Method in a telecommunications system, telecommunications system, user equipment and base station and methods therein

Abstract

A method in a telecommunications system, a telecommunications system, a user equipment, and a base station, and methods therein are provided.A method in a telecommunications system including a UE (User Equipment) on which a local portion of an application that generates application data runs, and a base station on which a remote portion of the application runs, the method comprising: when the local portion of the application in the UE starts running, the UE sending a connect message to the base station to initiate a connection with the base station, the connect message including application parameters generated by the local portion of the application; and the base station passing the application parameters to the remote portion of the application.

Description

[Technical field]

[0001] The present invention relates to a method in a wireless communication network. The present invention further relates to a user equipment and a base station and a method therein.

[0002] In particular, but not exclusively, the present invention relates to techniques for adapting methods and elements in wireless communication networks, such as 4G (LTE-A), 5G (also known as New Radio, NR) wireless communication networks, for use in resource-intensive applications, such as AI (Artificial Intelligence) applications, video processing, real-time control, and the like. [Background technology]

[0003] Wireless communication networks are widely known in which terminals (also called user equipment or UE, subscriber or mobile station) communicate over an air interface with base stations within range of the terminal. This part of the network is known as the Radio Access Network (RAN). Furthermore, communication may occur within the core network between the base stations and other entities, usually over wired connections.

[0004] A geographic area served by one or more base stations is commonly referred to as a cell, and typically a number of base stations are provided in suitable locations to form a system with adjacent and / or overlapping cells covering a wide geographic area more or less seamlessly.

[0005] Typically, many transmission points are provided at suitable locations to form a network covering a wide geographical area more or less seamlessly with adjacent and / or overlapping cells (herein the terms "system" and "network" are used synonymously). For each cell, a base station providing or at least managing the transmission points divides the available bandwidth, i.e. frequency and time resources, into individual resource allocations for the terminals served by the cell. In this way, signals transmitted in a cell and scheduled by the base station have a specific location in the frequency and time domain. The terminals are mobile and therefore may move between cells, prompting the need for handover of the terminal's connection to the network when the terminal moves between adjacent cells. A terminal may be in range of (i.e. be able to detect signals from and / or communicate with) several cells at the same time, but in the simplest case the terminal communicates with one "serving" cell. A terminal can usually only access a single network provided by a single operator. However, networks / operators may cooperate.

[0006] A Radio Access Technology, RAT, is the basic physical connection method for radio-based (wireless) communication networks. One Radio Access Technology, RAT, or one type of radio system is based on a set of standards called Long-Term Evolution, LTE, or LTE-A (Advanced) for later versions. In the system topology, in LTE (used generally here for LTE and LTE-A), each terminal, i.e., UE, connects wirelessly to a base station in the form of an enhanced Node B or eNB via a radio / air interface (Uu). This base station may be functionally separated into a baseband unit (BBU), which generates and processes digitized baseband RF signals, and a remote radio head, which generates an analog transmit RF signal from the baseband signal and feeds it to the antenna, while digitizing the RF receive signal.

[0007] It should be noted that the above refers by default to UEs operated by human users, for example in the form of mobile phones, laptop computers and PDAs or tablets. However, wireless communication networks may also be used for so-called Machine Type Communication (MTC) used in the Internet of Things (IoT), where MTC is a form of data communication involving one or more entities that does not necessarily require human interaction. Entities involved in MTC, hereafter referred to as MTC devices (or terminals), are considered as a type of UE, unless the context requires otherwise. Applications of MTC devices include sensors, video / imaging devices, fleet management, smart metering, product tracking, home automation, e-health, etc. MTC or IoT devices, in contrast to mobile devices of human users, are often in a fixed location or may be located, for example, in a drone or other vehicle.

[0008] The next generation of wireless access systems following LTE, known as "5G" or "NR" (New Radio), are necessary to meet modern requirements for increased bandwidth, efficiency and flexibility.

[0009] As part of the Physical Layer 1 (PHY) design in 5G, the traditional concept of both scheduling resources and housing physical antennas (either for human use or as part of the IoT) for wireless communication with terminals is becoming more fluid. The terminology used for 5G / NR includes a "gNB" (next-generation Node B) that manages (locally or remotely) at least one transmission point. Such a transmission point may also function as a reception point (replacing the RRH of LTE) and is typically called a TRP or TRxP (Transmission / Reception Point). Another change in 5G is that in 5G, the original BBU (Base Band Unit) of 4G / LTE is now split into a CU (Central Unit) and a DU (Distributed Unit).

[0010] In the 4G Core Network (CN), called EPC (Evolded Packet Core), protocols and reference points (interfaces) are defined for each entity, such as MME (Mobility Management Entity), S-GW (Serving Gateway), P-GW (Packet Data Network Gateway), etc., as described above. Alternatively, in the 5G Core (NGC), protocols and reference points (interfaces) are defined for each NF (Network Function). NFs may be implemented either as network elements on dedicated hardware, as software instances running on dedicated hardware, or as virtualized functions (not limited to specific hardware) instantiated on a suitable platform, e.g., cloud infrastructure.

[0011] FIG. 1 depicts a non-roaming reference 5G system architecture.

[0012] Specification 3GPP® TS 23.501 describes the architecture for 5G systems. The 5G architecture is defined as service-based, and interactions between network functions are expressed in two ways: A service-based representation, where a network function (NF) in the control plane (e.g., AMF) allows other authorized network functions to access their services. This representation also includes point-to-point reference points, if necessary. A reference point representation, which shows the interactions that exist between NF services in the network function described by a point-to-point reference point (e.g., N11) between any two network functions (e.g., AMF and SMF).

[0013] Service-based interfaces are used within the control plane. The 5G system architecture consists of the following Network Functions (NFs) (not all of them are shown in Figure 1): AUSF (Authentication Server Function) AMF (Access and Mobility Management Function) DN (Data Network, e.g. operator services, internet access, or third-party services) UDSF (Unstructured Data Storage Function) NEF (Network Exposure Function) NRF (Network Repository Function) NSSAAF(Network Slice Specific Authentication and Authorization Function) NSSF(Network Slice Selection Function) PCF (Policy Control Function) SMF (Session Management Function) Unified Data Management (UDM) Unified Data Repository (UDR) UPF (User Plane Function) UCMF(UE radio Capability Management Function) AF (Application Function) UE (User Equipment) (R)AN ((Radio) Access Network) 5G-EIR(5G-Equipment Identity Register) NWDAF (Network Data Analytics Function) CHF (CHarging Function)

[0014] The 5G system architecture also includes the following network entities: SCP(Service Communication Proxy) SEPP (Security Edge Protection Proxy)

[0015] The UE is shown at the bottom left and is connected to the RAN (gNB). There is also a direct connection between the UE and the Access and Mobility Management Function (AMF) via the N1 interface, mainly for non-access stratum (NAS) signaling. The User Plane Function (UPF) is one of the network functions (NFs) of the 5G Core Network (5GC) mentioned above. The UPF is responsible for packet routing and forwarding, packet inspection, QoS processing, and external PDU sessions to interconnect with the Data Network (DN) in the 5G architecture. The gNB (RAN) connects with the UPF via the N3 interface (the N9 interface is between two different UPFs (not shown)). The N4 interface connects the UPF to the Session Management Function (SMF), and the N2 interface connects the RAN (i.e., the gNB) to the AMF. The N6 interface is for the Data Network (DN), which is where the connection to the server / entity running the AI ​​or other algorithmic applications is likely to be located.

[0016] In the architecture shown in FIG. 1, UP (User Plane) data is transferred to DN (Data Network) via UPF (User Plane Function).

[0017] A recent development in RAN architecture is the O-RAN architecture (Open Radio Access Network).

[0018] The decomposition of the base station to separate the RRHs and the BBU creates the need for a fronthaul transport network between these two entities capable of carrying the antenna samples. This fronthaul may use the Common Public Radio Interface (CPRI) or Open Base Station Architecture Initiative (OBSAI) protocols.

[0019] CPRI is the more popular option and defines a simple synchronization protocol for mapping user data transmitted from the radio units. This simplicity allows for cost-effective small form factor designs and allows for coordination with base stations and features like CoMP, but the bandwidth requirements are high. Another concern is with MIMO scaling, and CPRI requires a dedicated link for every antenna. An alternative solution is needed, and this is where O-RAN comes in.

[0020] O-RAN provides a distinct approach to mobile fronthaul and midhaul networks, built on cloud principles, enabling interoperability and standardization of RAN elements. O-RAN works with existing technologies such as 4G, 5G (NR) and interfaces with EPC or NGC.

[0021] A simplified O-RAN architecture is shown in Figure 2 using an example of NR. Here, the RRH, DU, and CU in 5G are replaced with O-RU, O-DU, O-CU-CP (O-RAN Central Unit - Control Plane), and O-CU-UP (O-RAN Central Unit - User Plane), and the CU is divided into a control plane and a user plane. The key point in fronthaul design is the "functional split" that separates the functions of layer 1 (physical layer), layer 2 (data link layer), and layer 3 (transport layer) present in the RRH (O-RU) from the functions among these processing functions that exist in the BBU. O-RAN encompasses different splits.

[0022] The O-RU is an O-RAN distributed unit that is a logical node, e.g., hosts the RF / low physical layer based on the lower layer function split.

[0023] The O-RAN Distributed Unit (O-DU) is connected to the O-RU and contains the remaining lower layer functional splits (one of the various different splits between the O-RU and O-DU) and the RLC and MAC layers. The two adjacent elements in the diagram are the O-CU-CP and the O-CU-UP. Both units are connected to the O-RAN near real-time RAN intelligent controller (RT RIC) and there is also a non-RT RIC (O-RAN non-real-time RAN intelligent controller). The non-RT functions include service and policy management, RAN analysis, and model training for the near-RT RAN functions. The trained models and real-time control functions generated in the non-RT of the RIC are distributed to the near-RT of the RIC for runtime execution. The O-RAN reference architecture provides next generation RRM with embedded intelligence while optionally accommodating legacy Radio Resource Management (RRM). This resides in the near-RT functional layer of the RIC.

[0024] Turning to the interfaces between these elements, E2 is a logical interface that connects the quasi-RT RIC with the E2 nodes (O-DU, O-CU-CP, O-CU-UP). The E2 interface supports control plane protocols (no UP).

[0025] The E2 supports both quasi-RT RIC services (REPORT, INSERT, CONTROL, and POLICY) and quasi-RT RIC support functions, including E2 interface management (E2 setup, E2 reset, and general error condition reporting) and quasi-RT RIC service updates (i.e., capability exchange regarding the list of E2 node functions exposed on the E2).

[0026] The E2 interface includes O-RAN RRC Information Element (IE) data collection, including UE measurement and configuration information, which is achieved by directly accessing the RRC IEs from the RRC protocol running in the O-CU in the O-RAN-gNB.

[0027] The A1 interface between quasi-RT RIC and non-RT RIC supports O-RAN RRC IE data collection, including network measurements and context information. Summary of the Invention

[0028] It is desirable to adapt current wireless technologies to enable bandwidth and data efficient use of applications running on terminals and / or base stations, particularly computationally intensive applications such as AI applications.

[0029] According to a first aspect of the present invention, there is provided a method in a telecommunications system including a UE (User Equipment) on which a local part of an application that generates application data is executed and a base station on which a remote part of the application is executed, the method comprising: when the local part of the application in the UE starts executing, the UE sending an initial connection message to the base station to initiate a connection with the base station, the initial connection message including application parameters generated by the local part of the application; and the base station passing the application parameters to the remote part of the application.

[0030] The application parameters may be used in a number of different ways in the remote application part: for example, they may simply contain useful information, they may control how the remote and / or local application parts perform (e.g., based on knowledge at the UE of the local situation), or they may be used to control / change some aspects of the data transmission to / from the UE, such as changing the allocated data rate / resources.

[0031] Thus, in response to receiving the application parameters, the base station may send feedback to the UE from the remote portion of the application, and / or modify execution of the application, and / or modify one or more transmission parameters in response to the application parameters.

[0032] An application is assumed to be distributed, including at least one local part at a UE and at least one remote part at a base station. In some arrangements, there are multiple local parts (one per UE) that feed into a single remote part, for example, in a base station and / or a core network connected to the base station. In one arrangement within the scope of the claims, a distributed application may include separate applications executed locally and remotely that are combined to provide the functionality of a single distributed application.

[0033] The application parameters may be transmitted on a data channel together with the application data. In another arrangement, the application parameters may be transmitted on a control channel. For example, this may be the control channel used to set up the UE connection with the base station. Thus, the application parameters may be embedded within a Radio Resource Control (RRC) message.

[0034] The method may further include performing a transfer of application data generated by the application between the UE and the base station, which may be in parallel with the application parameters.

[0035] A remote application may run directly on the base station or on an entity in a data network directly connected to the base station. In one arrangement, the application itself runs in the data network, but is provided with data from an algorithm running in the base station. Similarly, a local application may run directly on the UE or on a device directly connected to the UE (e.g., a device connected to the UE via a wireless network such as a LAN, or via Bluetooth or an equivalent protocol).

[0036] The method is suitable for use in any wireless communication system. In one arrangement, the telecommunications system is a New Radio (NR) (or 5G) system and the method further includes routing the application data from the local application to the remote application via a User Plane Function (UPF).

[0037] The base station may be any base station, in particular an O-RAN (Open Radio Access Network) base station, in which case the application parameters may be passed to the remote application or to an algorithm used by the remote application via an O-CU-UP (O-RAN Central Unit - User Plane).

[0038] The O-RAN base station may pass application parameters to the application via the E2 interface from the O-RAN Central Unit - User Plane (O-CU-UP) to the application. The algorithms used in the remote application may run in the data network or potentially in the Non-RT RIC (Non Real Time Radio Access Network Intelligent Controller). Filtering occurs in the data network or in the O-RAN Central Unit - User Plane (O-CU-UP) and the data is forwarded to the remote application via the UPF.

[0039] The (distributed) application may be a Machine Learning (ML) application and the application parameters may be Artificial Intelligence (AI) indicators. In one example, the AI ​​indicators are used in the inference and / or filtering stages of the ML algorithm, preferably to control filtering of application data transmitted between the UE and the base station.

[0040] According to a further aspect of the present invention, there is provided a telecommunications system including a UE (User Equipment) and a base station, wherein a local part of an application that generates application data is executed in the UE and a remote part of the application is executed in the base station, and when the local part of the application starts to execute, the UE is configured to send an initial connection message to the base station to initiate a connection with the base station, the initial connection message including application parameters generated by the local part of the application, and the base station is configured to pass the application parameters to the remote part of the application.

[0041] According to yet another aspect of the present invention, there is provided a User Equipment (UE) in a telecommunications system including a UE on which a local part of an application that generates application data is executed, and a base station on which a remote part of the application is executed, wherein when the local part of the application starts to execute, the UE is configured to send an initial connection message to the base station to initiate a connection with the base station, the initial connection message including application parameters generated by the local part of the application.

[0042] According to a corresponding method aspect, there is provided a method in a User Equipment (UE) in a telecommunications system including a UE on which a local part of an application that generates application data is executed and a base station on which a remote part of the application is executed, wherein when the local part of the application starts executing, the UE sends an initial connection message to the base station to initiate a connection with the base station, the initial connection message including application parameters generated by the local part of the application.

[0043] According to yet another aspect of the present invention, there is provided a base station in a telecommunications system including a UE (User Equipment) on which a local part of an application that generates application data is executed, and a base station on which a remote part of the application is executed, the base station being configured to receive an initial connection message sent from the UE to initiate a connection with the base station when the local part of the application starts execution, the initial connection message including application parameters generated by the local part of the application, and the base station being configured to pass the application parameters to the remote part of the application.

[0044] According to a corresponding method aspect, there is provided a method in a base station in a telecommunications system including a UE (User Equipment) in which a local part of an application that generates application data is executed, and a base station in which a remote part of the application is executed, wherein the base station receives an initial connection message sent from the UE to initiate a connection with the base station when the local part of the application starts execution, the initial connection message including application parameters generated by the local part of the application, and the base station passes the application parameters to the remote part of the application.

[0045] A UE / system / base station according to preferred embodiments of the present invention may include any combination of the method aspects described above. Methods according to embodiments of the invention may be described as computer implemented in that they require processing and memory capabilities.

[0046] An apparatus according to a preferred embodiment is configured or arranged to perform certain functions. This configuration or arrangement can be through the use of hardware or middleware, or any other suitable system. In a preferred embodiment, the configuration or arrangement is through software.

[0047] According to a further aspect, there is provided a computer program comprising instructions, when executed by a UE, a base station, or a network, for configuring the base station or the network to perform the steps of the method according to any of the preceding method definitions or any combination thereof. The computer program may be stored in a non-transitory computer readable medium.

[0048] In general, the hardware referred to may include elements configured or arranged to provide a defined functionality, for example, this hardware may include a receiver, a transmitter (or a combined transceiver), a processor, memory / storage media, a user interface, and other hardware components commonly found in a terminal.

[0049] The invention may be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or a combination thereof. The invention may be implemented as a computer program or computer program product, i.e. a computer program tangibly embodied in a non-transitory information carrier, for example a machine-readable storage device, or in a propagating signal, for execution by or for controlling the operation of one or more hardware modules. The computer program may be in the form of a stand-alone program, computer program part, or multiple computer programs, written in any form of programming language, including compiled or interpreted languages, and deployed in any form, as a stand-alone program, or as a module, component, subroutine, or other unit suitable for use in a data processing environment. The computer program may be deployed to be executed on one module or multiple modules at one site, or distributed across multiple sites on a vehicle or in a back-end system, and interconnected by a communication network.

[0050] The steps of the methods of the present invention may be performed by one or more programmable processors executing computer programs to perform the functions of the present invention by operating on input data and generating output data.

[0051] The present invention has been described with reference to specific embodiments. Other embodiments are within the scope of the following claims. For example, the steps of the invention may be performed in a different order and still achieve desirable results. [Brief description of the drawings]

[0052] Reference is made, by way of example only, to the accompanying drawings, in which:

[0053] [Figure 1] A conceptual diagram of 5G core architecture. [Diagram 2] A conceptual diagram of O-RAN architecture. [Diagram 3] FIG. 2 is a block diagram illustrating application processing shared between remote and local applications. [Figure 4] FIG. 4 is a block diagram of the same components as FIG. 3, showing the AI ​​data filtering configuration in the UE and the AI ​​algorithm. [Diagram 5] FIG. 1 is a block diagram of an example O-RAN with data flow. [Figure 6] FIG. 1 is a block diagram of an example of O-RAN with existing ML processing. [Figure 7] FIG. 13 is a block diagram of an example of O-RAN with a new transfer of application parameters to a remote AI module. [Figure 8] FIG. 8 is the block diagram of FIG. 7 including explicit parameters available in the E2 interface. [Figure 9] FIG. 1 is a block diagram of an AI / ML algorithm as part of an O-RAN gNB. [Figure 10] FIG. 1 is a block diagram of an AI / ML algorithm with E2 and A1 information. [Figure 11] FIG. 1 is a block diagram of an AI / ML algorithm with separation of remote applications from the base station. [Figure 12] FIG. 1 is a block diagram of an AI / ML algorithm with PDCP / SDAP modifications. [Figure 13]1 is an overall flowchart. [Figure 14a] 4 is a flowchart of an initial connection of a UE to a base station with contention resolution. [Figure 14b] 1 is a flowchart of an initial connection of a UE to a base station without contention resolution. [Figure 15] 4 is an exemplary structure of application parameters transmitted with data. [Figure 16] 13 is an exemplary structure of an application parameter that is not transmitted with the data. [Figure 17] FIG. 2 is a block diagram of protocol stacks for the user plane and control plane. [Figure 18] 1 is a block diagram of a base station or a UE. [Figure 19] FIG. 2 is a block diagram of a computing device, such as a server at DN, on which an algorithmic application may execute. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0054] The arrangements herein include procedures and signaling in the wireless network to help control AI / ML (or other computationally intensive) processing that occurs in close cooperation with the base station.

[0055] This includes controlling the inference / filtering of applications using AI / ML. The method herein uses the concept of distributed applications, as shown in Figure 3, where a local application part 10 connected to a UE 20 and modem is connected over the air interface to a remote application part running on a base station 30 (denoted as ORAN-gNB). The local application running on the UE generates data, which is sent over the air interface via the UE modem to the ORAN-gNB, where the remote part of the application is running.

[0056] The advantage of this type of distributed application is that the processing of data may be shared between local processing in the devices connected to the network and the network equipment itself. It is expected that the amount of data processing performed locally and remotely may change dynamically at any time. The advantage of performing processing locally is that there is less delay between the generation of data and its subsequent processing, and the disadvantage is that the processing uses local processing resources, which are typically scarce, and possibly power resources for battery-operated local devices. The advantage of remote processing is that more intensive data processing may be easily performed, and remote processing is generally not power limited. Remote processing may also more easily use knowledge of global system parameters and settings from other devices connected to the network, as well as network operating conditions.

[0057] One particular application envisioned for this is a machine learning system as shown in Figure 4. The following examples all refer to AI / ML, but it will be understood that they apply equally to other applications running on the UE and base station / DN.

[0058] In Fig. 4, the local application part generates data, which may be filtered before transmitting it over the network using the UE modem and the air interface as before. This data is then forwarded to an AI / ML processing block / module, which is either in the base station or closely coupled to the base station. It is configured to perform further data processing as required by the entire application (including both local and remote elements).

[0059] The data may undergo filtering / inference at the UE local application and at the remote application. Not all data is required to perform data prediction / classification as the output of the application, and filtering may be seen as part of the processing of the ML algorithm. Filtering may be seen as a possible first step in inference, as some data is filtered out and not used in the inference processing step that provides a prediction. In this specification, references to filtering / inference are to any filtering and / or inference that occurs. Both of these processes may act to reduce the amount / volume of data to be transferred.

[0060] Typically, in an ML system, filtering / inference is performed after training has occurred. Inference requires less computing resources and bandwidth than training. Filtering / inference is the part that uses smaller batches of real-time data after training and efficiently generates predictions or classifications based on the data. ML inference is modeled as both filtering and processing in the ML model. By doing this modeling, some of the inference type of processing may be easily moved to different physical nodes in the system. Assuming that training has already occurred, the arrangements herein may control application execution in the base station and / or UE, e.g., supporting the use of small (smaller) packets of data from local applications without the need for extensive higher layer processing and signaling to occur.

[0061] One problem addressed herein is the transfer of initial application data from a local application to a remote application.

[0062] Figure 5 is an overall view of the preferred arrangement. Application data is shown flowing between a local part of the application running on or near the UE via the base station, and a remote (and potentially computationally intensive) part of the application shown as an AI / ML algorithm. The double arrows between the UE and the base station and between the base station and the AI / ML algorithm represent the transmission of application parameters when the UE first initiates a connection to the base station when the application begins to execute. In an alternative arrangement, the remote part of the application runs partially or entirely within the base station.

[0063] Typically, in any application, the startup phase for the application includes the setup of a User Plane (UP) data pipe for the transfer of application data. This can typically take up to 100 ms, significantly contributing to performance degradation for many applications. The arrangements herein solve this problem by embedding the remote application in the base station (or attached to the base station in a DN) and allowing it to have direct access to initial connection messages such as radio parameters from the Radio Resource Configuration (RRC) process that occurs in the base station. The arrangements address the transmission of initial control information from the local application to the remote application. They solve the latency problem by allowing the remote application to have local application information as soon as possible. This is made possible by embedding the initial control information in the very first wireless communication message between the local device and the base station.

[0064] The arrangement here optionally uses real-time (or near real-time) control settings of base stations, for example in the ORAN context, to optimize the performance of AI / ML applications. In this arrangement, real-time (or near real-time) is used to refer to control that occurs simultaneously as data is processed without significant delay. For example, optimization is expected to occur for every packet received as the packet of data is being processed.

[0065] A preferred arrangement may also use parameters sent when a connection is initiated for network controlled configuration of application data filtering to process data streams between applications in the system. The advantage of controlling application data filtering with network signaling is that the needs of the application are more closely coupled with the capabilities and available resources of the device transferring application data over the air interface. For example, a video processing application that may send a high data rate stream to a pattern recognition AI application can reduce the volume of the application data stream by controlling the filtering of the application data. The advantage of this is that the processing needs of the device used for transferring data over the air interface are reduced.

[0066] Figure 6 shows the current status of ML application data processing from a UE / Modem connected to a wireless network.

[0067] In this example and the following examples, a local application (or another device running a local application if the local application does not run directly on the UE) is connected to the UE, and the UE provides data from the local part of the AI ​​application or a similar application to the remote application via the RAN and gNB. The data is routed through the UPF. Filtering is shown as part of the local and remote application, which can be in the UE / modem and / or optionally in the gNB. Although filtering and inference are shown, other processing may be substituted. The remote application has access to the ML algorithms and filtering / inference, but may be provided separately from them. It may be, for example, a parent application in the cloud. The application may include calls to processing in the gNB.

[0068] The illustrated example is an O-RAN example, with the open fronthaul control and user plane shown as the CUS plane in the diagram. The UE modem connects to an O-RAN Radio Unit (O-RU), which is a logical node that hosts the Low-PHY layer and RF processing, based on a preferred hierarchical functional split.

[0069] The data filtered by the local application is transmitted through the base station to the remote AI / ML algorithm in the base station. In fact, in this diagram, there are two logically different functions that include the ML algorithm. The first one is the local application that passes the information to the UE / Modem. The UE / Modem can be a user terminal (e.g., a mobile phone) or a modem (which can be a standalone UE, an MTC terminal or a communication module). The UPF function shown in the AI / ML algorithm module represents the 5G core network function for the user plane function. This UPF is needed because in 5G, the inventors have come to realize that there is no easy access to the UP data flows in the ORAN-gNB and the easiest way to access the UP data is via the UPF. The data is passed to the O-CU-UP via the O-RU, O-DU.

[0070] Several arrangements here take advantage of the architecture and signaling framework opportunities from the use of the O-RAN architecture introduced above (see https: / / www.o-ran.org / , the relevant portions of which are incorporated herein by reference). This is illustrated in Figure 7 with the same portions of local and remote AI applications as described in Figure 6. The basic O-RAN architecture is as shown in Figure 2.

[0071] The bidirectional line connecting the AI / ML algorithm and the E2 interface represents a novel and valuable part of this arrangement, since it is advantageous to be able to access information such as UE measurements and configuration information that is not available from the O-CU-UP connection to the UPF in the AI / ML algorithm. Application parameters from the connection messages are passed via this interface.

[0072] Figure 8 differs from Figure 7 in that it shows explicit information element (IE) transfer, including RAN parameter values ​​and parameter names. These may include RAN capability name, RAN parameter type, RAN parameter ID, RAN parameter value, and RAN parameter name.

[0073] Figure 8 also shows how the O-CU-UP includes processing applied to UP data as part of its Packet Data Convergence Protocol (PDCP) and Service Data Adaptation Protocol (SDAP) functionality. This shows that UP data is available within the O-CU-UP entity, but only at the PDCP and SDAP protocol level (but not as an Internet Protocol (IP) flow, e.g., TCP / IP). That is, UP data cannot be terminated at the gNB (anchor point), but packet inspection is possible.

[0074] Figure 9 shows another arrangement option for the AI / ML algorithms, where the AI / ML algorithms run within the same entity as the ORAN-gNB and are in fact implemented in the near real-time RIC. This allows the AI / ML algorithms to act directly on the information available on the E2 interface without the need for an external interface. In this arrangement, the E2 interface between the O-CU-UP and the near real-time RIC is used to transport application data as well as the connection from the O-CU-UP to the UPF at the PDCP / SDAP protocol level.

[0075] As a further refinement of the architecture, Figure 10 shows how information from the A1 interface, and therefore any data collected in the non-real-time RIC, can be used as part of improving the operation of AI / ML algorithms. The non-real-time RIC contains the service and policy management network management functions. The non-real-time RIC receives and acts on reliable data from the CU and DU in a standardized format over the A1 interface. Algorithms in the non-real-time RIC provide the ability to modify RAN operation based on the operational requirements set by the network operator.

[0076] In another architecture shown in Figure 11, the remote application is moved from the ORAN-gNB and connected to the UPF function of the 5G core network. In this architecture, the ML algorithms reside in the non-real-time RIC and control the filtering / inference of UP data as an additional function placed after the PDCP / SDAP function in the O-CU-U entity. The filtered and / or inferred ML algorithms are provided to the remote application via the UPF.

[0077] A final implementation option is shown in FIG. 12, where the filtering / reasoning functionality is combined as part of the PDCP / SDAP entity, thus allowing this functionality to be included simply by 3GPP specification modifications to the operation of either the PDCP or SDAP protocols.

[0078] What is shared between all these different implementations is the way in which the device initially connects to the network, allowing additional information to be carried in the initial data sent from the device to the base station, as shown in Figure 13.

[0079] At S10, a local AI / ML application is started. At S20, the application generates both application data and application parameters (herein also specifically referred to as "AI indicators") in parallel. At S30, when the UE makes an initial connection to the network (typically moving from an idle state such as RRC_IDLE, RRC_INACTIVE state to a connected state such as RRC connected state), the application parameters are sent to the AI / ML application for algorithmic control. At S40, the base station passes on the application parameters to the AI / ML application at S40, potentially using the E2 interface. At optional step S50, there may be feedback to the local application. For example, information regarding radio bearer or other service and policy management, and RAN analysis information may help fine-tune the AI / ML application.

[0080] The connection in S30 may vary depending on the situation. Typically, when a device connects to a 5G network, it may use one of two different signaling methods, as shown in Figures 14a and 14b.

[0081] At S100 in Figure 14a, the UE sends an AI indicator with MSGA (PRACH preamble and PUSC transmission) and waits for a gNB MSGB response or sends MSG3 for an RRC connection request. At S110, there is a resulting MSGB / MSG4 contention resolution from the gNB to the UE. At S120, optional data transfer in one or both directions follows, followed by an RRC release at S130.

[0082] Meanwhile, in S200 of Fig. 14b, the UE sends an AI indicator together with Configured Grant Type 1, a grant-free uplink transmission also known as TWG (Transmission without grant), i.e. data transmission without resource request. In S210, the gNB responds with DL data transmission on CRNTI, and the UE data may immediately start data transmission. Steps S220 and S230 are the same as S120 and S130.

[0083] Essentially, the two different methods, with or without configured grants (CG), both involve a new AI indicator, which is intended to provide information to the gNB that allows application data to be routed directly to AI / ML processing in the gNB.

[0084] More generally, a similar indicator could be used to indicate that some kind of local processing is performed at the gNB, not necessarily using AI / ML techniques.

[0085] This AI indicator has many different possible use cases, including using this early indication as a signal for AI / ML algorithms running in the gNB.

[0086] Other examples of uses include controlling filtering / inference in AI / ML applications, e.g., switching to a set of known pre-trained inferences without the need for training. Another example is the use of an indicator to indicate that retraining is required, and the application may move to a training mode of operation, where typically more data is needed to facilitate future filtering and inference settings.

[0087] Specifically, the AI ​​indicators / parameters may be structured as one or more bits, depending on the type of information these bits are designed to convey.

[0088] Typically, the design of the parameters is flexible to allow for different uses of the control information used by a particular (e.g., a particular AI / ML) application. For example, in an AI indicator specification, a number of bits may be defined, and the purpose of the bits may not be defined in the specification, but rather left to be defined by the network / service operator.

[0089] An example structure is shown in Figure 15. The application indicator / parameters are sent along with the data used for the application, but are differentiated by using an application data end indicator in the application data packet, for example, inserted after the data and after the parameters, or vice versa. In this case, the parameters are sent in the same data packet as the application data.

[0090] Alternatively, the application parameters are transmitted as separate data, as shown in Figure 16. In this example, the transport of the application parameters may use a control channel, which may be defined separately from the particular wireless communication method used.

[0091] For example, in the case of a 5G / 4G system, the control channel may be a radio resource channel (RRC) in which a control plane (CP) is used to transfer RRC information, as shown in FIG. 17.

[0092] The intent of the AI ​​indication or other parameters is primarily to act as a control signal (not as data) from the UE to the network (where the network includes, for example, architectural components as shown in FIG. 7). The network then acts on receipt of the control information from the UE. The network action may target one specific UE. In another scenario, the network action may depend on receipt of the AI ​​indication from multiple UEs. For example, multiple UEs may run local parts of an application and transmit data to the same base station (or different base stations, but the same remote part of the application). For example, if the AI ​​indicator is transmitted from a set of cameras connected to a UE transmitting data to the network, a network action taken as a result of the parameter, for example, an increase / decrease in the scheduled data rate, may be applied to all cameras in the set.

[0093] Specific examples of AI indicators that act as AI controls include the following:

[0094] The UE sends a request to the AI / ML application that its internal algorithms have detected the need for AI / ML retraining (e.g., the network subsequently sends information to the UE to only send known test data until retraining is complete). The UE transmits an AI indicator that immediately switches the AI / ML algorithms to a different (known) mode of operation, e.g., the UE transmits a different amount of data (i.e., bit rate) from the connected device to the network. The UE may send an AI indicator to request the network to take immediate action, such as disabling filtering of the data (e.g., immediately moving to full rate video data when a video stream is being transmitted). The UE sends an AI control to indicate a problem in device processing, and an AI / ML entity in the network may then attempt, diagnose, and optionally request some additional help / support from network operation. AI (or other non-AI) control may be used to act as a more general indicator of the amount of data required by an application, e.g., when multiple devices are used to transmit data, the devices themselves may determine that they require more bandwidth for their data, and this application level bandwidth request may be processed by an AI / ML algorithm, resulting in a subsequent change in the provision of UL data rates by the base station. This request may be transmitted as part of the data received by and conveyed to the non-real-time RIC, and used by the base station scheduler to determine the appropriate bandwidth. In the case of such bandwidth changes, the information in the base station is also available as part of the radio bearer setup information, e.g., it is assumed that the specific different bandwidth is already known to the application and also known to the bearer management function in the base station.

[0095] Such an indication may be a single or multiple bit, depending on the granularity of the information required.

[0096] Hardware Implementation

[0097] 18 shows a hardware structure of a base station or UE 800 suitable for use in embodiments of the present invention, including an antenna 802, a transmission and reception unit 804, a controller 806, and a storage medium or memory 808. The base station / UE 800 may be an example of the UE 20 or base station 30 in FIG. 5 and elsewhere.

[0098] A base station or UE belongs to at least one RAT, and a base station may, for example, be described as an eNB, may take the form of a NR / 5G, may be described as a gNB or an ORAN-gNB, etc. Other / different base stations may take other forms of different RATs, as long as they are suitable for transmitting and receiving signals from other base stations.

[0099] In any embodiment, the controller 806 may be, for example, a microprocessor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other logic circuitry programmed or otherwise configured to perform the various functions described above. For example, the various functions described above may be embodied in the form of a computer program stored in the storage medium 808 and executed by the controller 806.

[0100] Figure 19 is a block diagram of a computing device that may be used to implement the methods in a telecommunications system as described herein. For example, the computing device may store and execute local and remote portions of the applications shown in Figure 3 and elsewhere. The computing device includes a processor 993 and memory 994. Optionally, the computing device also includes a network interface 997 for communication with other computing devices, e.g., other computing devices of invention embodiments.

[0101] For example, an embodiment may be comprised of a network of such computing devices. Optionally, the computing devices also include one or more input mechanisms, such as a keyboard and mouse 996, and one or more display units, such as a monitor 995. The components may be connected to each other via a bus 992.

[0102] The memory 994 may include a computer-readable medium. The term may refer to a single medium or multiple media (e.g., centralized or distributed databases and / or associated caches and servers) configured to carry computer-executable instructions or store data structures. Computer-executable instructions may include, for example, instructions and data accessible by a general purpose computer, a special purpose computer, or a special purpose processing device (e.g., one or more processors) to perform one or more functions or operations. Thus, the term "computer-readable storage medium" may also include any medium capable of storing, encoding, or carrying a set of instructions for execution by a machine and causing the machine to perform any one or more of the methods of the present disclosure. Thus, the term "computer-readable storage medium" may be interpreted to include, but is not limited to, solid-state memory, optical media, and magnetic media. By way of example, and not limitation, such computer-readable media may include non-transitory computer-readable storage media including random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), compact disk read-only memory (CD-ROM) or other optical disk storage devices, magnetic disk storage devices or other magnetic storage devices, flash memory devices (e.g., solid-state memory devices).

[0103] The processor 993 is configured to control the computing device and perform processing operations, such as executing code stored in the memory to implement a variety of different functions such as running complex algorithms using data from one or more UEs and / or determining resources to be used for transmitting data to and from those UEs, as described herein.

[0104] The memory 994 may store data read and written by the processor 993, for example, data from an AI application running on the UE and / or data corresponding to a network configuration. As referred to herein, a processor may include one or more general-purpose processing devices, such as a microprocessor, a central processing unit, etc. A processor may include a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, or a processor implementing other instruction sets or a combination of instruction sets. A processor may also include one or more special-purpose processing devices, such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), a network processor, etc. In one or more embodiments, the processor is configured to execute instructions to perform the operations and steps discussed herein.

[0105] The network interface (network I / F) 997 may be connected to a network such as a DN as part of the Internet, and may be connected to other computing devices via the network. The network I / F 997 may control data input / output with other devices via the network.

[0106] The (AI) algorithmic applications and / or decision applications shown in the drawings in the DN may be a processor 993 (or a plurality thereof) executing processing instructions (programs) stored in memory 994 and exchanging data with a base station via a network I / F 997. In particular, the processor 993 may execute processing instructions to receive data and / or possible network configurations from a base station via the network I / F, determine at least one preferred network configuration from the possible network configurations, and report the preferred network configuration to the base station for use by an application on the UE.

[0107] Methods embodying aspects of the present invention may be implemented on a computing device such as that illustrated in Figure 19. Such a computing device need not have all of the components illustrated in Figure 19, but may be comprised of a subset of those components. Methods embodying aspects of the present invention may be implemented by a single computing device in communication with one or more data storage servers over a network. The computing device may be the data storage device itself, which stores and processes data, as well as a preferred selected network configuration.

[0108] Methods embodying aspects of the present invention may be performed by a plurality of computing devices operating in cooperation with one another, one or more of which may be data storage servers that store at least a portion of the preferred selected network configuration.

[0109] summary

[0110] The methods herein relate to the transfer of initial application level parameters for control of AL / ML or other algorithms executed on or closely connected to the base station.

[0111] The methodology introduces procedures and signaling into wireless networks to help optimize control of inference-based data filtering using AI / ML. In a preferred arrangement, base station (non)real-time control settings are used to optimize performance of AI / ML or other applications.

[0112] One main set of problems addressed is the shortcomings and limitations of current control of inference-based data filtering using AI / ML, namely, that within an architecture such as O-RAN, the data flow in applications using AI / ML cannot be rapidly processed since the information is not transmitted as in the present method using air interface information from the real-time control unit of the base station.

[0113] An advantage of the arrangements herein is improved processing such as initial inference / filtering, which may be achieved, inter alia, by fast forwarding of application layer information to embedded applications in the base station or DN. The arrangements herein detail initial signaling from the application to the base station in the context of an initial data connection. These arrangements support the use of ML algorithms to enable improved remote inference / filtering for AI / ML applications with respect to specific desired characteristics for a particular local application.

[0114] The present disclosure includes the following inventions. (Appendix 1) 1. A method in a telecommunications system including a User Equipment (UE) on which a local part of an application that generates application data is executed, and a base station on which a remote part of the application is executed, comprising: When the local portion of the application in the UE starts execution, the UE sends a connection message to the base station to initiate a connection with the base station, the connection message including application parameters generated by the local portion of the application; the base station passing the application parameters to the remote portion of the application. (Appendix 2) The method according to claim 1, wherein the application parameters are transmitted in a control channel used to set up a connection with the base station, for example embedded in a Radio Resource Control (RRC) message. (Appendix 3) 3. The method of any one of Supplementary Notes 1-2, further comprising transferring application data generated by the application between the UE and the base station. (Appendix 4) 4. The method of claim 1, wherein the base station modifies execution of the application in response to receiving the application parameters, and / or sends feedback from the remote part of the application to the UE, and / or modifies one or more transmission parameters in response to the application parameters. (Appendix 5) 5. The method of claim 1, wherein the remote part of the application is executed directly on the base station or on an entity in a data network directly connected to the base station, and the local part of the application is executed directly on the UE or on a device directly connected to the UE. (Appendix 6) The method according to any one of appendixes 1 to 5, wherein the telecommunications system is a New Radio (NR) system, and the method further comprises routing application data from the local part of the application to the remote part of the application via a User Plane Function (UPF). (Appendix 7) 7. The method according to any one of appendixes 1 to 6, wherein the base station is an Open Radio Access Network (O-RAN) base station and the application parameters are passed to the remote part of the application or to an algorithm used in the remote part of the application via an O-RAN Central Unit - User Plane (O-CU-UP). (Appendix 8) 8. The method of claim 7, wherein the O-RAN base station passes the application parameters to the application via an E2 interface from the O-CU-UP to the application. (Appendix 9) 9. The method of claim 7 or 8, wherein an algorithm used in the remote part of the application is executed in a Non-RT RIC (on Real Time Radio Access Network Intelligent Controller) with filtering in the O-CU-UP, and data is forwarded to the remote part of the application via UPF. (Appendix 10) A method according to any one of Supplementary Notes 1 to 9, wherein the application is a Machine Learning (ML) application and the application parameters are Artificial Intelligence (AI) indicators used in the inference and / or filtering stages of an ML algorithm, the AI ​​indicators preferably controlling filtering of application data transmitted between the UE and the base station. (Appendix 11) A telecommunications system including a UE (User Equipment) and a base station, a local portion of an application that generates application data executes in a UE and a remote portion of the application executes in the base station; When the local portion of the application starts to execute, the UE is configured to send a connection message to the base station to initiate a connection with the base station, the connection message including application parameters generated by the local portion of the application; The base station is configured to pass the application parameters to the remote part of the application. (Appendix 12) A User Equipment (UE) on which a local part of an application for telecommunications and generating application data is executed, The UE is configured to send a connection message to a base station for electrical communication when the local portion of the application begins execution to initiate a connection with the base station, the connection message including application parameters generated by the local portion of the application for use in the remote portion of the application. (Appendix 13) A base station for telecommunications and in which a remote part of an application is executed, The base station is configured to receive a connection message for electrical communication sent from a UE to initiate a connection with the base station, the connection message including application parameters generated by a local part of the application when the local part of the application starts execution; The base station is configured to pass the application parameters to the remote part of the application. (Appendix 14) A method of telecommunications in a User Equipment (UE) where a local part of an application that generates application data is executed, comprising: A telecommunications method, wherein when the local part of the application starts execution, the UE sends a connection message for telecommunications to the base station to initiate a connection with the base station, the connection message including application parameters generated by the local part of the application. (Appendix 15) A method in a base station where a remote part of an application is executed, comprising the steps of: The base station receives a connection message for electrical communication sent from a UE to initiate a connection with the base station, the connection message including application parameters generated by a local portion of the application when the local portion of the application starts execution; The base station passes the application parameters to the remote part of the application.

Claims

1. 1. A method in a telecommunications system comprising a User Equipment (UE) on which a local part of an application that generates application data is executed, and a base station on which a remote part of said application is executed, comprising: the UE starting execution of the local portion of the application, generating application parameters by the local portion of the application, and sending a connection message to the base station when initiating an initial connection with the base station, the connection message including the application parameters; the base station passing the application parameters to the remote part of the application; The method, wherein the application is a Machine Learning (ML) application, the application parameters are Artificial Intelligence (AI) indicators used in the inference and / or filtering stages of an ML algorithm, and the AI ​​indicators control filtering of application data transmitted between the UE and the base station.

2. The method of claim 1 , wherein the application parameters are transmitted in a control channel used to set up an initial connection with the base station, e.g., embedded in a Radio Resource Control (RRC) message.

3. The method according to any one of claims 1 to 2, further comprising transferring application data generated by the application between the UE and the base station.

4. 2. The method of claim 1, wherein the base station modifies execution of the application in response to receiving the application parameters, and / or sends feedback from the remote part of the application to the UE, and / or modifies one or more transmission parameters in response to the application parameters.

5. 2. The method of claim 1, wherein the remote portion of the application is executed directly on the base station or on an entity in a data network directly connected to the base station, and the local portion of the application is executed directly on the UE or on a device directly connected to the UE.

6. 2. The method of claim 1, wherein the telecommunications system is a New Radio (NR) system, and the method further comprises routing application data from the local portion of the application to a remote portion of the application via a User Plane Function (UPF).

7. 2. The method of claim 1, wherein the base station is an Open Radio Access Network (O-RAN) base station and the application parameters are passed to the remote part of the application or to an algorithm used in the remote part of the application via an O-RAN Central Unit - User Plane (O-CU-UP).

8. The method of claim 7, wherein the O-RAN base station passes the application parameters to the application via an E2 interface from the O-CU-UP to the application.

9. 9. The method according to claim 7 or 8, wherein algorithms used in the remote part of the application are executed in a Non-RT on Real Time Radio Access Network Intelligent Controller (RIC) with filtering in the O-CU-UP, and data is forwarded to the remote part of the application via a UPF.

10. A telecommunications system including a UE (User Equipment) and a base station, a local part of an application that generates application data executes in a UE and a remote part of the application executes in the base station; the UE is configured to initiate execution of the local portion of the application, generate application parameters by the local portion of the application, and, upon initiating an initial connection with the base station, send a connection message to the base station, the connection message including the application parameters; the base station is configured to pass the application parameters to the remote part of the application; The application is a Machine Learning (ML) application, the application parameters are Artificial Intelligence (AI) indicators used in the inference and / or filtering stages of an ML algorithm, and the AI ​​indicators control filtering of application data transmitted between the UE and the base station.

11. A UE (User Equipment) on which a local part of an application for telecommunications and generating application data is executed, configured to initiate execution of the local portion of the application, generate application parameters by the local portion of the application, and upon initiating an initial connection with a base station, send a connection message for electrical communication to the base station, the connection message including the application parameters for use in a remote portion of the application; The application is a Machine Learning (ML) application, the application parameters are Artificial Intelligence (AI) indicators used in the inference and / or filtering stages of an ML algorithm, and the AI ​​indicators control filtering of application data transmitted between the UE and the base station. A UE.

12. A base station for telecommunications and in which a remote part of an application is executed, The base station is configured to receive a connection message for telecommunication sent from a UE, the connection message being sent when the UE starts execution of a local part of the application, generates application parameters by the local part of the application, and starts an initial connection with the base station, the connection message including the application parameters; the base station is configured to pass the application parameters to the remote part of the application; The application is a Machine Learning (ML) application, the application parameters are Artificial Intelligence (AI) indicators used in the inference and / or filtering stages of an ML algorithm, and the AI ​​indicators control filtering of application data transmitted between the UE and the base station.

13. A telecommunications method in a User Equipment (UE) in which a local part of an application that generates application data is executed, comprising: initiating execution of the local portion of the application, generating application parameters by the local portion of the application, and upon initiating an initial connection with a base station, sending a connection message for electrical communication to the base station, the connection message including the application parameters for use in a remote portion of the application; The application is a Machine Learning (ML) application, the application parameters are Artificial Intelligence (AI) indicators used in an inference and / or filtering stage of an ML algorithm, and the AI ​​indicators control filtering of application data transmitted between the UE and the base station.

14. A method in a base station where a remote part of an application is executed, comprising the steps of: The base station receives a connection message for telecommunication sent from a UE, the connection message being sent when the UE starts execution of a local portion of the application, generates application parameters by the local portion of the application, and starts an initial connection with the base station, the connection message including the application parameters; the base station passes the application parameters to the remote part of the application; The application is a Machine Learning (ML) application, the application parameters are Artificial Intelligence (AI) indicators used in an inference and / or filtering stage of an ML algorithm, and the AI ​​indicators control filtering of application data transmitted between the UE and the base station.

Images