagram illustrates a flowchart of an example operation of a control system according to one or more embodiments.

[0166] Flowchart 1300 illustrates the operation of the control system (control system 900) discussed above. In some embodiments, the operation of the control system may be implemented by a device for transmitting control signals (e.g., control input 912) via a communication network (e.g., wireless communication network 106), the communication network being configured to support deterministic communication based on a time-sensitive networking (TSN) mechanism. In some embodiments, the device may include a processor and a memory communicatively coupled to the processor. In some embodiments, the memory stores one or more instructions that, when executed by the processor, cause the processor to perform the following operations as shown in flowchart 1300.

[0167] The operation includes operation 1310 for sending quality of service requirement information for transmitting control signals to an end device (e.g., end device 102). Figure 10 As discussed above, operation 1310 can be performed by the control application host (e.g., operation 1010). Quality of Service (QoS) requirement information may include data path requirement information as a network slice SLA to the wireless communication network 106, and provisioning is instantiated using network functions 126 such as AMF and SMF, and TSN-AF 124. In some embodiments, QoS requirements may include TSN gating. In some embodiments, QoS requirement information includes computational latency information and communication latency information required for transmitting control signals. In some embodiments, the wireless communication network 106 provides network function 126 to configure data path provisioning for transmitting control signals generated by the controller based on control laws, based on data path requirements. In some embodiments, control signals are configured to control the operation of end devices in a system associated with the communication network. As discussed above, control signals are generated based on the controller's control laws to control the operation of end devices (e.g., sensors, actuators, etc.).

[0168] The operation includes operation 1320, which polls the communication network to obtain operational parameter information related to the network's operational parameters. For example... Figure 10 As discussed herein, operation 1320 can be performed by the AI / ML control application 910 (e.g., operations 1020 and 1030). In some embodiments, operational parameter information related to the operational parameters of the communication network includes network state information regarding bandwidth, latency, and predicted load along the data path to the end device.

[0169] The operation includes operation 1330 for sending control signals to the end device based on quality of service (QoS) requirement information. In some embodiments, the communication network may configure the data path for sending control signals based on QoS requirements. The control signals are based on a control law. For example... Figure 10As discussed herein, operation 1330 can be performed by AI / ML control application 910 (e.g., operation 1040).

[0170] The operation includes operation 1340 for receiving operation parameter information from components of the communication network in response to polling of operation parameters, and receiving feedback information (e.g., feedback information 935) from the end device. Figure 10 As discussed herein, operation 1340 can be performed by AI / ML control application 910 (e.g., operation 1050).

[0171] The operation includes operation 1350, which uses artificial intelligence (AI) functionality to update the control law of the device for transmitting the control signal based on operating parameter information and feedback information, in order to compensate for delays in the transmission of the control signal. For example... Figure 10 As discussed herein, operation 1350 can be performed by AI / ML system 920 (e.g., operation 1060).

[0172] like Figure 10 As discussed herein, in some embodiments, the operation further includes operations for generating new data path requirements by the AI / ML control application 910 based on the updated control law (e.g., operation 1070).

[0173] As discussed above, in some embodiments, the communication network includes multiple communication network (CN) components, such as components 110, 118, 120, 122, 124, 126, 140, 190, etc. Each CN component is configured to provide discrete functionality for transmitting control signals to end devices across the communication network. In some embodiments, at least one of the multiple CN components is configured to support time-sensitive deterministic communication as a TSN block in the TSN network, according to TSN scheduling of the TSN mechanism.

[0174] As discussed above, in some embodiments, the communication network includes a TSN configuration module (e.g., a TSN scheduler, configuration module 215) configured to update the TSN schedule based on the updated control law. Each of the plurality of CN components is configured as a TSN block with the updated TSN schedule.

[0175] In some implementations, the communication network includes a wireless network configured according to standards defined by 3GPP.

[0176] In some implementations, the Neural Network Exchange Format (NNEF) is used in the AI ​​functionality, such as Figure 12 The subject of discussion.

[0177] As discussed above, in some implementations, AI functionality is distributed across and implemented using the resources of the communication network. In some implementations, these resources may include user equipment (UE), radio access network (RAN), multi-access edge computing (MEC), core, and cloud processors.

[0178] like Figure 11 In some implementations discussed, the AI ​​functionality is provided by: a data collection module (e.g., data collection module 1110) for collecting input data and generating training and inference data based on the input data, the input data including information related to the operating parameters of the communication network and feedback information from end devices; a model training module (e.g., model training module 1120) for training an AI model based on the training data; a model inference module (e.g., model inference module 1130) for predicting data based on the inference data using the trained AI model; and an actuator module (actuator module 1140) for outputting feedback data based on the predicted data.

[0179] The operations discussed above can be cycled periodically to adapt to changes. In some implementations, AI / ML trust metrics can be used. AI / ML trust metrics can be indicators of how much people trust a control system, which can be important for life-or-death control systems such as those in cars and airplanes. Trust in a control system can be determined based on performance feedback metrics, such as variation around a target value, rise time, settling time, and overshoot percentage. Trust will depend on how these metrics vary with different inputs and environmental conditions. Therefore, Figure 13 The operations described can be applied to real-time control scenarios defined by 3GPP, such as "AI-assisted signaling," predicting radio channel states, MIMO beam management, and positioning. These scenarios are provided by the following research projects in 3GPP: SA Working Group 1 (SA1): Research and work projects on the migration of AI / ML models in 5G; SA Working Group 2 (SA2): Research projects on 5G system support for AI / ML-based services; SA Working Group 3 (SA3): Research projects on security and privacy of AI / ML-based services and applications; SA Working Group 4 (SA4): Research projects on AI / ML for media; SA Working Group 5 (SA5): Research projects on AI / ML management; RAN Working Group 3 (RAN3): Work projects on AI / ML for next-generation RAN (NG-RAN); and RAN Working Groups 1, 2, and 4 (RAN1, RAN2, and RNA4): Research projects on AI / ML for new radio (NR) air interfaces.

[0180] Figure 14A block diagram of an example control system according to one or more embodiments is shown.

[0181] like Figure 13 As discussed herein, the control system can be implemented by a device 1400 for transmitting control signals (e.g., control input 912) via a communication network (e.g., wireless communication network 106), which is configured to support deterministic communication based on a time-sensitive networking (TSN) mechanism. Figure 14 As shown, device 1400 includes a processor 1410 and a memory 1420 communicatively coupled to the processor 1410. As... Figure 13 As discussed herein, memory 1420 stores one or more instructions that, when executed by processor 1410, cause processor 1410 to perform the following operations as shown in flowchart 1300. Processor 1410 may be a microprocessor or multi-core processor, integrated circuit, field-programmable gate array, etc., that executes instructions (e.g., instructions stored in memory device 1420) to update the control laws of device 1400 based on artificial intelligence (AI) capabilities. Memory 1420 stores computer program instructions in a machine-readable or computer-readable medium (also referred to as a computer-readable storage medium, machine-readable medium, or machine-readable storage medium).

[0182] In some embodiments, an apparatus (e.g., apparatus 1400) for transmitting control signals via a communication network configured to support time-sensitive deterministic communication based on a Time-Sensitive Networking (TSN) mechanism includes: a processor (e.g., processor 1410); and a memory (e.g., memory 1420) communicatively coupled to the processor; and one or more instructions stored in the memory, which, when executed by the processor, cause the processor to: send quality of service requirement information to an end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; poll components of the communication network to obtain operating parameter information related to operating parameters of the communication network; transmit the control signal based on the quality of service requirement information, wherein the control signal is based on a control law; receive the operating parameter information from components of the communication network and receive feedback information from the end device in response to polling of the operating parameters; and update the control law of the apparatus for transmitting the control signal based on the operating parameter information and the feedback information, based on artificial intelligence (AI) functionality, to compensate for delays in the transmission of the control signal.

[0183] In some embodiments, a method for transmitting a control signal via an associated communication network configured to support time-sensitive deterministic communication based on a Time-Sensitive Networking (TSN) mechanism includes: sending quality of service (QoS) requirement information to an end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; polling components of the communication network to obtain operating parameter information related to operating parameters of the communication network; transmitting the control signal to the end device based on the QoS requirement information, wherein the control signal is based on a control law; receiving the operating parameter information from components of the communication network and receiving feedback information from the end device in response to polling of the operating parameters; and updating the control law of the device for transmitting the control signal based on the operating parameter information and the feedback information using artificial intelligence (AI) functionality to compensate for delays in the transmission of the control signal.

[0184] In some embodiments, a non-transitory computer-readable medium storage device is configured to have one or more instructions that, when executed by a processor, cause the processor to perform the methods described above.

[0185] 3GPP-designated components can be controlled objects (target elements controlled by a control system). As an example, a 3GPP-designated RAN can be controlled in a manner that provides deterministic operation (reducing frame delay variations and jitter). Specifically, the parameter control defined in Technical Specification (TS) 38.322 Radio Link Control (RLC) protocol specification—transmission of upper-layer PDUs, sequence numbering, error correction via ARQ, segmentation and resegmentation of RLC SDUs, reassembly of SDUs, duplicate detection, RLC SDU discarding, RLC reconstruction, and error detection—can be dynamically adjusted during operation to produce stable periodic delays as requested by the TSN scheduler for a specific TSN flow passing through the 3GPP system.

[0186] Those skilled in the art will understand that the various illustrative blocks, modules, elements, components, methods, and algorithms described herein can be implemented as electronic hardware, computer software, or a combination of both. To illustrate this interchangeability between hardware and software, the various illustrative blocks, modules, elements, components, methods, and algorithms have been described above generally according to their functionality. Whether this functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the system as a whole. The described functionality can be implemented in different ways for each specific application. Without departing from the scope of the subject matter, the various components and blocks can be arranged in different ways (e.g., arranged in different orders, or partitioned in different ways).

[0187] It should be understood that the specific order or hierarchy of steps in the disclosed process is an illustration of the example method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged. Some of these steps may be performed simultaneously. The appended method claims present the elements of each step in an illustrative order and are not intended to limit one to the presented specific order or hierarchy.

[0188] The preceding description is provided to enable any person skilled in the art to practice the various aspects described herein. The preceding description provides various examples of the subject matter, and the subject matter is not limited to these examples. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may also be applied to other aspects. Therefore, the claims are not intended to limit the aspects shown herein, but are intended to conform to the full scope consistent with the language of the claims, wherein reference to an element in the singular is not intended to mean “one and only one,” but rather “one or more,” unless specifically stated otherwise. Unless specifically stated otherwise, the term “some” refers to one or more. Masculine pronouns (e.g., his) include feminine and neutral genders (e.g., her and its), and vice versa. Titles and subtitles (if any) are used for convenience only and do not limit the disclosure described herein.

[0189] The predicates “configured to,” “operable for,” and “programmed to” do not imply any specific tangible or intangible modification to the subject, but are intended to be used interchangeably. For example, a processor configured to monitor and control operations or components can also mean that the processor is programmed to monitor and control operations, or that the processor is operable for monitoring and controlling operations. Similarly, a processor configured to execute code can be interpreted as being programmed to execute code or being operable for executing code.

[0190] As used herein, the term "automatic" can include actions performed by a computer or machine without user intervention; for example, actions performed in response to instructions made by a predicate action by a computer or machine or other initiation mechanism. The word "example" is used herein to mean "serving as an example or illustration." Any aspect or design described herein as an "example" is not necessarily to be construed as being more preferred or advantageous than other aspects or designs.

[0191] For example, phrases such as "aspect" do not imply that such an aspect is essential to the art, or that such an aspect applies to all configurations of the art. Disclosure relating to one aspect may apply to all configurations or one or more configurations. One aspect may provide one or more examples. For example, phrases such as "one aspect" may refer to one or more aspects, or vice versa. For example, phrases such as "embodiment" do not imply that such an embodiment is essential to the art, or that such an embodiment applies to all configurations of the art. Disclosure relating to one embodiment may apply to all embodiments or one or more embodiments. One embodiment may provide one or more examples. For example, phrases such as "one embodiment" may refer to one or more embodiments, or vice versa. For example, phrases such as "configuration" do not imply that such a configuration is essential to the art, or that such a configuration applies to all configurations of the art. Disclosure relating to one configuration may apply to all configurations or one or more configurations. One configuration may provide one or more examples. For example, phrases such as "one configuration" may refer to one or more configurations, or vice versa.

[0192] All structural and functional equivalents of the elements throughout the various aspects described in this disclosure that are known or subsequently known to those skilled in the art are expressly incorporated herein by reference and are intended to be covered by the claims. Furthermore, nothing disclosed herein is intended to be offered to the public, whether or not such disclosure is expressly stated in the claims. No claim element shall be construed in accordance with 35 USC § 112(f) unless the element expressly states the phrase “means for” or, in the case of a method claim, uses the phrase “step for”.

Claims

1. An apparatus for transmitting control signals via a communication network, said communication network being configured to support time-sensitive deterministic communication based on a time-sensitive networking (TSN) mechanism, said apparatus comprising: processor; as well as A memory, communicatively coupled to the processor, stores one or more instructions that, when executed by the processor, cause the processor to: Sending quality of service requirement information to the end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; Poll the components of the communication network to obtain operating parameter information related to the operating parameters of the communication network; The control signal is sent to the terminal device based on the service quality requirement information, wherein the control signal is based on a control law; Responding to polling of the operating parameters, the system receives operating parameter information from components of the communication network and receives feedback information from the end device; and Based on artificial intelligence (AI) functionality, the control law of the device used to transmit the control signal is updated based on the operating parameter information and feedback information to compensate for the delay in the transmission of the control signal.

2. The device according to claim 1, wherein, The processor is further configured to: New data path requirements are generated based on the updated control law.

3. The device according to claim 1, wherein, The quality of service requirement information includes computational delay information and communication delay information required for transmitting the control signals.

4. The device according to claim 1, wherein, The operational parameter information related to the operational parameters of the communication network includes network state information regarding bandwidth, latency, and predicted load along the data path to the end device.

5. The device according to claim 1, wherein, The communication network includes: Multiple communication network (CN) components, each CN component being configured to provide discrete functions for transmitting control signals to the end device across the communication network. At least one of the plurality of CN components is configured to support time-sensitive deterministic communication as a TSN block in the TSN network, according to the TSN scheduling of the TSN mechanism.

6. The device according to claim 5, wherein, The communication network includes: The TSN configuration module is configured to update the TSN schedule based on the updated control law, and each of the plurality of CN components is configured as a TSN block with the updated TSN schedule.

7. The device according to claim 1, wherein, The communication network includes wireless networks configured according to standards defined by 3GPP.

8. The device according to claim 1, wherein, The AI ​​functions are distributed across the resources of the communication network and implemented through those resources.

9. The device according to claim 8, wherein, The resources include user equipment (UE), radio access network (RAN), multi-access edge computing (MEC), core and cloud processors.

10. The device according to claim 1, wherein, The Neural Network Exchange Format (NNEF) is used in the AI ​​function.

11. The device according to claim 1, wherein, The AI ​​functionality is provided by the following: The data collection module is used to collect input data and generate training data and inference data based on the input data. The input data includes the information related to the operating parameters of the communication network and the feedback information from the terminal device. The model training module is used to train an AI model based on the training data; A model inference module is used to predict data based on the inference data using a trained AI model; as well as The actuator module is used to output feedback data based on the predicted data.

12. A method for transmitting control signals via an associated communication network, said communication network being configured to support time-sensitive deterministic communication based on a time-sensitive networking (TSN) mechanism, said method comprising: Sending quality of service requirement information to the end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; Poll the components of the communication network to obtain operating parameter information related to the operating parameters of the communication network; The control signal is sent to the terminal device based on the service quality requirement information, wherein the control signal is based on a control law; Responding to polling of the operating parameters, the system receives operating parameter information from components of the communication network and receives feedback information from the end device; and Based on artificial intelligence (AI) functionality, the control law of the device used to transmit the control signal is updated based on the operating parameter information and feedback information to compensate for the delay in the transmission of the control signal.

13. The method of claim 12, comprising: New data path requirements are generated based on the updated control law.

14. The method according to claim 12, wherein, The information related to the data path requirements includes computational delay information and communication delay information required for transmitting the control signals.

15. The method according to claim 12, wherein, The information related to the operating parameters of the communication network includes network status information regarding bandwidth, latency, and predicted load along the data path to the end device.

16. The method according to claim 12, wherein, The communication network includes: Multiple communication network (CN) components, each CN component being configured to provide discrete functions for transmitting control signals to the end device across the communication network. At least one of the plurality of CN components is configured to support time-sensitive deterministic communication as a TSN block in the TSN network, according to the TSN scheduling of the TSN mechanism.

17. The method according to claim 16, wherein, The communication network includes: The TSN configuration module is configured to update the TSN schedule based on the updated control law, and each of the plurality of CN components is configured as a TSN block with the updated TSN schedule.

18. The method of claim 12, wherein the communication network comprises a wireless network configured according to standards defined by 3GPP.

19. The method according to claim 12, wherein, The AI ​​functions are distributed across the resources of the communication network and implemented through those resources.

20. A non-transitory computer-readable medium storage device, the non-transitory computer-readable medium storage device storing one or more instructions, the one or more instructions causing the processor, when executed by a processor, to: Sending quality of service requirement information to the end device for transmitting the control signal, the control signal being configured to control the operation of the end device in a system associated with the communication network; Poll the components of the communication network to obtain operating parameter information related to the operating parameters of the communication network; The control signal is sent to the terminal device based on the service quality requirement information, wherein the control signal is based on a control law; Responding to polling of the operating parameters, the system receives operating parameter information from components of the communication network and receives feedback information from the end device; and Based on artificial intelligence (AI) functionality, the control law of the device used to transmit the control signal is updated based on the operating parameter information and feedback information to compensate for the delay in the transmission of the control signal.

Images