Method for enhancing AIML application traffic over d2d communication

Abstract

A first wireless transmit / receive unit (WTRU) may trigger a proximity service (ProSe) discovery procedure indicating an application mode of operation and / or ProSe capabilities. The first WTRU may discover an artificial intelligence machine learning (AIML) application assistance service. The ProSe capabilities may include one or more AIML application assistance services associated with at least one of data storage, data processing, and / or data aggregation at one or more peer WTRUs and / or relay WTRUs. The first WTRU may generate a discovery code during the ProSe discovery process based on the AIML application assistance service indicated by the peer WTRU and / or relay WTRU. The first WTRU may select a second WTRU from the peer-to-peer WTRU (s) and / or relay WTRU (s) based on the discovery code. The first WTRU may trigger ProSe direct communication with the selected second WTRU such that the first WTRU can connect to the second WTRU to provide the one or more AIML application assistance services.

Description

Cross - Reference to Related Applications

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 422,310, filed on November 3, 2022, U.S. Provisional Patent Application No. 63 / 452,878, filed on March 17, 2023, and U.S. Provisional Patent Application No. 63 / 583,393, filed on September 18, 2023, the entire contents of which are incorporated herein by reference. Background Art

[0002] The 3GPP 5G system may not be able to support the features imposed by specific service characteristics that an AIML operation running on a 3GPP network may include. Summary of the Invention

[0003] One or more Prose - based AIML application service embodiments are described herein.

[0004] Prose provisioning, discovery, and / or direct communication in support of volunteer operations are described herein.

[0005] A Prose provisioning mechanism is described herein that associates performance metrics with application service characteristics and / or volunteer services (e.g., enabling a WTRU to perform services such as store - and - forward data and / or process - and - forward data and / or aggregate and / or forward data).

[0006] A Prose discovery mechanism is described herein that generates discovery codes based on volunteer services provided by peer WTRUs and / or relay WTRUs.

[0007] A Prose direct communication mechanism is described herein that enables a WTRU operating as a volunteer - consumer to connect to other WTRUs operating as volunteer - providers by using discovery codes associated with AIML service characteristics.

[0008] An associated destination Layer 2 information element is described herein that enables a WTRU to operate as a store - and - forward and / or process - and - forward and / or Layer 3 repeater. A volunteer - consumer WTRU can be a WTRU that uses services from another WTRU (e.g., a volunteer - producer WTRU), which can provide point - to - point connections and / or additional functionality beyond WTRU - to - network relay. For example, this type of functionality can be store - and - forward information and / or process - and - forward information.

[0009] This document may describe one or more extensions to a Quality of Service (QoS) monitoring process to enable an Application Function (AF) to request QoS monitoring of a (multiple) connection over a PC5 link (e.g., also perform QoS monitoring).

[0010] A first WTRU may trigger a Proximity Service (ProSe) discovery process by indicating an application assistance operation mode and one or more ProSe capabilities. The application assistance operation mode may be associated with one or more AI and / or ML applications. The first WTRU may be configured to discover one or more Artificial Intelligence Machine Learning (AI / ML) application assistance services. One or more of the ProSe capabilities may include one or more AI / ML application assistance services associated with at least one of data storage, data processing, and / or data aggregation at one or more peer WTRUs and / or one or more relay WTRUs. The first WTRU may generate one or more discovery codes. For example, the first WTRU may generate one or more discovery codes based on one or more AI / ML application assistance services indicated by one or more peer WTRUs and / or one or more relay WTRUs during the ProSe discovery process. The first WTRU may select a second WTRU from one or more peer WTRUs and / or one or more relay WTRUs. For example, the first WTRU may select the second WTRU from one or more peer WTRUs and / or one or more relay WTRUs based on one or more discovery codes. The first WTRU may trigger ProSe direct communication with the selected second WTRU such that the first WTRU can connect to the second WTRU to provide one or more AI / ML application assistance services.

[0011] Triggering the ProSe discovery process may include a WTRU configured to send a solicitation message to one or more peer WTRUs and / or one or more relay WTRUs. The WTRU may use one or more discovery codes to send the solicitation message. The solicitation message may request a first ProSe capability among one or more ProSe capabilities. The solicitation message may indicate one or more of the following: a ProSe application identifier, an operation mode, and / or metadata associated with the operation mode.

[0012] The WTRU may receive a response message from one or more of the following: one or more peer WTRUs and / or one or more relay WTRUs. The response message may indicate that the corresponding WTRU is configured to perform the first ProSe capability requested in the solicitation message. The first ProSe capability may include store-and-forward and / or process-and-forward. Selecting another (e.g., second) WTRU may be based on the response message indicating the first ProSe capability.

[0013] Triggering ProSe direct communication may include a first WTRU being configured to send a ProSe direct communication request to a second WTRU. The ProSe direct communication request may indicate an intention to use a first ProSe capability via the second WTRU.

[0014] The first WTRU may send a provision request indicating one or more ProSe capabilities. The provision request may include one or more of the following: assistant service code, user information identifier, operating mode, and / or one or more quality of service (QoS) characteristics. The first WTRU may receive a QoS monitoring request that indicates monitoring of one or more application traffic characteristics associated with an application assistance operating mode and / or one or more ProSe capabilities. The first WTRU may send a QoS monitoring response that indicates one or more QoS monitoring results associated with the one or more application traffic characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1A is a system diagram illustrating an example communication system in which one or more of the disclosed embodiments may be implemented.

[0016] Figure 1B is a system diagram illustrating an example wireless transmit / receive unit (WTRU) that may be used within the Figure 1A illustrated communication system, according to an embodiment.

[0017] Figure 1C is a system diagram illustrating an example radio access network (RAN) and an example core network (CN) that may be used within the Figure 1A illustrated communication system, according to an embodiment.

[0018] Figure 1D is a system diagram illustrating a further example RAN and a further example CN that may be used within the Figure 1A illustrated communication system, according to an embodiment.

[0019] Figure 2 illustrates an example of an artificial intelligence (AI) / machine learning (ML) model download via a 5G system (e.g., model distribution).

[0020] Figure 3 illustrates an example of an AI / ML download from a volunteer WTRU.

[0021] Figure 4 illustrates a flowchart depicting an example process by which a network function (NF) retrieves the computed integrated proximity service (ProSe) capabilities of a given WTRU (e.g., a relay WTRU and / or a remote WTRU).

[0022] Figure 5 The figure illustrates a flow chart that depicts an example process of one or more changes to the computed integrated ProSe capabilities of an NF subscription for a given WTRU (e.g., a relay WTRU and / or a remote WTRU).

[0023] Figure 6 The figure illustrates a flow chart that depicts an example process of an NF retrieving the authorization(s) of a given WTRU for computed integrated ProSe capabilities.

[0024] Figure 7 The figure illustrates a flow chart that depicts an example process of an NF discovering a relay WTRU with authorization(s) for computed integrated ProSe capabilities that can be in the vicinity of a given WTRU.

[0025] Figure 8 The figure illustrates a flow chart that depicts an example ProSe parameter provisioning.

[0026] Figure 9 The figure illustrates a flow chart that depicts an example ProSe discovery request (e.g., Model B).

[0027] Figure 10 The figure illustrates a flow chart that depicts an example ProSe direct communication that supports a volunteer function.

[0028] Figure 11 The figure depicts a flow chart that illustrates an example of quality of service (QoS) monitoring of a PC5 link.

[0029] Figure 12A The figure is a diagram of an example system environment that can implement an artificial intelligence (AI) and / or machine learning (ML) model.

[0030] Figure 12B The figure illustrates an example of a neural network.

[0031] Figure 12C The figure is a diagram of an example system environment for training and / or implementing an AI / ML model that includes a neural network (NN). Specific implementation

[0032] Figure 1AFIG. is a diagram illustrating an example communication system 100 in which one or more of the disclosed embodiments may be implemented. The communication system 100 may enable multiple wireless users to access such content by sharing system resources, including wireless bandwidth. For example, the communication system 100 may employ one or more channel access methods such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), single carrier FDMA (SC-FDMA), zero-tail unique word DFT-spread OFDM (ZT UW DTS-s OFDM), unique word OFDM (UW-OFDM), resource block filtered OFDM, filter bank multicarrier (FBMC), etc.

[0033] As Figure 1A shown, the communication system 100 may include wireless transmit / receive units (WTRUs) 102a, 102b, 102c, 102d, RAN 104 / 113, CN 106 / 115, public switched telephone network (PSTN) 108, Internet 110, and other networks 112, although it will be appreciated that the disclosed embodiments contemplate any number of WTRUs, base stations, networks, and / or network elements. Each of the WTRUs 102a, 102b, 102c, 102d may be any type of device configured to operate and / or communicate in a wireless environment. By way of example, the WTRUs 102a, 102b, 102c, 102d (any of which may be referred to as a “station” and / or “STA”) may be configured to transmit and / or receive wireless signals and may include user equipment (UE), mobile stations, fixed or mobile subscriber units, subscription-based units, pagers, cellular telephones, personal digital assistants (PDA), smart phones, laptop computers, netbooks, personal computers, wireless sensors, hotspots or Mi-Fi devices, Internet of Things (IoT) devices, watches or other wearable devices, head-mounted displays (HMD), vehicles, drones, medical devices and applications (e.g., remote surgery), industrial devices and applications (e.g., robots and / or other wireless devices operating in industrial and / or automation processing chain scenarios), consumer electronic devices, devices operating on commercial and / or industrial wireless networks, etc. Any of the WTRUs 102a, 102b, 102c, 102d may be interchangeably referred to as a UE.

[0034] The communication system 100 may further include base station 114a and / or base station 114b. Each of base stations 114a, 114b may be any type of device configured to wirelessly interface with at least one of WTRUs 102a, 102b, 102c, 102d to facilitate access to one or more communication networks such as CN 106 / 115, the Internet 110, and / or other networks 112. By way of example, base stations 114a, 114b may be base transceiver stations (BTSs), Node-Bs, eNode Bs, home Node Bs, home eNode Bs, gNBs, NR NodeBs, site controllers, access points (APs), wireless routers, etc. Although base stations 114a, 114b are depicted as single elements, it will be appreciated that base stations 114a, 114b may include any number of interconnected base stations and / or network elements.

[0035] Base station 114a may be part of RAN 104 / 113, which may also include other base stations and / or network elements (not shown) such as base station controllers (BSCs), radio network controllers (RNCs), relay nodes, etc. Base station 114a and / or base station 114b may be configured to transmit and / or receive wireless signals on one or more carrier frequencies, which may be referred to as cells (not shown). These frequencies may be in licensed spectrum, unlicensed spectrum, or a combination of licensed and unlicensed spectrum. A cell may provide coverage of a particular geographical area for wireless services, which may be relatively fixed or may change over time. A cell may be further divided into cell sectors. For example, the cell associated with base station 114a may be divided into three sectors. Thus, in one embodiment, base station 114a may include three transceivers, i.e., one transceiver for each sector of the cell. In one embodiment, base station 114a may employ multiple-input multiple-output (MIMO) technology and may use multiple transceivers for each sector of the cell. For example, beamforming may be used to transmit and / or receive signals in a desired spatial direction.

[0036] Base stations 114a, 114b may communicate with one or more of WTRUs 102a, 102b, 102c, 102d via air interface 116, which may be any suitable wireless communication link (e.g., radio frequency (RF), microwave, millimeter wave, infrared (IR), ultraviolet (UV), visible light, etc.). Air interface 116 may be established using any suitable radio access technology (RAT).

[0037] More specifically, as noted above, the communication system 100 can be a multi-access system and can employ one or more channel access schemes, such as, CDMA, TDMA, FDMA, OFDMA, SC-FDMA, etc. For example, the base stations 114a in the RAN 104 / 113, and the WTRUs 102a, 102b, 102c can implement radio technologies such as Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access (UTRA), which can use Wideband CDMA (WCDMA) to establish the air interfaces 115 / 116 / 117. WCDMA can include communication protocols such as High Speed Packet Access (HSPA) and / or Evolved HSPA (HSPA+). HSPA can include High Speed Downlink (DL) Packet Access (HSDPA) and / or High Speed UL Packet Access (HSUPA).

[0038] In one embodiment, the base station 114a and the WTRUs 102a, 102b, 102c can implement radio technologies such as Evolved UMTS Terrestrial Radio Access (E-UTRA), which can use Long Term Evolution (LTE) and / or Advanced LTE (LTE-A) and / or Advanced LTE Pro (LTE-A Pro) to establish the air interface 116.

[0039] In one embodiment, the base station 114a and the WTRUs 102a, 102b, 102c can implement radio technologies such as NR radio access, which can use New Radio (NR) to establish the air interface 116.

[0040] In one embodiment, the base station 114a and the WTRUs 102a, 102b, 102c can implement multiple radio access technologies. For example, the base station 114a and the WTRUs 102a, 102b, 102c can implement LTE radio access and NR radio access together, for example, using the Dual Connectivity (DC) principle. Thus, the air interfaces utilized by the WTRUs 102a, 102b, 102c can be characterized by multiple types of radio access technologies and / or transmissions sent to / from multiple types of base stations (e.g., eNBs and gNBs).

[0041] In other embodiments, the base station 114a and the WTRUs 102a, 102b, 102c may implement radio technologies such as IEEE 802.11 (i.e., Wireless Fidelity (WiFi)), IEEE 802.16 (i.e., Worldwide Interoperability for Microwave Access (WiMAX)), CDMA2000, CDMA2000 1X, CDMA2000 EV-DO, Interim Standard 2000 (IS-2000), Interim Standard 95 (IS-95), Interim Standard 856 (IS-856), Global System for Mobile Communications (GSM), Enhanced Data Rate for GSM Evolution (EDGE), GSM EDGE (GERAN), etc.

[0042] Figure 1A The base station 114b in Figure 1A may be, for example, a wireless router, a home node B, a home eNode B, or an access point, and may utilize any suitable RAT to facilitate wireless connection in a local area such as a commercial venue, a home, a vehicle, a campus, an industrial facility, an air corridor (e.g., for drones), a road, etc. In one embodiment, the base station 114b and the WTRUs 102c, 102d may implement radio technologies such as IEEE 802.11 to establish a Wireless Local Area Network (WLAN). In one embodiment, the base station 114b and the WTRUs 102c, 102d may implement radio technologies such as IEEE 802.15 to establish a Wireless Personal Area Network (WPAN). In yet another embodiment, the base station 114b and the WTRUs 102c, 102d may utilize a cellular-based RAT (e.g., WCDMA, CDMA2000, GSM, LTE, LTE-A, LTE-A Pro, NR, etc.) to establish a pico cell or a femto cell. As Figure 1A shown, the base station 114b may have a direct connection to the Internet 110. Thus, the base station 114b may not need to access the Internet 110 via the CN 106 / 115.

[0043] The RAN 104 / 113 may communicate with the CN 106 / 115, which may be any type of network configured to provide voice, data, applications, and / or Voice over Internet Protocol (VoIP) services to one or more of the WTRUs 102a, 102b, 102c, 102d. The data may have different Quality of Service (QoS) requirements such as different throughput requirements, latency requirements, error tolerance requirements, reliability requirements, data throughput requirements, mobility requirements, etc. The CN 106 / 115 may provide call control, billing services, location-based services, prepaid calls, Internet connectivity, video distribution, etc. and / or perform advanced security functions such as user authentication. Although Figure 1AAlthough not shown in the figure, it will be appreciated that RAN 104 / 113 and / or CN 106 / 115 may communicate directly or indirectly with other RANs that employ the same RAT or a different RAT as RAN 104 / 113. For example, in addition to being connected to RAN 104 / 113 that may utilize NR radio technology, CN 106 / 115 may also communicate with another RAN (not shown) that employs GSM, UMTS, CDMA 2000, WiMAX, E-UTRA, or WiFi radio technology.

[0044] CN 106 / 115 may also act as a gateway for the WTRUs 102a, 102b, 102c, 102d to access the PSTN 108, the Internet 110, and / or other networks 112. The PSTN 108 may include a circuit-switched telephone network that provides plain old telephone service (POTS). The Internet 110 may include a global system of interconnected computer networks and devices that use common communication protocols such as the Transmission Control Protocol (TCP), the User Datagram Protocol (UDP), and / or the Internet Protocol (IP) in the TCP / IP Internet protocol suite. The network 112 may include wired and / or wireless communication networks owned and / or operated by other service providers. For example, the network 112 may include another CN that is connected to one or more RANs, and the RAN may employ the same RAT or a different RAT as RAN 104 / 113.

[0045] Some or all of the WTRUs 102a, 102b, 102c, 102d in the communication system 100 may include multi-mode capabilities (e.g., the WTRUs 102a, 102b, 102c, 102d may include multiple transceivers for communicating with different wireless networks over different wireless links). For example, Figure 1A the illustrated WTRU 102c may be configured to communicate with a base station 114a that may employ a cellular-based radio technology and a base station 114b that may employ IEEE 802 radio technology.

[0046] Figure 1B is a system diagram illustrating an example WTRU 102. As Figure 1B shown, the WTRU 102 may include a processor 118, a transceiver 120, a transmit / receive element 122, a speaker / microphone 124, a keyboard 126, a display / touchpad 128, a non-removable memory 130, a removable memory 132, a power supply 134, a Global Positioning System (GPS) chipset 136, and / or other peripheral devices 138, etc. It will be appreciated that the WTRU 102 may include any sub-combination of the above elements while remaining consistent with the embodiments.

[0047] The processor 118 can be a general-purpose processor, a dedicated processor, a conventional processor, a digital signal processor (DSP), multiple microprocessors, one or more microprocessors associated with a DSP core, a controller, a microcontroller, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) circuit, any other type of integrated circuit (IC), a state machine, etc. The processor 118 can perform signal encoding, data processing, power control, input / output processing, and / or any other functions that enable the WTRU 102 to operate in a wireless environment. The processor 118 can be coupled to the transceiver 120, which can be coupled to the transmit / receive element 122. Although Figure 1B the processor 118 and the transceiver 120 are depicted as separate components, it will be appreciated that the processor 118 and the transceiver 120 can be integrated together in an electronic package or chip.

[0048] The transmit / receive element 122 can be configured to transmit signals to a base station (e.g., base station 114a) or receive signals from a base station via the air interface 116. For example, in one embodiment, the transmit / receive element 122 can be an antenna configured to transmit and / or receive RF signals. In one embodiment, the transmit / receive element 122 can be a transmitter / detector configured to transmit and / or receive, for example, IR, UV, or visible light signals. In yet another embodiment, the transmit / receive element 122 can be configured to transmit and / or receive both RF and optical signals. It will be appreciated that the transmit / receive element 122 can be configured to transmit and / or receive any combination of wireless signals.

[0049] Although the transmit / receive element 122 is depicted as a single element in Figure 1B the WTRU 102 can include any number of transmit / receive elements 122. More specifically, the WTRU 102 can employ MIMO technology. Thus, in one embodiment, the WTRU 102 can include two or more transmit / receive elements 122 (e.g., multiple antennas) for transmitting and receiving wireless signals via the air interface 116.

[0050] The transceiver 120 can be configured to modulate the signals to be transmitted by the transmit / receive element 122 and demodulate the signals received by the transmit / receive element 122. As noted above, the WTRU 102 can have multi-mode capabilities. Thus, for example, the transceiver 120 can include multiple transceivers for enabling the WTRU 102 to communicate via, for example, multiple RATs such as NR and IEEE 802.11.

[0051] The processor 118 of the WTRU 102 can be coupled to a speaker / microphone 124, a keyboard 126, and / or a display / touchpad 128 (e.g., a liquid crystal display (LCD) display unit or an organic light emitting diode (OLED) display unit), and can receive user input data from them. The processor 118 can also output user data to the speaker / microphone 124, the keyboard 126, and / or the display / touchpad 128. Additionally, the processor 118 can access information from any type of suitable memory (such as non-removable memory 130 and / or removable memory 132), and store data in that memory. The non-removable memory 130 can include random access memory (RAM), read-only memory (ROM), a hard disk, or any other type of memory storage device. The removable memory 132 can include a subscriber identity module (SIM) card, a memory stick, a secure digital (SD) memory card, etc. In other embodiments, the processor 118 can access information from a memory that is not physically located on the WTRU 102 (such as on a server or a home computer (not shown)), and store data in that memory.

[0052] The processor 118 can receive power from a power supply 134, and can be configured to distribute and / or control the power to other components in the WTRU 102. The power supply 134 can be any suitable device for powering the WTRU 102. For example, the power supply 134 can include one or more dry cell batteries (e.g., nickel cadmium (NiCd), nickel zinc (NiZn), nickel metal hydride (NiMH), lithium ion (Li-ion), etc.), solar cells, fuel cells, etc.

[0053] The processor 118 can also be coupled to a GPS chipset 136, which can be configured to provide location information (e.g., longitude and latitude) about the current location of the WTRU 102. In addition to or instead of the information from the GPS chipset 136, the WTRU 102 can receive location information from a base station (e.g., base stations 114a, 114b) via an air interface 116 and / or determine its location based on the timing of signals received from two or more nearby base stations. It will be appreciated that the WTRU 102 can obtain location information by any suitable location determination method while remaining consistent with the embodiments.

[0054] The processor 118 may be further coupled to other peripheral devices 138, which may include one or more software and / or hardware modules that provide additional features, functionality, and / or wired or wireless connections. For example, the peripheral devices 138 may include an accelerometer, an electronic compass, a satellite transceiver, a digital camera (for photos and / or videos), a universal serial bus (USB) port, a vibration device, a television transceiver, a hands-free headset, modules, a frequency modulation (FM) radio unit, a digital music player, a media player, a video game player module, an Internet browser, a virtual reality and / or augmented reality (VR / AR) device, an activity tracker, etc. The peripheral devices 138 may include one or more sensors, which may be one or more of the following: a gyroscope, an accelerometer, a Hall effect sensor, a magnetometer, an orientation sensor, a proximity sensor, a temperature sensor, a time sensor, a geographical location sensor, an altimeter, a light sensor, a touch sensor, a magnetometer, a barometer, a gesture sensor, a biometric sensor, and / or a humidity sensor.

[0055] The WTRU 102 may include a full-duplex radio, for which the transmission and reception of some or all signals (e.g., associated with a particular subframe for both UL (e.g., for transmission) and downlink (e.g., for reception)) may be concurrent and / or simultaneous. The full-duplex radio may include an interference management unit 139 to reduce and / or substantially eliminate self-interference via signal processing performed by hardware (e.g., a choke) or via a processor (e.g., a separate processor (not shown) or via the processor 118). In one embodiment, the WTRU 102 may include a half-duplex radio, for which the transmission and reception of some or all signals (e.g., associated with a particular subframe for UL (e.g., for transmission) or downlink (e.g., for reception)).

[0056] Figure 1C is a system diagram illustrating a RAN 104 and a CN 106 according to one embodiment. As noted above, the RAN 104 employs E-UTRA radio technology to communicate with the WTRU 102a, 102b, 102c via the air interface 116. The RAN 104 may also communicate with the CN 106.

[0057] The RAN 104 may include eNode-Bs 160a, 160b, 160c, although it will be appreciated that the RAN 104 may include any number of eNode-Bs while remaining consistent with the embodiments. The eNode-Bs 160a, 160b, 160c may each include one or more transceivers for communicating with the WTRUs 102a, 102b, 102c over the air interface 116. In one embodiment, the eNode-Bs 160a, 160b, 160c may implement MIMO technology. Thus, the eNode-B 160a, for example, may use multiple antennas to transmit wireless signals to and / or receive wireless signals from the WTRU 102a.

[0058] Each of the eNode-Bs 160a, 160b, 160c may be associated with a particular cell (not shown) and may be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and / or DL, etc. As Figure 1C shown, the eNode-Bs 160a, 160b, 160c may communicate with each other via the X2 interface.

[0059] Figure 1C The CN 106 shown may include a Mobility Management Entity (MME) 162, a Serving Gateway (SGW) 164, and a Packet Data Network (PDN) Gateway (or PGW) 166. Although each of the foregoing elements is depicted as part of the CN 106, it will be appreciated that any of these elements may be owned and / or operated by an entity other than the CN operator.

[0060] The MME 162 may be connected to each of the eNode-Bs 160a, 160b, 160c in the RAN 104 via the S1 interface and may act as a control node. For example, the MME 162 may be responsible for authenticating users of the WTRUs 102a, 102b, 102c, bearer activation / deactivation, selecting a particular serving gateway during the initial attachment of the WTRUs 102a, 102b, 102c, etc. The MME 162 may provide control plane functions for handover between the RAN 104 and other RANs (not shown) employing other radio technologies such as GSM and / or WCDMA.

[0061] The SGW 164 can be connected to each of the eNode Bs 160a, 160b, 160c in the RAN 104 via the S1 interface. The SGW 164 can generally route and forward user data packets to / from the WTRUs 102a, 102b, 102c. The SGW 164 can perform other functions such as anchoring the user plane during handovers between eNode Bs, triggering paging when DL data is available for the WTRUs 102a, 102b, 102c, managing and storing the context of the WTRUs 102a, 102b, 102c, etc.

[0062] The SGW 164 can be connected to the PGW 166, which can provide the WTRUs 102a, 102b, 102c with access to a packet switched network (such as the Internet 110) to facilitate communication between the WTRUs 102a, 102b, 102c and IP-enabled devices.

[0063] The CN 106 can facilitate communication with other networks. For example, the CN 106 can provide the WTRUs 102a, 102b, 102c with access to a circuit switched network (such as the PSTN 108) to facilitate communication between the WTRUs 102a, 102b, 102c and traditional landline communication devices. For example, the CN 106 can include an IP gateway (e.g., an IP Multimedia Subsystem (IMS) server) or can communicate with the IP gateway, which serves as an interface between the CN 106 and the PSTN 108. Additionally, the CN 106 can provide the WTRUs 102a, 102b, 102c with access to other networks 112, which can include other wired and / or wireless networks owned and / or operated by other service providers.

[0064] Although the WTRU is described as a wireless terminal in Figures 1A to 1D it is envisioned that in some representative embodiments, such a terminal can use (e.g., temporarily or permanently) a wired communication interface to the communication network.

[0065] In a representative embodiment, the other network 112 can be a WLAN.

[0066] In an infrastructure basic service set (BSS) mode, a WLAN can have an access point (AP) for the BSS and one or more stations (STAs) associated with the AP. The AP can have access or an interface to a distribution system (DS) or another type of wired / wireless network that carries traffic into and / or out of the BSS. Traffic destined for an STA from outside the BSS can reach the STA through the AP and can be delivered to the STA. Traffic from an STA to a destination outside the BSS can be sent to the AP for delivery to the corresponding destination. Traffic between STAs within the BSS can be sent through the AP. For example, a source STA can send traffic to the AP, and the AP can deliver the traffic to the destination STA. Traffic between STAs within the BSS can be considered and / or referred to as peer traffic. Peer traffic can be sent between the source and destination STAs (e.g., directly between them) using direct link setup (DLS). In some representative embodiments, DLS can use 802.11e DLS or 802.11z tunnel DLS (TDLS). A WLAN using an independent BSS (IBSS) mode may not have an AP, and STAs within the IBSS or using the IBSS (e.g., all STAs) can communicate directly with each other. The IBSS communication mode can sometimes be referred to in this document as an "ad-hoc" communication mode.

[0067] When using an 802.11ac infrastructure operation mode or a similar operation mode, the AP can transmit beacons on a fixed channel (such as the primary channel). The primary channel can be of a fixed width (e.g., a 20 MHz wide bandwidth) or a width dynamically set via signaling. The primary channel can be the operating channel of the BSS and can be used by STAs to establish a connection with the AP. In some representative embodiments, carrier sense multiple access with collision avoidance (CSMA / CA) can be implemented in, for example, an 802.11 system. For CSMA / CA, STAs including the AP (e.g., each STA) can sense the primary channel. If the primary channel is sensed / detected and / or determined to be busy by a particular STA, the particular STA can back off. Only one STA (e.g., only one station) can transmit in a given BSS at any given time.

[0068] High throughput (HT) STAs can communicate using a 40 MHz wide channel, for example, via a combination of a primary 20 MHz channel and an adjacent or non-adjacent 20 MHz channel to form a 40 MHz wide channel.

[0069] A very high throughput (VHT) STA can support channels that are 20 MHz, 40 MHz, 80 MHz, and / or 160 MHz wide. The 40 MHz and / or 80 MHz channels can be formed by combining contiguous 20 MHz channels. The 160 MHz channel can be formed by combining eight contiguous 20 MHz channels or by combining two non - contiguous 80 MHz channels (which can be referred to as an 80+80 configuration). For the 80+80 configuration, after channel coding, the data can be passed through a segment parser that can divide the data into two streams. The inverse fast Fourier transform (IFFT) processing and time - domain processing can be done separately on each stream. The streams can be mapped to two 80 MHz channels, and the data can be transmitted by the STA that is performing the transmission. At the receiver of the STA that is performing the reception, the above operations for the 80+80 configuration can be reversed, and the combined data can be sent to the media access control (MAC).

[0070] The operating modes below 1 GHz are supported by 802.11af and 802.11ah. The channel operating bandwidth and carriers are reduced in 802.11af and 802.11ah compared to those used in 802.11n and 802.11ac. 802.11af supports 5 MHz bandwidth, 10 MHz bandwidth, and 20 MHz bandwidth in the TV white space (TVWS) spectrum, and 802.11ah supports 1 MHz bandwidth, 2 MHz bandwidth, 4 MHz bandwidth, 8 MHz bandwidth, and 16 MHz bandwidth using non - TVWS spectrum. According to a representative embodiment, 802.11ah can support meter - type control / machine - type communication, such as MTC devices in a macro - coverage area. The MTC devices can have certain capabilities, for example, limited capabilities, including supporting (e.g., only supporting) certain and / or limited bandwidths. The MTC devices can include a battery whose battery life is above a threshold (e.g., to maintain a very long battery life).

[0071] A WLAN system that can support multiple channels and channel bandwidths (such as 802.11n, 802.11ac, 802.11af, and 802.11ah) includes a channel that can be designated as the primary channel. The primary channel can have a bandwidth equal to the maximum common operating bandwidth supported by all STAs in the BSS. The bandwidth of the primary channel can be set and / or restricted by the STA that supports the minimum bandwidth operation mode among all STAs operating in the BSS. In the example of 802.11ah, for an STA that supports (e.g., only supports) the 1MHz mode (e.g., an MTC type device), the primary channel can be 1MHz wide, even if the AP and other STAs in the BSS support 2MHz, 4MHz, 8MHz, 16MHz, and / or other channel bandwidth operation modes. Carrier sensing and / or network allocation vector (NAV) settings can depend on the state of the primary channel. If the primary channel is busy, for example, due to an STA (which only supports the 1MHz operation mode) transmitting to the AP, the entire available frequency band may be considered busy, even if most of the frequency band remains idle and available.

[0072] In the United States, the available frequency band that 802.11ah can use is from 902MHz to 928MHz. In Korea, the available frequency band is from 917.5MHz to 923.5MHz. In Japan, the available frequency band is from 916.5MHz to 927.5MHz. Depending on the country code, the total bandwidth available for 802.11ah is 6MHz to 26MHz.

[0073] Figure 1D FIG. is a system diagram illustrating RAN 113 and CN 115 according to an embodiment. As noted above, RAN113 can employ NR radio technology to communicate with WTRUs 102a, 102b, 102c via air interface 116. RAN 113 can also communicate with CN 115.

[0074] The RAN 113 may include gNBs 180a, 180b, 180c, although it will be appreciated that the RAN 113 may include any number of gNBs while remaining consistent with the embodiments. The gNBs 180a, 180b, 180c may each include one or more transceivers for communicating with the WTRUs 102a, 102b, 102c over the air interface 116. In one embodiment, the gNBs 180a, 180b, 180c may implement MIMO technology. For example, the gNBs 180a, 180b may utilize beamforming to transmit signals to and / or receive signals from the gNBs 180a, 180b, 180c. Thus, the gNB 180a, for example, may use multiple antennas to transmit wireless signals to the WTRU 102a and / or receive wireless signals from the WTRU 102a. In one embodiment, the gNBs 180a, 180b, 180c may implement carrier aggregation technology. For example, the gNB 180a may transmit multiple component carriers to the WTRU 102a (not shown). A subset of these component carriers may be on unlicensed spectrum, while the remaining component carriers may be on licensed spectrum. In one embodiment, the gNBs 180a, 180b, 180c may implement coordinated multi-point (CoMP) technology. For example, the WTRU 102a may receive a coordinated transmission from the gNB 180a and the gNB 180b (and / or gNB 180c).

[0075] The WTRUs 102a, 102b, 102c may use transmissions associated with scalable numerology. For example, the OFDM symbol interval and / or the OFDM subcarrier interval may vary for different transmissions, different cells, and / or different portions of the radio transmission spectrum. The WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c using various or scalable length subframes or transmission time intervals (TTIs) (e.g., containing different numbers of OFDM symbols and / or lasting for different lengths of absolute time).

[0076] gNBs 180a, 180b, 180c may be configured to communicate with WTRUs 102a, 102b, 102c in stand-alone configuration and / or non-stand-alone configuration. In stand-alone configuration, WTRUs 102a, 102b, 102c may communicate with gNBs 180a, 180b, 180c without accessing another RAN (e.g., such as eNode-Bs 160a, 160b, 160c). In stand-alone configuration, WTRUs 102a, 102b, 102c may utilize one or more of gNBs 180a, 180b, 180c as a mobility anchor. In stand-alone configuration, WTRUs 102a, 102b, 102c may communicate with gNBs 180a, 180b, 180c using signals in the unlicensed band. In non-stand-alone configuration, WTRUs 102a, 102b, 102c may communicate with / connect to gNBs 180a, 180b, 180c while also communicating with / connecting to another RAN (such as eNode-Bs 160a, 160b, 160c). For example, WTRUs 102a, 102b, 102c may implement the DC principle to communicate with one or more gNBs 180a, 180b, 180c and one or more eNode-Bs 160a, 160b, 160c substantially simultaneously. In non-stand-alone configuration, eNode-Bs 160a, 160b, 160c may act as the mobility anchor for WTRUs 102a, 102b, 102c, and gNBs 180a, 180b, 180c may provide additional coverage and / or throughput to serve WTRUs 102a, 102b, 102c.

[0077] Each of gNBs 180a, 180b, 180c may be associated with a specific cell (not shown) and may be configured to handle radio resource management decisions, handover decisions, scheduling of users in UL and / or DL, support for network slicing, dual connectivity, networking between NR and E-UTRA, routing of user plane data towards user plane functions (UPFs) 184a, 184b, routing of control plane information towards access and mobility management functions (AMFs) 182a, 182b, etc. As Figure 1D shown, gNBs 180a, 180b, 180c may communicate with each other via the Xn interface.

[0078] Figure 1DThe illustrated CN 115 may include at least one AMF 182a, 182b, at least one UPF 184a, 184b, at least one Session Management Function (SMF) 183a, 183b, and possibly Data Networks (DN) 185a, 185b. Although each of the foregoing elements is depicted as part of the CN 115, it will be appreciated that any of these elements may be owned and / or operated by an entity other than the CN operator.

[0079] The AMF 182a, 182b may be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 113 via the N2 interface and may act as a control node. For example, the AMF 182a, 182b may be responsible for authenticating users of the WTRUs 102a, 102b, 102c, supporting network slicing (e.g., handling different PDU sessions with different requirements), selecting a particular SMF 183a, 183b, managing the registration area, terminating NAS signaling, mobility management, etc. Network slicing may be used by the AMF 182a, 182b to customize the CN support for the WTRUs 102a, 102b, 102c based on the type of service utilized by the WTRUs 102a, 102b, 102c. For example, different network slices may be established for different use cases, such as services relying on Ultra-Reliable Low Latency (URLLC) access, services relying on Enhanced Massive Mobile Broadband (eMBB) access, services for Machine-Type Communication (MTC) access, etc. The AMF 162 may provide control plane functions for handovers between the RAN 113 and other RANs (not shown) employing other radio technologies such as LTE, LTE-A, LTE-A Pro, and / or non-3GPP access technologies such as WiFi.

[0080] The SMF 183a, 183b may be connected to the AMF 182a, 182b in the CN 115 via the N11 interface. The SMF 183a, 183b may also be connected to the UPF 184a, 184b in the CN 115 via the N4 interface. The SMF 183a, 183b may select and control the UPF 184a, 184b and configure the traffic routing through the UPF 184a, 182b. The SMF 183a, 183b may perform other functions such as managing and allocating UE IP addresses, managing PDU sessions, controlling policy enforcement and QoS, providing downlink data notifications, etc. The PDU session type may be IP-based, non-IP-based, Ethernet-based, etc.

[0081] UPF 184a and 184b can be connected to one or more of gNBs 180a, 180b, 180c in the RAN 113 via the N3 interface, and these gNBs can provide access to a packet switched network (such as the Internet 110) to the WTRUs 102a, 102b, 102c to facilitate communication between the WTRUs 102a, 102b, 102c and IP-enabled devices. UPF 184a, 184b can perform other functions, such as routing and forwarding packets, implementing user plane policies, supporting multi-homed PDU sessions, handling user plane QoS, buffering downlink packets, providing mobility anchoring, etc.

[0082] The CN 115 can facilitate communication with other networks. For example, the CN 115 can include an IP gateway (e.g., an IP Multimedia Subsystem (IMS) server) or can communicate with the IP gateway, which serves as an interface between the CN 115 and the PSTN 108. Additionally, the CN 115 can provide access to other networks 112 to the WTRUs 102a, 102b, 102c, and the other networks 112 can include other wired and / or wireless networks owned and / or operated by other service providers. In one embodiment, the WTRUs 102a, 102b, 102c can be connected to local data networks (DNs) 185a, 185b via the N3 interface to the UPF 184a, 184b and the N6 interface between the UPF 184a, 184b and the DNs 185a, 185b through the UPF 184a, 184b.

[0083] In view of Figures 1A to 1D and Figures 1A to 1D In view of the corresponding descriptions of, one or more or all of the functions described in the text regarding one or more of the following can be performed by one or more emulation devices (not shown): WTRUs 102a to 102d, base stations 114a to 114b, eNode-Bs 160a to 160c, MME 162, SGW 164, PGW 166, gNBs 180a to 180c, AMFs 182a to 182b, UPFs 184a to 184b, SMFs 183a to 183b, DNs 185a to 185b, and / or any other device(s) described herein. The emulation device(s) can be one or more devices configured to emulate one or more or all of the functions described herein. For example, the emulation device(s) can be used to test other devices and / or simulate network and / or WTRU functions.

[0084] Emulation devices can be designed to implement one or more tests of other devices in a laboratory environment and / or in an operator network environment. For example, one or more emulation devices can perform one or more or all functions while being fully or partially implemented and / or deployed as part of a wired and / or wireless communication network in order to test other devices within the communication network. One or more emulation devices can perform one or more or all functions while being temporarily implemented / deployed as part of a wired and / or wireless communication network. Emulation devices can be directly coupled to another device for testing purposes and / or can perform tests using over-the-air wireless communication.

[0085] One or more emulation devices can perform one or more or all functions while not being implemented / deployed as part of a wired and / or wireless communication network. For example, emulation devices can be used to test test scenarios in a laboratory and / or non-deployed (e.g., test) wired and / or wireless communication networks in order to implement testing of one or more components. One or more emulation devices can be test devices. Direct RF coupling and / or wireless communication via an RF circuit (e.g., which can include one or more antennas) can be used by emulation devices to transmit and / or receive data.

[0086] Embodiments are described herein that enhance AI / ML application services via device-to-device (D2D) communication. For example, embodiments can relate to enhancing (e.g., 5G) proximity-based service (ProSe) communication to define a (e.g., new) assistant WTRU with enhanced ProSe capabilities. The WTRU can indicate the (multiple) capabilities of the WTRU to provide volunteer services as a consumer (e.g., consumer WTRU) and / or a producer (e.g., producer WTRU). The WTRU can indicate one or more performance parameters for requesting radio access network (RAN) resources in a policy container. The WTRU can use a service request procedure to update ProSe preferences regarding AI / ML application operation requirements / characteristics.

[0087] A policy control function (PCF) can generate performance values, e.g., to support AI / ML application services via a PC5 link. The PCF can use access type preferences to indicate whether a volunteer-consumer WTRU should use a PC5 link and / or a Uu link when requesting volunteer services (e.g., store and forward and / or process or forward).

[0088] The Application Function (AF) may request an update to the performance requirements of a specific AI / ML application operation using AF-based service parameter provisioning. For example, a WTRU may send a provisioning request that indicates one or more ProSe capabilities (e.g., including one or more AI / ML application assistance services). The provisioning request may include one or more of the following: assistant service code, user information identifier, operation mode, and / or one or more Quality of Service (QoS) characteristics.

[0089] Embodiments for discovering an AI / ML application assistant are described herein. A WTRU desiring to signal its intent to act as a volunteer-consumer WTRU may use a specific user information (info) identifier (ID) that meets this criterion. A WTRU acting as a WTRU-to-network (WTRU-to-NW) relay may (e.g., also) be able to signal on its WTRU's layer 3 WTRU-to-network relay service that it may support, for example, a "store and forward" service, a "process and forward" service, or both a "store and forward" service and a "process and forward" service. A WTRU desiring to signal its intent to act as a volunteer-producer WTRU may use a specific user information identifier (ID) that meets this criterion. The WTRU may indicate an assistant service code, user InfoID, and / or operation code, and a validity timer associated with the WTRU provider, application ID in the policy container. For example, during a WTRU configuration update service operation, the WTRU may provide an assistant service code, user Info ID, and / or operation code, and a validity timer associated with the WTRU provider, application ID in the policy container.

[0090] Embodiments for Prose discovery to support volunteer operations are described herein. The WTRU may provide a Prose application ID and / or operation mode and / or metadata supporting a specific operation mode. The WTRU may provide the operation mode (e.g., the relevant mode) that the WTRU is looking for and the metadata associated with that operation mode. A volunteer WTRU may receive a solicitation message. The WTRU (e.g., the volunteer WTRU) may use a discovery filter obtained from a Direct Discovery Name Management Function (DDNMF) to match it against the operation mode and / or metadata characteristics / requirements.

[0091] Embodiments for Prose-based connection establishment to support AI / ML operation services are described herein. A WTRU executing as a volunteer-consumer may select a volunteer-producer and / or may determine the applicable destination layer 2 ID and / or assistant service code to use. A WTRU executing as a volunteer-consumer may send a unicast direct communication request, e.g., to signal the WTRU's intent to use "store and forward" and / or "process and forward".

[0092] This document describes embodiments for one or more processes to enable 5GS assistance for AI / ML operations in the application layer for communication over a PC5 link. A WTRU may participate in (multiple) application layer communications such as AI / ML federated learning and / or transfer learning. For example, AI / ML (e.g., 1209) may include one or more algorithms configured for supervised and / or unsupervised learning as described herein. The WTRU may receive a Quality of Service (QoS) monitoring request that indicates monitoring of one or more application traffic characteristics associated with an application assistance operation mode and / or one or more ProSe capabilities. The WTRU may provide one or more Quality of Service (QoS) measurements for the (multiple) QoS monitoring requests from the Core Network (CN). For example, the WTRU may send a QoS monitoring response that indicates one or more QoS monitoring results associated with one or more application traffic characteristics. The WTRU may provide these measurements directly or via an Access Network (AN) to the CN. The WTRU may receive (e.g., from a Session Management Function (SMF), a Policy Control Function (PCF), and / or a Network Exposure Function (NEF)) an Application Identifier (ID) and / or a Group ID and / or a Prose Identifier. If the WTRU receives an Application ID and / or a Group ID and / or a prose identifier from an Application Function (AF), then for example, packet format information may be signaled to monitor which PC5 links and / or which PC5 QoS flows. The WTRU may use the QoS monitoring policy and / or the Prose Identifier and / or the Application ID / Group ID to identify the PC5 QoS flows that may (e.g., may need to) be monitored by the WTRU and / or the eNB, and / or what may (e.g., may need to) be monitored. For example, if the Prose Identifier is associated with a vertical federated learning operation that operates over a PC5 link using a specific Packet Format Information (PFI), the WTRU may monitor whether the allocated Guaranteed Flow Bit Rate (GFBR) and / or one or more other QoS parameters have crossed one or more (e.g., any) thresholds, and if so, report its (multiple) status. For example, the WTRU may report the values of one or more (e.g., relevant) parameters that have crossed the threshold, such as the Packet Error Rate (PER), the Maximum Data Burst Volume (MDBV), and / or the Packet Delay Budget (PDB).

[0093] The SMF may retrieve session management (SM) context from the Unified Data Manager (UDM) using the WTRU-ID and / or export one or more associated packet data unit (PDU) sessions for those WTRUs. The SMF may use the application ID and / or group ID and / or Prose identifier and / or PFI (if provided by the AF) to signal to the WTRU which (if any) PC5 links and / or which PC5 QoS flows may (e.g., may be required to) be monitored. The SMF may use one or more authorized QoS policies provided by the PCF during the policy session procedure to construct a QoS monitoring request.

[0094] The NEF and / or PCF may determine to request QoS monitoring from the access network (e.g., gNB) and / or from the WTRU, e.g., based on one or more operator policies.

[0095] The embodiments described herein may include enhancements for supporting specific service characteristics that may be exploited by AI / ML operations running on a (e.g., 3GPP) network and / or other networks. As described herein, AI / ML operations may be split between AI / ML endpoints (e.g., between a WTRU and an AI / ML application server). AI / ML operations may include AI / ML model / data distribution and / or sharing on a 3GPP system (e.g., 5G system) and / or other networking systems. AI / ML operations may include distributed / federated learning (FL) on a 3GPP system (e.g., 5G system) and / or other networking systems.

[0096] One or more (e.g., some) telecommunication systems (e.g., 3GPP 5G systems) may require scenarios involving AI / ML services operating between one or more (e.g., multiple) WTRUs and application servers (e.g., connected via a network (e.g., 3GPP network or other network)).

[0097] Figure 2 An example of the download of an artificial intelligence (AI) / machine learning (ML) model 202 through a 5G system (e.g., model distribution) 200 is illustrated. The AI / ML model 202 (e.g., as depicted and / or described) may be downloaded through a cellular telecommunication system such as a 5G system (e.g., model distribution). One or more (e.g., some) systems operating according to such a system configuration have been introduced, which may assume that AI / ML services will be carried over a communication path across telecommunication systems (e.g., 5GS systems). Figures 12A to 12C The AI / ML model 202 (e.g., as depicted and / or described) may be downloaded through a cellular telecommunication system such as a 5G system (e.g., model distribution). One or more (e.g., some) systems operating according to such a system configuration have been introduced, which may assume that AI / ML services will be carried over a communication path across telecommunication systems (e.g., 5GS systems).

[0098] One or more embodiments described herein may be collectively referred to as assisting AI / ML operations in the application layer. One or more embodiments described herein may include one or more (e.g., new) monitoring events to estimate the performance of the connection between an AI / ML application function (AI / ML AF) and one or more WTRU parts of an AF session (e.g., session inactivity and / or traffic exchanged between the WTRU and the AI / ML AF). This information may enable (e.g., the AI / ML AF) to run one or more federated learning operations to schedule one or more participating WTRUs to be part of such operations, e.g., based on information provided by the 5GS.

[0099] The 5GS may be able to provide assistance to one or more AI / ML AFs, e.g., through a feature called the member WTRU selection function. The member WTRU selection function may enable the down selection of member WTRUs, which is part of the AI / ML operation but may (e.g., may be required to) meet specific filtering criteria (e.g., such as certain minimum experience quality and / or minimum QoS requirements). The AI / ML AF may (e.g., also) request a time window to transmit one or more large amounts of data (e.g., transmit an ML model) using scheduled data transmission with QoS characteristics, which may enable such data to be transmitted at the least congested time of the 5GS. Additionally or alternatively, the AI / ML AF may request a multi-member AF session with the required QoS for a list of WTRUs identified by WTRU addresses, which may be a way to ensure the operation (e.g., successful operation) of one or more 5GS resources for applying AI / ML operations. One or more (e.g., new) features described herein may enable the consideration of AF sessions running between one or more WTRUs and the AI / ML AF.

[0100] The transmission characteristics inherent to device-to-device communication may contribute to enhancing AI / ML operations (e.g., by providing reduced latency, access to localized data, and / or energy savings). In an example, one or more (e.g., some) WTRUs may attempt to download an AI / ML model in an area where the radio connection is weak (e.g., poor). There may be other WTRUs that have a stronger connection and / or stronger overall conditions (e.g., significantly stronger radio conditions and / or more powerful processing, memory, and / or battery resources). The WTRUs with a stronger connection and / or stronger overall conditions may enjoy privileged conditions that may provide assistance to the WTRUs with less favorable conditions, e.g., via a site link. Such assistance may be delivered in the form of local model downloads, processing offloading, and / or storage, processing, and / or forwarding capabilities.

[0101] Figure 3 Illustrated is an example of an AI / ML download 300 from a volunteer WTRU 302. In the example, as Figure 3 shown, the volunteer WTRU 302 may communicate with other WTRUs, e.g., via sidelink communication, to download one or more AI / ML models. One or more AI / ML models (e.g., 1209) may be executed as described herein. One or more other WTRUs that contact the volunteer WTRU 302 to download one or more AI / ML models (e.g., 1209, 1209a, and / or as Figures 12A to 12C described) may result in savings of radio resources, storage, and / or energy.

[0102] This document may describe enabling (multiple) WTRU-assisted (e.g., 5G-assisted) discovery and selection to provide AI / ML application assistance via device-to-device communication. For example, an application assistance operation mode and / or one or more ProSe capabilities may be associated with one or more AI and / or ML applications. The AI / ML applications may use performance statistics and / or predictions to determine whether available resources are suitable for enabling a particular AI / ML application operation to be successfully performed. For example, the suitability of available resources may cover an assessment to determine whether a current connection is likely to support model download. However, one or more (e.g., some) examples may be directed to connections over a PDU session. In an example, the performance and / or predictions performed may be customized according to the type of (multiple) connection, which may exclude device-to-device communication.

[0103] When selecting members of a group of WTRUs participating in an AI / ML operation, e.g., based on the AI / ML operation, the different roles that these WTRUs may play may be different compared to those cases where communication is performed via a PDU session. For example, a group of WTRUs that need to download an AI / ML model may utilize the proximity to other WTRUs and / or the proximity to these (multiple) WTRUs when these (multiple) WTRUs are farther from the gNB. Based on one or more characteristics of this proximity, e.g., the WTRU members of the group may be closer to each other but farther from the gNB; the WTRU members of the group may be discovered and / or selected.

[0104] One or more mechanisms may be included to enable an AI / ML application to monitor the predictions and / or performance statistics of resources supporting an AI / ML session, e.g., via device-to-device communication. Mechanisms may be included to enable an AI / ML application to monitor the predictions and / or performance statistics of resources supporting AI / ML communication. Mechanisms may be included to assist the AI / ML application in discovering and / or selecting members of a group and / or possibly the roles that these members should play in the group.

[0105] WTRU-to-network relay can be enabled to provide storage / processing and forwarding capabilities as described herein.

[0106] Volunteer WTRUs that are well-connected to the base station can assist in receiving and / or storing AI / ML models (e.g., receiving and / or storing AI / ML models first). Other WTRUs can (e.g., then) download one or more AI / ML models from the volunteer WTRUs via direct device connection. A mechanism can be implemented that enables an intermediate node (such as a WTRU-to-network relay) to act as an assistant and / or volunteer in a manner that other WTRUs can benefit from its temporary privileged capabilities. In one or more (e.g., some) examples, the WTRU-to-network relay may not store and forward information, and / or process and forward information. Rather, the WTRU-to-network relay can forward the content of a message received from a remote node to the WTRU(s) connected to the relay.

[0107] Embodiments are described herein for enabling an AI / ML application to determine whether a WTRU can provide WTRU-to-network relay capabilities while also being able to store and forward and / or process and forward AI / ML application traffic and / or any other application traffic.

[0108] One or more examples are described herein for enabling a WTRU to discover whether a WTRU-to-network relay can store and forward information, process and store information, and / or both.

[0109] One or more examples are described herein for enabling a WTRU to negotiate additional services (e.g., "store and forward", "process and forward", and / or both) that can be provided at the discovered WTRU-to-network relay WTRU.

[0110] Embodiments are described herein for enabling a (multi) 5GS WTRU-to-network relay to be discovered based on its computing capabilities. A compute-integrated relay can be included as a type of relay service in other (e.g., future) radios, where the relay WTRU can aggregate data received from one or more (e.g., multiple) WTRUs (e.g., and / or aggregate with the relay WTRU's own data), generate aggregated data, and forward the aggregated data to one or more other entities (e.g., network functions, application functions, and / or another WTRU). For example, the relay WTRU can perform model aggregation for a federated learning task.

[0111] One or more ProSe services in 5GS may be communication-oriented and / or may not support such type(s) of WTRU-to-NW relay services. For example, if the WTRU-to-NW relay is unable to perform (e.g., the required) calculations, the WTRU-to-NW relay may not be selected as a relay node even if it can provide communication-oriented relaying.

[0112] This document may describe enhancements to 5G ProSe capabilities. Although the enhancements described herein may focus on WTRU-to-NW relay services, they may also (e.g., directly) apply to WTRU-to-WTRU relay services (e.g., by replacing WTRU-to-NW with WTRU-to-WTRU in one or more of the enhancements described herein).

[0113] This document describes one or more embodiments for implementing 5G ProSe capabilities. Although one or more of the embodiments described herein may focus on WTRU-to-NW relay services, for example one or more of the embodiments may (e.g., directly) apply to WTRU-to-WTRU relay services (e.g., WTRU-to-WTRU may be substituted for WTRU-to-NW in one or more of the embodiments described herein).

[0114] The embodiments described herein may enable AI / ML application traffic carried over device-to-device communication to leverage one or more unique characteristics that can be provided by one or more WTRUs acting as a remote WTRU and / or a WTRU-to-network relay WTRU to other WTRUs. The remote WTRU may be a WTRU on the other side of the relay from the WTRU-to-network relay. For example, when performing processing-intensive, latency-sensitive AI / ML operations, the remote WTRU may be able to discover, select, and / or negotiate specific services from peer WTRUs and / or a WTRU-to-network relay WTRU. Peer WTRUs may be defined as two or more WTRUs that are connected to each other, e.g., via a direct PC5 connection. Prose services may be one or more (e.g., any) services (e.g., games, extended reality services, etc.) and / or they may be AI / ML operations that may include: local AI / ML model downloads, processing of offline inferences, and / or relaying interim training results and / or relaying models downloaded from edge / cloud servers. The embodiments herein may include mechanisms and / or processes to allow a WTRU to proactively volunteer as an assistant, e.g., to support AI / ML application traffic and / or one or more (e.g., any) other traffic, while also acting as / being a peer WTRU and / or a WTRU-to-network relay WTRU and / or providing preprocessing and / or aggregation of traffic from two or more WTRUs, e.g., before the traffic can be further forwarded to one or more other WTRUs and / or forwarded to the network.

[0115] For example, a WTRU-to-network relay may provide the download and / or storage of one or more ML models that may be (e.g., further) downloaded by one or more other WTRUs connected to the WTRU-to-network relay. The WTRU-to-network relay may provide ML model aggregation, which may enable one or more WTRUs connected to the WTRU-to-network relay to send intermediate models that may be further processed (e.g., aggregated) before sending the intermediate ML models to the AI / ML AF. One or more (e.g., new) mechanisms are described herein to enable the AF to request QoS monitoring for WTRUs participating in communication on a PC5 link.

[0116] This document describes embodiments that enable enhancement of ProSe services (e.g., 5G ProSe services) to support AI / ML services. During an initial registration process, as part of its capabilities (e.g., 5GMM capabilities), a WTRU may provide an indication as to whether it can process and / or forward packets from a peer / remote WTRU, store and forward packets from a remote / peer WTRU, and / or both. The store-and-forward functionality may be useful for one or more AI / ML operations related to model distribution and / or sharing, where a WTRU-to-network relay WTRU may store models that are then distributed (e.g., immediately) and / or at a later stage to remote WTRUs that request these models during a ProSe direct communication phase. The process-and-forward functionality may be useful for AI / ML operations related to federated learning operations, where training models from remote WTRUs may be further processed at a WTRU-to-network relay WTRU and then forwarded to an AI / ML application server.

[0117] The information described herein can be added to Prose capabilities (e.g., 5G Prose capabilities). For example, Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports direct discovery (e.g., 5G Prose direct discovery). Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose direct communication (e.g., 5G Prose direct communication). Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose direct communication (e.g., 5G Prose direct communication) to support store-and-forward. Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose direct communication (e.g., 5G Prose direct communication) to support process-and-forward. Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose direct communication (e.g., 5G Prose direct communication) to support both store-and-forward and process-and-forward. Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 2 WTRU-to-network relay (e.g., 5G Prose layer 2 WTRU-to-network relay). Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 3 WTRU-to-network relay (e.g., 5G Prose layer 3 WTRU-to-network relay). Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 3 WTRU-to-network relay (e.g., 5G Prose layer 3 WTRU-to-network relay) to support store-and-forward. Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 3 WTRU-to-network relay (e.g., 5G Prose layer 3 WTRU-to-network relay) to support process-and-forward. Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 3 WTRU-to-network relay (e.g., 5G Prose layer 3 WTRU-to-network relay) to support both store-and-forward and process-and-forward. Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 2 remote WTRU (e.g., 5G Prose layer 2 remote WTRU). Prose capabilities (e.g., 5G Prose capabilities) can indicate whether the WTRU supports Prose layer 3 remote WTRU (e.g., 5G Prose layer 3 remote WTRU).

[0118] The WTRU may indicate whether the capabilities are used as volunteer-provider and / or volunteer-consumer. For example, the WTRU may indicate in a policy container sent to the AMF in a registration message whether the capabilities are used as volunteer-provider and / or volunteer consideration.

[0119] The WTRU may require assistance from one or more other WTRUs, e.g., to perform ML model download on its own behalf and / or to perform ML model aggregation before an intermediate model is sent from the assisting WTRU to the AF. The WTRU may provide assistance information to one or more other WTRUs, e.g., to store one or more ML models and / or to make one or more ML models available to one or more other WTRUs and / or to perform one or more model aggregation operations before the intermediate model ML is sent to the central AF. When the WTRU declares its intention to assist one or more other WTRUs, e.g., to store and forward and / or process and forward data from the requesting assisting WTRU, the WTRU may be referred to as a volunteer-provider. A WTRU seeking assistance from a volunteer-provider WTRU may be referred to as a volunteer-consumer WTRU.

[0120] Depending on the Prose capabilities, e.g., the WTRU may provide parameters (e.g., specific parameters) in a policy container to request relevant resources from the next generation (NG) RAN. The AMF may authorize one or more capabilities, e.g., via the UDM (e.g., authorization obtained from subscriber records and / or subscription data).

[0121] The AMF may store the capabilities, and / or may use the capabilities to discover (e.g., via the Network Repository Function (NRF)) the PCF that can process them. The Prose NR WTRU-PC5 Aggregate Maximum Bit Rate (AMBR) may also be provided to the AMF, e.g., as during one or more (e.g., current) Prose procedures. If the WTRU is authorized to use the capabilities, then e.g., the AMF may include the capabilities in an NG Application Protocol (NGAP) message sent to the RAN.

[0122] The RAN may operate (e.g., continue to operate) with a similar functionality. The AMF may select different NR WTRU-PC5-AMBR values, e.g., depending on whether the WTRU has signaled its intention to be a volunteer-consumer and / or volunteer-producer. The AMF may obtain the PC5 QoS parameters generated by the PCF during the policy association establishment and / or modification procedure.

[0123] The PCF may use parameter values provided by the WTRU in the policy container (e.g., new) to generate values associated with applying AI / ML operations services, which may include support via the PC5 link. The PCF may use access type preferences to indicate whether the WTRU should use a volunteer-provider WTRU as a preferred access. The PCF may use access type preferences to indicate whether the WTRU should use a volunteer-provider WTRU as a preferred access, and / or whether the role should function as a WTRU-to-network relay and / or a peer WTRU on the PC5 link.

[0124] The parameters described herein may vary depending on whether the WTRU operates as a volunteer-provider and / or a volunteer-consumer. The AMF may relay the parameters to the NG-RAN and / or the WTRU.

[0125] The WTRU may use the service request procedure to update Prose preferences regarding AI / ML application operation requirements. The AI / ML application function may use the service-specific information provisioning function (e.g., AF-based service parameter provisioning) to request an update to the AI / ML application parameter values. The AI / ML application function may use the service-specific information provisioning function to support AI / ML application characteristics for a specific AI / ML application operation. Depending on the type of AI / ML operation (e.g., FL and / or model delivery), the AF and / or the application server may update one or more requirements (e.g., characteristics that may be referred to herein) for one or more different AI / ML operations. For example, vertical federated learning operations may have different latency requirements than model distribution / model delivery operations. The AI / ML application server (AS) may provide path preferences associated with specific QoS characteristics for a certain (certain) application. For example, the AI / ML application characteristics that the AI / ML application server (AS) may update may include one or more of the following: expected transmission delay characteristics / requirements, DL / uplink (UL) bitrates of traffic traveling via the PC5, failure to meet QoS parameter values (e.g., guaranteed flow bitrate (GFBR) / maximum flow bitrate (MFBR)), and / or (a) specific PQI value(s). The AI / ML AS may provide path preferences associated with specific QoS characteristics for a certain application, e.g., forcing traffic to a volunteer WTRU that supports those characteristics. Network analysis may be used to collect WTRU information to support Prose-based communication (e.g., by extending the DCAF function). For example, the analysis may be collected depending on whether the PC5 link is used to provide store-and-forward and / or process-and-forward capabilities.

[0126] This document describes one or more embodiments for supporting ProSe capabilities (e.g., enhanced 5G ProSe capabilities) for computational integrated ProSe (e.g., 5G). The ProSe capabilities (e.g., 5G ProSe capabilities) can indicate whether a WTRU (e.g., a relay WTRU and / or a remote WTRU) supports one or more of the following computational integrated ProSe capabilities: 5G ProSe computational integrated relay, i.e., the relay WTRU can support certain computational operations on one or more packets received from one or more remote WTRUs; and / or 5G ProSe computational integrated relay, i.e., the remote WTRU can be configured (e.g., can be willing) to compute its packets with one or more specific computational operations performed by the relay WTRU (e.g., WTRU-to-NW relay).

[0127] The computational operations can be, but are not limited to, one or more of the following.

[0128] A first example of a computational operation can include adding new context. Remote WTRU-A can send one or more packets to the relay WTRU. Each packet can include one or more (e.g., some) sensed information collected by remote WTRU-A. However, additional context information can be (e.g., can be required to be) added to each packet. The additional context information can include multimodal data, such as but not limited to the location of the relay WTRU and / or sensed information from one or more other remote WTRUs. For this example, the parameter 5G ProSe capabilities can (e.g., additionally) indicate the ability to add new context information.

[0129] A second example of a computational operation can include continuous AI model training. One or more packets from a remote WTRU can include a partial AI model (e.g., the output of the first few layers of a deep neural network model). The remote WTRU can send the partial AI model in the form of one or more (e.g., multiple) packets to the relay WTRU. The relay WTRU can receive the one or more packets. The relay WTRU can continue to train the AI model, e.g., using the partial AI model. For this example, the parameter 5G ProSe capabilities can (e.g., additionally) indicate the ability to perform continuous AI model training as a relay WTRU.

[0130] A third example of a computing operation may include AI model fine-tuning. One or more packets from a remote WTRU may include further improving the accuracy of the AI model. The remote WTRU may send the AI model to the relay WTRU. For example, the remote WTRU may send the AI model to the relay WTRU in the form of one or more (e.g., multiple) packets. The relay WTRU may receive one or more packets. The relay WTRU may retrain, fine-tune, and / or customize the AI model, e.g., using additional training data generated by the relay WTRU and / or received from one or more other remote WTRUs. The relay WTRU may retrain, fine-tune, and / or customize the AI model to produce another (e.g., new) AI model with higher accuracy. For this example, the parameter 5G ProSe capability may (e.g., additionally) indicate the ability to perform AI model fine-tuning as a relay WTRU.

[0131] A fourth example of a computing operation may include AI model pruning. One or more packets from a remote WTRU may include a large-sized AI model. Since the size of the AI model may be large, the remote WTRU may send the AI model to the relay WTRU in the form of one or more (e.g., multiple) packets. The relay WTRU may receive one or more packets. The relay WTRU may perform a computing operation on the received AI model to remove one or more (e.g., some) weight parameters of the AI model, thereby generating another (e.g., new) AI model with fewer parameters (e.g., reduced model size). For this example, the parameter 5G ProSe capability may (e.g., additionally) indicate the ability to perform AI model pruning as a relay WTRU.

[0132] A fifth example of a computing operation may include AI model quantization. One or more packets from a remote WTRU may include a large-sized AI model. Since the size of the AI model may be large, the remote WTRU may send the AI model to the relay WTRU in the form of one or more (e.g., multiple) packets. The relay WTRU may receive one or more packets. The relay WTRU may perform a computing operation on the received AI model to quantize each (multiple) weight parameter of the AI model, thereby reducing the model size. For this example, the parameter 5G ProSe capability may (e.g., additionally) indicate the ability to perform AI model quantization as a relay WTRU.

[0133] A relay WTRU can be a WTRU-to-NW relay WTRU and / or a WTRU-to-WTRU relay WTRU. Examples of computations that the relay WTRU can support include one or more of the following. The relay WTRU can compress the content included in a single packet from a remote WTRU. The relay WTRU can combine, aggregate, and / or compress the content included in one or more (e.g., multiple) packets from (e.g., the same) remote WTRU. The relay WTRU can combine, aggregate, and / or compress the content included in one or more (e.g., multiple) packets, e.g., where one or more packets (e.g., each) are from different remote WTRUs. The relay WTRU can combine, aggregate, and / or compress the content included in one or more (e.g., multiple) packets (e.g., each from a different remote WTRU) and additional content that the relay WTRU can generate and / or host locally. The relay WTRU can split the content included in a single packet from a remote WTRU and / or generate one or more (e.g., multiple) packets (e.g., (multiple) additional packets). The relay WTRU can append additional content to a single packet from a remote WTRU. The relay WTRU can analyze the content included in one or more (e.g., multiple) packets from the same remote WTRU (e.g., using an AI algorithm / tool / model) to extract (e.g., meaningful) information from the one or more (e.g., multiple) packets. The relay WTRU can analyze the content included in one or more (e.g., multiple) packets from (multiple) different remote WTRUs (e.g., using an AI algorithm / tool / model) to extract (e.g., meaningful) information from the one or more (e.g., multiple) packets.

[0134] Such computations can be performed on one or more packets from a single remote WTRU (e.g., only on). Additionally or alternatively, such computations can be performed on one or more packets from one or more (e.g., multiple) remote WTRUs. Additionally or alternatively, such computations can be performed on one or more packets from one or more remote WTRUs and the relay WTRU. The (multiple) additional packets generated by the WTRU when performing computations on one or more packets received from (multiple) remote WTRUs can modify, enhance, and / or supersede the semantics of the original content from the packets received from (multiple) remote WTRUs.

[0135] Additionally or alternatively, the 5G ProSe capability may indicate whether the relay WTRU and / or the remote WTRU support one or more of the following compute integrated ProSe capabilities: 5G ProSe compute integrated relay, where the relay WTRU may perform one or more computations on one or more packets from the same remote WTRU; 5G ProSe compute integrated relay, i.e., the remote WTRU may be configured (e.g., may be willing) to have one or more of its packets computed individually by the WTRU-to-NW relay; 5G ProSe compute integrated relay, where the relay WTRU may perform computations on one or more packets from one or more different remote WTRUs; 5G ProSe compute integrated relay, i.e., the remote WTRU may be configured (e.g., may be willing) to have one or more of its packets computed by the WTRU-to-NW relay together with one or more packets from one or more other remote WTRUs; 5G ProSe compute integrated relay, where the relay WTRU may perform computations on one or more packets from one or more remote WTRUs and packets generated (e.g., generated locally) by the relay WTRU; and / or 5G ProSe compute integrated relay, i.e., the remote WTRU may be configured (e.g., may be willing) to have one or more of its packets computed by the WTRU-to-NW relay together with one or more packets from one or more (e.g., any) other WTRUs.

[0136] Computation instructions may specify computations that a relay WTRU may perform. As described herein, one or more (e.g., each) computation scenarios may include one or more different computation instructions. The relay WTRU may perform application-independent and / or application-dependent computations. For example, when the relay WTRU performs (e.g., simple) computations (e.g., combination and / or aggregation of one or more packets from one or more / multiple remote WTRUs), the relay WTRU may not understand the content included in such one or more packets. Thus, for example, such (e.g., simple) computations may be application-independent. For example, the computation instructions may include an indication to combine every three packets from a remote WTRU into another packet. For one or more other computations (e.g., analyzing the content included in one or more packets to extract information), the relay WTRU may (e.g., may need to) understand the content included in one or more packets, which may be application-dependent. Such computations may include the relay WTRU obtaining (e.g., receiving) one or more (e.g., some) inputs (e.g., AI tools / models for analyzing the content included on one or more packets) from one or more applications residing on one or more remote WTRUs. For application-dependent computations, one or more computation instructions may depend on the applications on one or more WTRUs. These applications may specify and / or configure one or more computation instructions to the relay WTRU (e.g., out-of-band) (e.g., directly) and / or in one or more packets (e.g., in-band).

[0137] Although computation instructions may be application-dependent, for example, the computation instructions may include one or more (e.g., common) parameters such as: a computation scope, a computation level, and / or one or more target WTRUs. The computation scope may include a description of the (multiple) target WTRUs, i.e., the computation may be performed by and / or for the relay WTRU. The value of this parameter may include the following indications: one remote WTRU, across one or more (e.g., multiple) remote WTRUs, across one remote WTRU and / or the relay WTRU, and / or across one or more (e.g., multiple) remote WTRUs and / or the relay WTRU. The computation level may include a packet level (e.g., combining one or more / multiple packets). The computation level may include a content level (e.g., extracting information from that included in one or more / multiple packets), etc. The (multiple) target WTRUs may indicate a list of one or more remote WTRUs that corresponds to the computation scope.

[0138] The WTRU may send an indication of the computed integrated ProSe capability to the AMF. As described herein, by embedding the 5G ProSe capability parameter in a registration request message, a service request message, and / or a registration update message, the WTRU (e.g., a relay WTRU and / or a remote WTRU) may send one or more indications of the (multiple) computed integrated ProSe capabilities to the access and mobility management function (AMF) of the WTRU. The AMF may determine whether to authorize the WTRU to use the (multiple) computed integrated 5G ProSe relay services. For example, the AMF may determine whether to authorize the WTRU to use the (multiple) computed integrated 5G ProSe relay services based on its 5G ProSe capability and / or its subscription data that the AMF may retrieve from the unified data management (UDM). The AMF may store the (multiple) authorized computed integrated 5G ProSe services. Additionally or alternatively, the AMF may send the (multiple) authorized computed integrated 5G ProSe services to the policy control function (PCF). The PCF may determine one or more PC5 QoS parameters based on the authorized computed integrated ProSe capability and / or may send one or more PC5 QoS parameters to the AMF, which may be stored at the AMF as part of the WTRU context.

[0139] After the AMF stores the 5G ProSe capability with an indication of the computed integrated ProSe capability, for example, one or more other network functions (NFs) (e.g., the 5G direct discovery name management function (DDNMF)) may retrieve the 5G ProSe capability of a given WTRU and / or subscribe to one or more (e.g., any) (multiple) changes to the 5G ProSe capability of a given WTRU from the AMF, e.g., retrieve and / or subscribe to the (multiple) computed integrated ProSe capabilities of a given WTRU.

[0140] Figure 4 The figure illustrates a flow chart that depicts an example of the process by which an NF and / or an application function (AF) 402 retrieves the computed integrated ProSe capabilities of a given WTRU (e.g., a relay WTRU and / or a remote WTRU). As an example, the NF 402 (and / or the AF) may (e.g., may need to) select a relay WTRU for one or more (e.g., multiple) remote WTRUs. For this purpose, for example, the NF 402 may (e.g., first) obtain one or more computing requirements of the remote WTRU; the NF 402 may (e.g., may also) know one or more (e.g., some) potential relay WTRUs located near one or more remote WTRUs. The NF 402 may (e.g., then) use one or more processes described herein (e.g., Figure 4The (multiple) processes shown are used to retrieve the computed integrated ProSe capabilities of one or more (e.g., each) of these potential relay WTRUs. The NF 402 may (e.g., finally) select a potential relay WTRU as the relay WTRU for the remote WTRU, e.g., if its (multiple) computed integrated ProSe capabilities meet the (multiple) computing requirements of the remote WTRU.

[0141] In 404, the NF / AF 402 (e.g., AF, 5G DDNMF) may send a request to the AMF 406. The request may indicate that the NF / AF 402 wants to retrieve the computed integrated ProSe capabilities of a given WTRU. The request may include an identifier of the WTRU. The request 404 may include a specific computing level and / or computing scope to indicate that the NF / AF 402 wants to retrieve the capabilities of the WTRU related to this specific computing level and / or computing scope.

[0142] The AMF 406 may look up its WTRU context database against the identifier of the WTRU (e.g., and / or computing level and / or computing scope, if included here) and / or may find the computed integrated ProSe capabilities of the WTRU. In 408, the AMF 406 may send a response to the NF / AF 402, including the found computed integrated ProSe capabilities of the WTRU. If the WTRU does not support computed integrated ProSe, then e.g., this response 408 may indicate that the WTRU does not support computed integrated ProSe and / or this response 408 (e.g., merely) may include an empty message (e.g., the response may include only a header and no content body).

[0143] The Application Function (AF) (e.g., ProSe Application Server) may use one or more of the processes described herein to retrieve the computed integrated ProSe capabilities of the WTRU directly from the AMF (e.g., if the AF is trusted by the 5GS) and / or indirectly via the Network Exposure Function (NEF).

[0144] Figure 5The figure shows a flowchart depicting an example of a process 500 in which an NF and / or an Application Function (AF) 502 subscribes to one or more changes to the Computation Integrated ProSe capabilities of a given WTRU (e.g., a relay WTRU and / or a remote WTRU). As an example, the NF 502 (and / or AF) may select and / or approve a relay WTRU with Computation Integrated ProSe capabilities for one or more (e.g., some) remote WTRUs, but may ensure (e.g., may be required to ensure) that the relay WTRU can support (e.g., always support) the Computation Integrated ProSe capabilities as required by the remote WTRU. For this purpose, for example, the NF 502 may (e.g., first) use one or more of the processes described herein (e.g., Figure 5 one or more of the processes) to subscribe to one or more (e.g., any) changes to the Computation Integrated ProSe capabilities of the relay WTRU. If the relay WTRU decides not to support the Computation Integrated ProSe capabilities at a later time, then, for example, the relay WTRU may (e.g., be required to) reduce the energy consumption from computation. For example, the WTRU may send another (e.g., new) (e.g., 5G) ProSe capabilities indication to its serving AMF 504 (e.g., via a registration update), which may indicate that the relay WTRU does not support the Computation Integrated ProSe capabilities. The serving AMF 504 may send a notification including the other (e.g., new) (e.g., 5G) ProSe capabilities of the relay WTRU to the NF 502. The NF 502 (e.g., AF, DDNMF) may decide to select another relay WTRU and / or may notify one or more remote WTRUs that the relay WTRU no longer has the Computation Integrated ProSe capabilities.

[0145] In 506, the NF 502 (e.g., AF, DDNMF) may send a request to the AMF 504. The request 506 may indicate that the NF502 wants to subscribe to one or more changes to the (multiple) Computation Integrated ProSe capabilities of a given WTRU. The request 506 may include an identifier of the WTRU. The Application Function (AF) (e.g., ProSe Application Server) may use one or more of the processes described herein to subscribe to one or more changes to the Computation Integrated ProSe capabilities of the WTRU directly from the AMF (e.g., if the AF is trusted by the 5GS) and / or indirectly via the Network Exposure Function (NEF).

[0146] In 508, the WTRU's computed integrated ProSe capabilities can include changes (e.g., the WTRU can become non - supportive of computed integrated ProSe capabilities, a relay WTRU can need to reduce power consumption from computing, etc.). For example, the relay WTRU can send (e.g., a new) ProSe capabilities indication (e.g., a 5G ProSe capabilities indication) to its serving AMF (e.g., via a registration update). The ProSe capabilities indication can indicate that the relay WTRU will not support computed integrated ProSe capabilities.

[0147] In 510, the AMF 504 can send a notification to the NF 502, including the changed computed integrated ProSe capabilities of the WTRU.

[0148] This document describes systems, methods, and apparatuses for enhancing 5G ProSe authorization information to support computed integrated ProSe. Authorization can include WTRUs that are authorized to provide computed integrated ProSe capabilities.

[0149] After the WTRU (e.g., a relay WTRU and / or a remote WTRU) sends its 5G ProSe capabilities to its AMF, for example, the AMF can determine whether to authorize the WTRU to use (a) computed integrated 5G ProSe relay services based on its (multiple) 5G ProSe capabilities and / or the subscription data that the AMF can retrieve from the UDM. The AMF can store (a) authorized computed integrated 5G ProSe services and / or can send them to the PCF. The PCF can determine one or more PC5 QoS parameters based on the authorized computed integrated 5G ProSe capabilities and / or can send one or more PC5 QoS parameters to the AMF, which can be stored at the AMF as part of the WTRU context.

[0150] If a WTRU is authorized to provide and / or use a (multiple) computational integrated 5G ProSe relay service, for example, its AMF may send 5G ProSe authorization information to the NG-RAN, which may include one or more of the following authorized computational integrated ProSe capabilities: whether the WTRU is authorized to provide a (multiple) 5G ProSe computational integrated relay service as a WTRU-to-NW relay to perform computations on one or more packets received from one or more remote WTRUs; whether the WTRU is authorized to use a (multiple) 5G ProSe computational integrated relay service to have one or more of its packets sent to a relay WTRU (e.g., WTRU-to-NW relay) and / or computed by it; a list of (multiple) WTRU-to-NW relays from which the WTRU may utilize a (multiple) 5G ProSe computational integrated relay service (e.g., and / or a (multiple) 5G ProSe service); and / or a list of one or more remote WTRUs to which the WTRU may provide a (multiple) 5G ProSe computational integrated relay service (e.g., and / or a (multiple) 5G ProSe service).

[0151] Additionally or alternatively, the 5G ProSe authorization information may include one or more of the following authorized computational integrated ProSe capabilities: whether the WTRU is authorized to use a (multiple) 5G ProSe computational integrated relay service to have one or more of its packets computed by a WTRU-to-NW relay (e.g., individually); whether the WTRU is authorized to provide a (multiple) 5G ProSe computational integrated relay service as a WTRU-to-NW relay, where the WTRU may perform computations on one or more packets from one or more different remote WTRUs; whether the WTRU is authorized to use a (multiple) 5G ProSe computational integrated relay service to have one or more of its packets computed by a WTRU-to-NW relay together with one or more packets from one or more remote WTRUs; whether the WTRU is authorized to provide a (multiple) 5G ProSe computational integrated relay service as a WTRU-to-NW relay, where the WTRU may perform computations on one or more packets from one or more remote WTRUs and packets generated (e.g., generated locally) by the WTRU; and / or whether the WTRU is authorized to use a (multiple) 5G ProSe computational integrated relay service to have one or more of its packets computed by a WTRU-to-NW relay together with one or more packets from one or more (e.g., any other) (multiple) WTRUs.

[0152] Additionally or alternatively, the 5G ProSe authorization information may include one or more of the following authorized compute integrated ProSe capabilities: whether the WTRU is authorized to provide (a) 5G ProSe compute integrated relay service as a WTRU-to-NW relay, where the WTRU may perform computations based on one or more in-band instructions embedded in one or more packets from one or more remote WTRUs; whether the WTRU is authorized to use (a) 5G ProSe compute integrated relay service, where the WTRU may be willing to embed one or more in-band compute instructions in one or more of its packets and / or may be willing to have one or more of its packets computed by a WTRU-to-NW relay; whether the WTRU is authorized to provide (a) 5G ProSe compute integrated relay service as a WTRU-to-NW relay, where the WTRU may perform computations based on one or more out-of-band compute instructions (e.g., packet aggregation instructions) configured and / or provisioned to the WTRU; and / or whether the WTRU is authorized to use (a) 5G ProSe compute integrated relay service, where the WTRU may be willing to have one or more of its packets computed by a WTRU-to-NW relay based on one or more out-of-band compute instructions. Although the term 5G ProSe is used herein, one or more similar embodiments for ProSe communication may be implemented in one or more other systems.

[0153] After the AMF stores the 5G ProSe authorization information with the (a) compute integrated ProSe capability indication, for example, one or more other network functions (NFs) (e.g., 5G DDNMF) may retrieve the 5G ProSe authorization information of a given WTRU and / or may search for a WTRU that is authorized to support compute integrated ProSe and / or may be located in the vicinity of the given WTRU.

[0154] Figure 6 A flowchart is illustrated that depicts an example of a process 600 in which an NF retrieves the (a) authorized compute integrated ProSe capabilities of a given WTRU.

[0155] In 604, the NF 602 may send a request to the AMF 606. The request 604 may indicate that the NF 602 wants to retrieve the (a) authorized compute integrated ProSe capabilities of a given WTRU. The request 604 may include an identifier of the WTRU. The request 604 may include a specific compute level and / or compute scope to indicate that the NF 602 (e.g., only) wants to retrieve the authorized capabilities of the WTRU that are related to the specific compute level and / or compute scope. Additionally or alternatively, as described herein, the compute level may indicate one or more (e.g., simple) logical functions, one or more (e.g., more complex) functions, and / or one or more filters.

[0156] The AMF 606 can look up its WTRU context database based on the identifier of the WTRU (e.g., and / or calculation level and / or calculation scope, if included herein) and / or can find the authorized calculation integrated ProSe capabilities of the WTRU.

[0157] At 608, the AMF 606 can send a response to the NF 602, including the found authorized calculation integrated ProSe capabilities of the WTRU. If the WTRU is not authorized to support and / or use any calculation integrated ProSe capabilities, then for example, this response can include an indication that the WTRU does not have any authorized calculation integrated ProSe capabilities and / or this response can (e.g., merely) include an empty message (e.g., the response can only include a header and no content body).

[0158] The application function (AF) (e.g., ProSe application server) can use one or more similar (e.g., the same) procedures described herein to retrieve the authorized calculation integrated ProSe capabilities of the WTRU directly from the AMF 606 (e.g., if the AF is trusted by the 5GS) and / or indirectly via the network exposure function (NEF).

[0159] Figure 7 A flowchart is illustrated that depicts an example of a process 700 for an NF to discover a relay WTRU with authorized calculation integrated ProSe capabilities that may be in the vicinity of a given WTRU.

[0160] In 704, the NF 702 may send a request to the AMF 706. The request may indicate that the NF 702 wants to discover relay WTRUs that satisfy one or more (e.g., some) of the relay WTRU conditions. For example, a relay WTRU may be authorized to support compute integrated ProSe capabilities (e.g., as requested). For example, the relay WTRU may be located in the vicinity of a given WTRU. For example, the relay WTRU may support one or more AI / ML models and / or AI / ML functions (e.g., as described herein, AI / ML models 1209, 1209 and / or as described herein). Thus, for example, the request may include an identifier of the given WTRU and / or the requested compute integrated ProSe capabilities. As described herein, one or more relay WTRU conditions may be included (e.g., included in the request). One or more other relay WTRU conditions may be specified by the NF 702 in one or more future requests to the AMF 706. The NF 702 may dynamically indicate one or more other conditions. For example, the NF 702 may (e.g., at any time) issue one or more other requests indicating one or more different conditions (e.g., different given WTRUs, the relay WTRU should be located in the vicinity of multiple given WTRUs, etc.).

[0161] In 708, the AMF 706 may look up 5G authorized ProSe information of one or more WTRUs, i.e., the AMF 706 maintains its WTRU context against one or more (e.g., two) conditions as described herein. The AMF 706 may find one or more (e.g., multiple) relay WTRUs that satisfy one or more (e.g., two) of the conditions as described herein.

[0162] In 710, the AMF 706 may send a response to the NF 702. The response 710 may include one or more identifiers of the one or more found relay WTRUs as described herein. The response 710 may include an identifier of the WTRU. The NF 702 may be tasked with selecting a relay WTRU for one or more (e.g., several) remote WTRUs. If the NF 702 is tasked with selecting a relay WTRU for one or more (e.g., several) remote WTRUs, then for example, the NF 702 may send the identifier of the WTRU as the relay WTRU to one or more remote WTRUs.

[0163] This document describes the process for discovering a helper WTRU that provides AI / ML application service support via device-to-device communication. This document can provide authorization and provisioning of AI / ML application operation parameters. The AI / ML AS can provision (e.g., via a PDU session) parameter values specific to AI / ML application operations for parameters such as the required PQI, the required time window, and / or a specific geographical location (e.g., (a) tracking area and / or cell ID). One or more parameters can be provisioned on the supported user plane and / or associated with a specific application assistance operation mode (e.g., as described herein with respect to Table 1). Provisioning of parameters specific to AI / ML application operations can be established for communication between the WTRU and the AIML AS (e.g., explicitly). One or more parameters can be provisioned in the mobile device (UE) via the PC1 reference point and / or via the PCF, configured in the universal integrated circuit card (UICC), and / or both.

[0164] To assist the WTRU in determining whether to trigger a discovery mechanism to find the helper WTRU, embodiments can include one or more operation modes that are associated with the context of authorizing the WTRU to use the AI / ML application operation service assistant. For (a) WTRU(s) capable of accessing (a) AI / ML application operation enhancement(s) for device-to-device communication, one or more operation modes (e.g., application assistance operation modes) can be described herein. The WTRU can use the application assistance operation mode and / or the associated configuration to trigger a discovery request (e.g., as specified in Table 1). For example, the WTRU can trigger a ProSe discovery procedure that indicates the application assistance operation mode and one or more ProSe capabilities. The WTRU can be configured to discover one or more AI and / or ML application assistance services (e.g., including one or more AI and / or ML application assistance services associated with the AI / ML 1209 and / or 1209a shown). Figures 12A to 12C shown).

[0165] A WTRU may trigger a discovery request for an assistant if, for example, one or more of the following conditions are met: if the WTRU is not served by the NG-RAN; based on a PQI value = 95, the WTRU needs to meet the AI / ML application operation service requirements; the WTRU is capable of supporting Prose Layer 3 WTRU-to-network relay (e.g., 5G Prose Layer 3 WTRU-to-network relay), supporting store-and-forward; and / or if the WTRU is authorized for this service while being in a certain location and / or a Public Land Mobile Network (PLMN). A relay service code may be associated with the WTRU, which may support one or more (e.g., any one) of the Prose capabilities provided herein (e.g., for relay). One or more ProSe capabilities may include one or more AI / ML application-assisted services, which are associated with at least one or more of the following: data storage, data processing, and / or data aggregation at one or more peer WTRUs and / or one or more relay WTRUs.

[0166] Available application-assisted operation modes may include NG-RAN services with WTRU-to-NW relay support. Available application-assisted operation modes may include NG-RAN services with WTRU peer support. Available application-assisted operation modes may include non-NG-RAN services with WTRU-to-NW relay support. Available application-assisted operation modes may include non-NG-RAN services with WTRU peer support.

[0167] To determine the operation mode to be used, the WTRU may receive provisioning information associating the application-assisted operation mode with the requirements from the AI / ML application operation service (e.g., in the form of a specific PQI value, which is defined to support a Guaranteed Bit Rate (GBR) with a specific latency budget and / or packet error rate) to operate on the PC5 link (e.g., or any other application regarding this). Table 1 depicts an example association of the application-assisted operation modes. Table 1. Application-Assisted Operation Modes

[0168] In addition to the parameters for Prose direct discovery (e.g., 5G Prose direct discovery), the embodiments described herein may also include a Layer 2 ID and / or a Prose discovery WTRU ID (PDUID) that may be assigned to a WTRU supporting the capabilities described herein. When acting as an assistant WTRU, the WTRU may use one or more parameters (e.g., the associated Layer 2 ID) to distinguish the traffic to be relayed and / or to be stored and forwarded or processed and forwarded.

[0169] This document describes embodiments for discovering AI / ML application operation assistants. For WTRU-to-network relay, the user infoID can be enhanced to specify operation modes that the WTRU can identify as volunteer-consumers and / or volunteer-producers. For example, a WTRU that wants to signal an intent to perform as a volunteer-producer WTRU can use a specific user Info ID that meets this criterion (e.g., the intent for the WTRU to act as a volunteer-producer and / or volunteer-consumer). Additionally or alternatively, the Relay Service Code (RSC) can be enhanced and / or extended, and / or updated to also indicate the following Assistant Service Code (HSC) capabilities. The RSC can be extended such that a WTRU performing as a WTRU-to-network relay can (e.g., also) be able to signal what services it can support (e.g., store-and-forward, process-and-forward, and / or both services above) on top of its Layer 3 WTRU-to-network relay service. The RSC can be extended such that one or more non-relay WTRUs can indicate their ability to provide store-and-forward, process-and-forward, and / or both services.

[0170] PDU session parameters (e.g., single network slice selection assistance information (S-NSSAI), data network name (DNN)) can be associated with the HSC that supports specific AI / ML application service characteristics. A WTRU that aims to provide (a) volunteer-producer service(s) can use the (a) PDU session parameter(s) associated with that service. For example, a WTRU that aims to provide one or more volunteer-producer services can establish a PDU session based on the S-NSSAI and / or DNN that supports that service. Additionally, a WTRU acting as a volunteer-producer WTRU can (e.g., in turn) contact other volunteer-producer WTRUs (e.g., via a chain consisting of two peer WTRUs and / or a WTRU-to-network relay WTRU). A WTRU can use an extended RSC that indicates a store-and-forward service to (e.g., also) indicate that the WTRU can use the services of other WTRUs from the same application layer group ID. A mechanism can be used to signal the availability of other services available at the relay WTRU (e.g., specific characteristics of the relay including extended reality (XR) and / or virtual reality (VR) capabilities).

[0171] Embodiments for supporting Prose policy provisioning requests for volunteer operations are described herein. A WTRU may trigger a policy provisioning procedure to retrieve Prose policies such that the WTRU can find an AI / ML application operation business assistant WTRU and / or signal an intention of the WTRU to provide volunteer-producer services for a particular AI / ML application business. The PCF may obtain relevant Prose parameters to be configured in the WTRU using operator-configured Prose parameters stored in the UDR and / or AF-based service parameter provisioning.

[0172] Figure 8 An example Prose parameter provisioning 800 is illustrated.

[0173] In 808, the WTRU 802 may send a WTRU policy provisioning request to the AMF 806. The WTRU 802 may send a policy provisioning request 808 regarding an initial registration. The policy provisioning request may indicate one or more ProSe capabilities (e.g., including one or more AI / ML 1209 application assistance services described herein and / or regarding the AI / ML 1209, 1209a in Figures 12A to 12C ). The policy provisioning request may include a ProSe policy container. The ProSe policy container may include one or more of the following: HSC, operation mode, time window, and / or metadata (e.g., transmission latency, UL / DL data rate, etc.). The procedure 800 may enable the WTRU 802 to signal in 808 its intention to request one or more volunteer services (e.g., act as a volunteer-consumer) and / or provide volunteer services (e.g., act as a volunteer-producer), by indicating an assistant service code, user information identifier (ID), and / or operation mode, and one or more QoS characteristics (e.g., QoS characteristics for vertical federated learning operations) of a validity timer (e.g., time window) associated with the application ID provided by the WTRU in the policy container and / or metadata specified by the WTRU (such as transmission latency and / or GFBR). For example, if the WTRU 802 aims to use an assistant service associated with federated learning (e.g., as described in Figures 12A to 12C ) and / or an assistant service associated with the duration for which specific model data is valid, the WTRU 802 may provide a time window during which requests for an assistant that provides a store-and-forward service or an assistant that provides a process-and-forward service may be valid.

[0174] In 810, the PCF 804 can receive a policy container from the AMF 806 and / or can export one or more services that can be enabled based on information provided by the WTRU 802 in the policy container and / or authorized services that the WTRU 802 has in its subscriber record. Based on the requested operating mode and / or the QoS requirements of the services requested by the WTRU 802 (e.g., handling and forwarding requirements), for example, the PCF 804 can determine the operating mode (e.g., volunteer producer / volunteer consumer) and / or the relevant PDUID corresponding to the authorized volunteer operating mode. In an example, the mapping of the Prose service and / or the HSC to the destination layer 2 ID can be associated with the operating mode of the WTRU 802 that supports store-and-forward and / or handling and forwarding.

[0175] In 812, the PCF 804 can provide the policy container to the WTRU 802, e.g., via the AMF 806 with the values described herein. For example, the PCF 804 can send a Namf_Communication_N1N2MessageTransfer to the AMF 806 (e.g., access type preference [preferred preference = "volunteer-provider", authorization for the volunteer operating mode (provider or consumer), PDUID corresponding to the authorized volunteer operating mode, (multiple) user InfoID, Prose service and mapping of the operating mode to the destination layer ID, (multiple) authorized HSC).

[0176] In 814, one or more network-triggered service request procedures can be initiated, e.g., to deliver a policy.

[0177] In 816, one or more WTRU policies can be delivered using a WTRU configuration update procedure. For example, in 816, the AMF 806 can send one or more WTRU policies to the WTRU 802.

[0178] In 818, the WTRU 802 can provide the result of the configured policy layer group ID. The WTRU 802 can send the delivery result of one or more WTRUs to the AMF 806 (e.g., in 818). The AMF 806 can send a message to the PCF 804 (e.g., in 820). Message 820 can include Namf_Communication_N1N2Mes sageNotify. The message can include the delivery result of one or more WTRU policies.

[0179] Embodiments for supporting Prose direct discovery for volunteer operations are described. A WTRU may be preconfigured to contact a DDNMF, e.g., via a PDU established for a specific DNN and / or S-NSSAI that meets a specific AIML application. The WTRU may provide a Prose application ID and / or an operation mode and / or metadata supporting a specific operation mode, e.g., to obtain a discovery filter corresponding to a Prose application code and / or a Prose restriction code. For example, the WTRU may generate one or more discovery codes based on one or more AI / ML application assistance services indicated by one or more peer WTRUs and / or one or more relay WTRUs during a ProSe discovery process.

[0180] An AI / ML application server (e.g., operating as a Prose application server) may negotiate with the DDNMF for an AI / ML application operation code for a specific AI / ML application operation type (e.g., this information may be provided in the form of metadata).

[0181] Figure 9 A flowchart is illustrated that depicts an example ProSe discovery request 900 (e.g., model B).

[0182] The ProSe discovery request 900 depicts an example of how a volunteer-consumer WTRU 902 may discover (e.g., a suitable) volunteer-provider WTRU by using a model B discovery process. As Figure 9As shown, the WTRU 902 may provide the relevant operation mode that the WTRU 902 is looking for (e.g., whether the WTRU is intended to provide volunteer-producer services and / or request volunteer-consumer services) and the metadata associated with the operation mode (e.g., the volunteer-provider, the metadata supporting the performance characteristics for supporting model distribution). For example, the volunteer-consumer WTRU 902 may provide one or more AI / ML models and / or AI / ML application assistance services (e.g., data aggregation, data processing, data storage, etc.), as described herein. Triggering ProSe discovery may include sending a solicitation message to one or more peer WTRUs and / or one or more relay WTRUs using one or more discovery codes at 904a, 904b, 904c, 904d. For example, the WTRU 902 may send one or more solicitation messages to one or more WTRU volunteer-producers (e.g., 906, 908, 910, 912) (e.g., at 904a, 904b, 904c, 904d). For example, the WTRU volunteer-consumer 902 may send a solicitation message to the WTRU volunteer-producer 906 at 904a. The WTRU volunteer-consumer 902 may send a solicitation message to the WTRU volunteer-producer 908 at 904b. The WTRU volunteer-consumer 902 may send a solicitation message to the WTRU volunteer-producer 910 at 904c. The WTRU volunteer-consumer 902 may send a solicitation message to the WTRU volunteer-producer 912 at 904d. The solicitation message may request one or more ProSe capabilities. One or more ProSe capabilities may include store-and-forward and / or process-and-forward. The solicitation message may indicate the ProSe application identifier (ID), the operation mode, and / or the metadata associated with the operation mode. For example, as part of the metadata associated with the operation mode, the WTRU may indicate that the operation mode is, for example, vertical federated learning and / or the service requires a certain amount of storage and processing capabilities to be executed in the WTRU acting as the volunteer-producer. For example, the WTRU may indicate that the operation mode includes one or more supervised (e.g., (multiple) neural networks) and / or unsupervised learning algorithms, as described herein. The WTRU may indicate the operation mode and / or (multiple) AI / ML models and / or AI / ML algorithms, as described herein (e.g., regarding Figures 12A to 12C ).

[0183] A volunteer WTRU (e.g., one of volunteer-producer WTRUs 906, 908, 910, 912) may receive a solicitation message (e.g., 904a, 904b, 904c, and / or 904d). For example, the solicitation message may be a broadcast message. In an example, the volunteer WTRU (e.g., one of volunteer-producer WTRUs 904a, 904b, 904c, and / or 904d) may use one or more discovery filters obtained from the DDNMF to match / correlate the WTRU with an operating mode and / or metadata characteristics (e.g., requirements). For example, the volunteer WTRU (e.g., a WTRU providing volunteer-producer services) may use one or more discovery filters to determine whether its processing and / or storage capabilities can match specific requirements of, for example, vertical federated learning operations specified in the metadata provided by the WTRU requesting one or more volunteer-producer services. For example, the volunteer-consumer WTRU 902 may generate one or more discovery codes based on one or more AIML application assistance services associated with the volunteer-producer WTRUs 906, 908, 910, 912.

[0184] If the information provided in the filter matches the request from the volunteer - consumer WTRU 902, the WTRU (e.g., the corresponding one of the volunteer - producer WTRU 906, 908, 910, 912) can reply to the volunteer - consumer WTRU 902. For example, in 914a, one of the volunteer - producer WTRUs (e.g., the WTRU volunteer - producer 906) can send a response message to the WTRU volunteer - consumer 902. The response message 914a can include one or more of the following: the matching operation mode, operation - mode - specific security protection, etc. For example, in 914b, another volunteer - producer WTRU (e.g., the WTRU volunteer - producer 908) can send a response message to the WTRU volunteer - consumer 902. The response message 914b can include one or more of the following: the matching operation mode, operation - mode - specific security protection, etc. The WTRU (e.g., the volunteer - consumer WTRU 902) can receive response messages (e.g., 914a, 914b) from one or more of the following: one or more peer WTRUs and / or one or more relay WTRUs. The response messages (e.g., 914a, 914b) can indicate that the corresponding WTRU is configured to perform the (e.g., first) ProSe capability requested in the solicitation message. For example, the response messages (e.g., 914a, 914b) can indicate that the corresponding WTRU (e.g., the volunteer - producer WTRU 906, 908, 910, 912) is configured to perform one or more AI / ML (e.g., AI / ML 1209, 1209a) functions as described herein (e.g., as described herein, with respect to Figures 12A to 12C , and / or includes supervised learning, unsupervised learning, including neural networks, etc.).

[0185] Embodiments are described herein to enable WTRU - to - network relay (e.g., 5GS WTRU - to - network relay) to provide storage / processing and forwarding capabilities. Prose - based connection establishment supporting AI / ML operation services can be described herein. PDU session parameters can be associated with the HSC that supports specific AI / ML application service characteristics. A WTRU intended to provide volunteer - producer services can use the PDU session parameters associated with that service, which can include DNN, S - NSSAI, and / or possibly, whether the traffic on the PDU session is subject to store - and - forward and / or process - and - forward processing.

[0186] Figure 10 Illustrated is Prose direct communication that supports, for example, the volunteer function process 1000.

[0187] As Figure 10As shown, the service authorization, provisioning, and / or discovery processes (e.g., at 1002a, 1002b, 1008, and / or 1010) can be performed according to one or more of the mechanisms described herein. For example, a user information ID can be enhanced to specify an operating mode that a WTRU can identify as a volunteer - consuming WTRU 1004 (e.g., such as the volunteer - consumer 902 shown in Figure 9 ), or a volunteer - producing WTRU 1006 (e.g., such as one of the volunteer producers 906, 908, 910, 912 shown in Figure 9 ). For example, a WTRU that wants to signal its intention to act as a volunteer - producing WTRU can use a specific user information ID that meets this criterion (e.g., to signal its intention to act like a volunteer - producing and / or volunteer - consuming WTRU).

[0188] At 1002a, the authorization and / or provisioning can be performed on the volunteer - producing WTRU 1006. At 1002b, the authorization and / or provisioning can be performed on the volunteer - consuming WTRU 1004. At 1008, a PDU session can be established between the volunteer producer WTRU 1006 and the network (e.g., NG - RAN 1020, AMF 1022, SMF 1024, and / or UPF 1026).

[0189] At 1010, the discovery process can be performed by the volunteer - consuming WTRU 1004 and / or the volunteer - producing WTRU 1006. For example, at 1010, the volunteer - consuming WTRU 1004 can trigger the discovery process. The discovery process can include Figure 9 shown in and / or regarding Figure 9 described in and / or herein one or more processes. For example, the discovery process can include the WTRU volunteer - consuming WTRU 1004 sending a corresponding solicitation message to one or more WTRU volunteer - producers (e.g., such as volunteer producers 906, 908, 910, and / or 912) (e.g., such as in Figure 9among 904a, 904b, 904c, and / or 904d as shown). The solicitation message can request one or more ProSe capabilities. For example, the solicitation message can indicate an application assistance operation mode and / or one or more ProSe capabilities. One or more ProSe capabilities can include store-and-forward and / or process-and-forward. The solicitation message can indicate a ProSe application identifier (ID), an operation mode (e.g., such as an application assistance operation mode), and / or metadata associated with the operation mode. For example, as part of the metadata associated with the operation mode, the volunteer-consumer WTRU 1004 can indicate that the operation mode is, for example, vertical federated learning and / or that the service requires a certain amount of storage and processing capabilities to be performed in the WTRU acting as a volunteer-producer.

[0190] The volunteer-producer WTRU 1006 (e.g., one of the volunteer-producer WTRUs 906, 908, 910, 912) can receive a solicitation message (e.g., such as one of the volunteer-producer WTRUs 904a, 904b, 904c, and / or 904d). For example, the solicitation message can be a broadcast message. In an example, the volunteer-producer WTRU 1006 (e.g., such as one of the volunteer-producer WTRUs 904a, 904b, 904c, and / or 904d) can use one or more discovery filters obtained from the DDNMF to match / correlate the volunteer-producer WTRU with the operation mode and / or metadata characteristics (e.g., requirements). For example, the volunteer-producer WTRU 1006 (e.g., the WTRU providing the volunteer-producer service) can use one or more discovery filters to determine whether its processing capabilities and / or storage capabilities can match the specific requirements of, for example, a vertical federated learning operation specified in the metadata provided by the volunteer-consumer WTRU 1004 that requests one or more volunteer-producer services. If the information provided in the discovery filter matches the request of the volunteer-consumer WTRU 1004, the volunteer-producer WTRU 1006 can reply to the volunteer-consumer WTRU 1004. For example, as part of the discovery process, the volunteer-producer WTRU 1006 can send a response message (e.g., such as in Figure 9In the 914a shown), it is sent to the volunteer - consumer WTRU 1004. The response message may include one or more of the following: a matching operation mode, operation - mode - specific security protection, etc. For example, the response message may indicate that the volunteer - producer WTRU 1006 is configured to perform the (e.g., first) ProSe capability requested in the solicitation message. For example, the response message (e.g., 914a, 914b) may indicate that the volunteer - producer WTRU 1006 (e.g., volunteer - producer WTRUs 906, 908, 910, 912) is configured to perform one or more AI / ML functions (e.g., using an AI / ML model, such as Figure 12A the AI / ML 1209 shown and / or Figure 12B and Figure 12C the neural network 1209a shown), as described herein (e.g., as described herein, with respect to Figures 12A to 12C , and / or including supervised learning, unsupervised learning, including neural networks, etc.).

[0191] The volunteer - consumer WTRU 1004 may be configured to generate one or more discovery codes based on one or more AI / ML functions supported by the volunteer - producer WTRU 1006.

[0192] In 1012, the volunteer - consumer WTRU 1004 may select the volunteer - producer WTRU 1006. In 1012, the volunteer - consumer WTRU 1004 may determine the applicable destination layer 2 ID and / or assistant service code to be used. For example, the first WTRU (e.g., the volunteer - consumer WTRU 1004) may select a second WTRU (the volunteer - producer WTRU 1006) from one or more peer WTRUs and / or one or more relay WTRUs based on one or more discovery codes, as described herein. Selecting the second WTRU may be based on the response message (e.g., such as Figure 9 the response messages 914a, 914b shown) received during the discovery process in 1010, which indicates the (e.g., first) ProSe capability, as described herein. For example, the volunteer - consumer WTRU 1004 may be based on one or more AI / ML models and / or one or more AI / ML algorithms (e.g., such as Figure 12A the AI / ML 1209 shown and / or Figure 12B and Figure 12COne or more ProSe capabilities associated with the neural network 1209a, etc., shown, are used to select the volunteer-producer 1006. In an example, additionally or alternatively, in 1012, the volunteer-consumer WTRU 1004 may send a unicast direct communication request, thereby also signaling the intention of the volunteer-consumer WTRU to use the store-and-forward service and / or the process-and-forward service from the volunteer-consumer WTRU 1004. For example, triggering ProSe direct communication may include the first WTRU sending a ProSe direct communication request to another (e.g., second) WTRU, the request indicating the intention to use (e.g., the first) ProSe capabilities via the other (e.g., second) WTRU. The volunteer-producer WTRU 1006 may generate an associated destination layer 2 ID for communication to transmit the store-and-forward and / or process-and-forward information, e.g., based on the HSC. In an example, the primary layer 2 ID may be used for the relay function. The (e.g., first) WTRU may trigger ProSe direct communication with the selected (e.g., second) WTRU so that the (e.g., first) WTRU can connect to the second WTRU to provide one or more AI / ML application assistance services.

[0193] In 1012a, the volunteer-producer WTRU 1006 may use one or more PDU session parameters (e.g., S-NSSAI, DNN) to establish a PDU session, which may be associated with the HSC supporting the business characteristics of a specific AI / ML application. For example, a WTRU intended to provide the volunteer-producer service may use the PDU session parameters associated with that service. For example, a WTRU (e.g., such as the volunteer-producer WTRU 1006) intended to provide the volunteer-producer service may establish a PDU session according to the S-NSSAI and / or DNN supporting that service.

[0194] If the link layer is modified, then, for example, the WTRU (e.g., the volunteer-producer WTRU 1006) may request a PDU session modification procedure to establish another (e.g., new) QoS flow and / or bind traffic to an existing QoS flow. From this point, for example, uplink and / or downlink relay may start. For example, in 1014, an IP address / prefix may be assigned to the WTRU volunteer-consumer 1004 and the WTRU volunteer-producer 1006. In 1016, the link layer may be modified (e.g., a modification of the layer 2 link modification). In 1016, the modification of the link layer may include the volunteer-producer WTRU 1006 modifying an existing PDU session in 1016a, e.g., for relay.

[0195] In 1017, the volunteer-producer WTRU 1006 may send a message to the SMF 1024. The message may include a remote WTRU report. The remote WTRU report may include a remote WTRU user ID and / or IP address.

[0196] Depending on the selected assistant service code, for example, a WTRU-to-network relay (e.g., the volunteer-producer WTRU 1006) may perform store-and-forward operations and / or process-and-forward operations (e.g., in 1018a, 1018b).

[0197] Embodiments of one or more processes are described herein that enable assistance for AI / ML operations in the application layer for communication over a PC5 link (e.g., 5GS).

[0198] QoS monitoring for communication over a PC5 link that supports AI / ML operations is described herein. The supported existing QoS monitoring assistance functions may be included in the QoS measurements of QoS parameters on the QoS flows supported on a PDU session between the WTRU and the AF. The QoS monitoring may be controlled by the PCF based on PCF policies (e.g., the authorized parameters to be measured). These policies may be delivered to the WTRU during the PDU session modification procedure, e.g., if requested by the AF.

[0199] One or more performance statistics and / or predictions that an AI / ML application may use to determine whether available resources are suitable for successfully performing a particular AI / ML application operation may (e.g., merely) support the connection on the PDU session; the performance and / or predictions provided by these embodiments may be customized according to these types of connections.

[0200] This document may describe extending the QoS monitoring function for (e.g., also for) connections over a PC5 link. The extended QoS monitoring function may include one or more of the following. The extended QoS monitoring function may enable the AF to request QoS monitoring of the link between one or more (e.g., two) specific WTRUs from a group to a WTRU connected to the network through a specific WTRU-to-network relay and between one or more (e.g., two) WTRUs connected through a specific WTRU-to-WTRU relay. The extended QoS monitoring function may enable the SMF and / or NEF to request QoS measurements directly from (e.g., any one of) the WTRU, from the NG RAN, and / or from an OAM entity. The extended QoS monitoring function may enable the NG RAN to provide QoS measurements on each WTRU / PFI (PC5 QoS flow identifier).

[0201] Figure 11 A flowchart is depicted that illustrates an example of QoS monitoring 1100 of a PC5 link.

[0202] In 1102, the AF 1104 (e.g., an AI / ML application function) may request QoS monitoring for a WTRU that may participate in proximity services via a PC5 link. For example, in 1102, the AF 1104 sends a QoS monitoring request to the NEF 1108. The QoS monitoring request may include one or more GPSIs, PFIs, and / or application layer ID / group IDs of the (multiple) WTRUs participating in ProSe communication. The AF 1104 may (e.g., also) be connected to each WTRU via a PDU session. The AF 1104 may use an application ID, an application layer group ID, a Prose identifier, and / or a WTRU identifier (e.g., the Generic Public Subscription Identifier (GPSI) of the (multiple) WTRUs) to request QoS monitoring of the WTRU using the application layer ID / group ID via the PC5 link. Additionally or alternatively, if the AF 1104 is connected to the (multiple) WTRUs via a (e.g., regular) PDU session, then, for example, the AF 1104 may obtain a PFI (PC5 QoS flow identifier) from one or more WTRUs (the AF 1104 is interested in requesting QoS monitoring of the PC5 communication for the WTRU) and / or may use the PFI in a request to the 5GS (e.g., to the NEF). The AF 1104 may (e.g., also) use a relay service code that is associated with a group of WTRUs connecting the WTRU to a network relay and / or with a WTRU associated with a direct WTRU-to-WTRU relay connection.

[0203] In 1106, the NEF 1108 may authorize the QoS monitoring request and / or the NEF 1108 may (e.g., also) convert an external identifier to one or more internal identifiers (e.g., GPSI to SUPI). The NEF 1108 may (e.g., also) use the (multiple) WTRU IDs to identify the SMF serving the WTRU for which QoS monitoring is requested.

[0204] In 1110, the NEF 1108 may send (e.g., forward) the QoS monitoring request to the identified one or more SMFs 1112, as described herein. Additionally or alternatively, the NEF 1108 may send the request directly to the WTRU 1114 (e.g., via the AMF 1116) and / or via the PCF. The PCF may use a QoS monitoring policy to determine the type of QoS monitoring that the AF may (e.g., be allowed to) request (e.g., reporting period, reporting when at a specific location, and / or reporting time window). The PCF and / or the NEF may determine to request QoS monitoring from the access network (e.g., gNB) and / or from the WTRU, e.g., based on one or more operator policies.

[0205] In 1118, the WTRU 1114 may receive a QoS monitoring request, such as from the SMF 1120. The QoS monitoring request may indicate to monitor one or more application service characteristics and / or one or more ProSe capabilities associated with an application assistance operation mode. The SMF 1120 may request QoS monitoring for one or more WTRU IDs requested by the AF 1104. The SMF 1120 may use the WTRU-ID to retrieve the SM context from the UDM and / or may derive the associated PDU session(s) for these one or more WTRUs. The SMF 1120 may use the application ID (e.g., application layer ID) and / or group ID and / or Prose identifier and / or PFI (if provided by the AF 1114) to signal to the WTRU which PC5 links and / or which PC5 QoS flows may (e.g., may be required to) be monitored, and / or what parameters need to be monitored. The SMF 1120 may use the authorized QoS policy provided by the PCF during the policy association process to construct the QoS monitoring request.

[0206] In 1122, additionally or alternatively, the SMF 1120 may request QoS monitoring from the eNB. The SMF 1120 may send the QoS monitoring request to the RAN 1128. The SMF 1120 may use the WTRU-ID to retrieve the SM context from the UDM and / or may derive the associated PDU session(s) for those WTRUs. The SMF 1120 may use the application ID / group ID and / or Prose identifier and / or PFI (if provided by the AF 1114) to signal to the gNB which PC5 links and / or which PC5 QoS flows may (e.g., may be required to) be monitored, and / or what parameters may (e.g., may be required to) be monitored. The SMF 1120 may use the authorized QoS policy provided by the PCF during the policy association process to construct the QoS monitoring request.

[0207] In 1124, the WTRU 1114 may send a QoS monitoring response to the SMF 1120, e.g., in response to a QoS monitoring request received in 1118. The QoS monitoring response may indicate one or more QoS monitoring results associated with one or more application traffic characteristics (e.g., indicated by the QoS monitoring request). The QoS monitoring response may include one or more QoS parameter performance measurements (e.g., PDB, PER, MDBV, etc.). The WTRU 1114 may use a QoS monitoring policy and / or a Prose identifier / Application ID / Group ID to identify the PC5 QoS flows that may (e.g., may need to) be monitored, and / or what parameters may (e.g., may need to) be monitored from the QoS flows. For example, if the Prose identifier is associated with a vertical federated learning operation running on the PC5 link using a specific PFI, the WTRU 1114 may monitor whether the allocated GMBR has crossed any thresholds and / or may report it. The WTRU 1114 may report the monitoring results to the SMF 1120 via the NAS using the control plane. Additionally or alternatively, the WTRU 1114 may use the application layer (e.g., using a data collection function) to report the QoS monitoring results to a data collection application function, from which the originally requesting AF 1104 (e.g., AI / ML AF) may retrieve the data.

[0208] In 1126, additionally or alternatively, the RAN may send a QoS monitoring response to the SMF 1120. The QoS monitoring response sent in 1126 may include one or more QoS parameter performance measurements (e.g., PDB, PER, MDBV, etc.). The gNB may use a QoS monitoring policy and a Prose identifier and / or an Application ID and / or a Group ID to identify the PC5 QoS flows that may (e.g., may need to) be monitored, and / or which (what) parameters may (e.g., may need to) be monitored from the QoS flows (e.g., using the QoS monitoring policy). For example, if the Prose identifier is associated with a vertical federated learning operation running on the PC5 link using a specific PFI, the gNB may monitor whether the allocated GMBR has crossed any thresholds, and if so, report it. The eNB may use the information received from the WTRU 1114 via the sidelink WTRU information NR message to derive the QoS monitoring information requested by the AF 1104.

[0209] The QoS measurement results can be forwarded to the AF 1104 via the NEF 1108. For example, in 1128, the SMF 1120 can send a QoS monitoring response to the NEF 1108. The QoS monitoring response sent in 1128 can include one or more QoS parameter performance measurements (e.g., PDB, PER, MDBV, etc.). In 1130, the NEF 1108 can send the QoS monitoring response to the AF 1104 (e.g., AI / ML AF). The QoS monitoring response sent in 1130 can include one or more QoS parameter performance measurements (e.g., PDB, PER, MDBV, etc.). The QoS monitoring information can be based on the authorized QoS monitoring policy provided by the PCF, as described herein. For example, the AF 1104 can receive QoS monitoring information specifying whether the PC5 link is congested and / or does not meet one or more QoS monitoring requirements. The AF 1104 can decide to no longer consider one or more WTRUs that are part of the PC5 link communication for which the QoS characteristics are no longer met.

[0210] Figure 12A is a diagram of an example system environment 1201 in which the AI / ML 1209 model can be implemented. The AI / ML model 1209 can be implemented at the WTRU and / or in the network (e.g., location management function). The AI / ML 1209 model can include model data and one or more algorithms and / or functions that are configured to learn from the received input data 1207 to train the AI / ML 1209 and / or generate an output 1215. The input data 1207 can be an input in one or more formats, such as an image format, an audio format (e.g., spectrogram or other audio format), a tensor format (e.g., including a one-dimensional or multi-dimensional array), and / or another data type that can be input into the AI / ML 1209 algorithm. The input data 1207 can be the result of a preprocessing 1205 that can be performed on the raw data 1203, or the input data 1207 can include the raw data 1207 itself. The raw data 1203 can include image data, text data, audio data, or another sequence of information, such as a sequence of network information related to a communication network and other types of data. The preprocessing 1205 can include format changes or other types of processing (e.g., averaging, time-domain filtering, and / or frequency-domain filtering) to generate the input data 1207 in a format for input into the AI / ML 1209 algorithm. The output 1215 can be generated by the AI / ML 1209 algorithm in one or more formats, such as a tensor, a text format (e.g., words, sentences, or other text sequences), a numerical format (e.g., predictions), an audio format, an image format (e.g., including a video format), another data sequence format, and / or another output format.

[0211] AI / ML 1209 can be implemented using software and / or hardware as described herein. AI / ML 1209 can be stored as computer-executable instructions on a computer-readable medium that is accessible by one or more processors for execution as described herein. Example AI / ML environments and / or libraries include TENSORFLOW, TORCH, PYTORCH, MATLAB, GOOGLE CLOUD AI, and AUTOML, AMAZON SAGEMAKER, AZURE MACHINE LEARNING STUDIO, and / or ORACLE MACHINE LEARNING.

[0212] AI / ML 1209 can include one or more algorithms configured for unsupervised learning. Unsupervised learning can be implemented using AI / ML 1209 algorithms that learn from input data 1207 without being trained for a specific target output. For example, during unsupervised learning, the AI / ML 1209 algorithms can receive unlabeled data as input data 1207 and determine patterns or similarities in the input data 1207 without additional intervention (e.g., updating parameters and / or hyperparameters). AI / ML 1209 algorithms configured for implementing unsupervised learning can include algorithms configured for identifying patterns, groupings, clusters, anomalies, and / or similarities or other associations in the input data 1207. For example, AI / ML can implement hierarchical clustering algorithms, k-means clustering algorithms, k-nearest neighbor (K-NN) algorithms, anomaly detection algorithms, principal component analysis algorithms, and / or prior algorithms. AI / ML 1209 algorithms configured for unsupervised learning can be implemented on a single device or distributed across multiple devices such that the output 1215 or portions thereof can be aggregated at one or more devices for further processing and / or implementation in other downstream algorithms or processes, as may be further described herein.

[0213] AI / ML 1209 may include one or more algorithms configured for supervised learning. Supervised learning may be implemented using AI / ML 1209 algorithms that are trained during a training process to determine a predictive model using known results. The AI / ML 1209 algorithms may be characterized by parameters and / or hyperparameters, which may be trained during the training process. Parameters may include values derived during the training process. Parameters may include weights, coefficients, and / or biases. AI / ML 1209 may also include hyperparameters. Hyperparameters may include values for controlling the learning process. Hyperparameters may include learning rate, number of epochs, batch size, number of layers, number of nodes in each layer, number of kernels (e.g., for CNN), stride size (e.g., for CNN), size of the kernel in the pooling layer (e.g., for CNN), and / or other hyperparameters. Some parameters and hyperparameters may be used interchangeably.

[0214] AI / ML 1209 may be trained during supervised learning by inputting training data into the AI / ML 1209 algorithms and adjusting the parameters and / or hyperparameters for a known target output 1215 while minimizing the loss or error in the output 1215 generated by the AI / ML 1209 algorithms. The raw data 1230 may include or be segmented into training data, validation data, and / or test data for training, validating, and testing the AI / ML 1209 algorithms, respectively, during supervised learning. The training data, validation data, and / or test data may be preprocessed from the raw data 1203 for input into the AI / ML 1209 algorithms. During supervised learning, the training data may be labeled before being input into AI / ML 1209. The training data may be labeled to teach the AI / ML 1209 algorithms to learn from the labeled data and to test the accuracy of AI / ML 1209 on unlabeled input data 1207 during production / implementation of the AI / ML 1209 algorithms or similar AI / ML 1209 algorithms with similar parameters and / or hyperparameters. The training data may be used to fit the parameters of the AI / ML 1209 model using an optimization function, such as a loss or error function. Generally, the training data includes input data 1207 pairs and corresponding target outputs 1215, and the parameters may be trained to generate that output (e.g., within a threshold loss and error). The trained or fitted AI / ML 1209 model may receive the validation data as input to evaluate the model fit on the training dataset while adjusting the hyperparameters of the AI / ML 1209 model. The AI / ML 1209 model may receive the test data to evaluate the final model fit on the training dataset and to evaluate the performance of the AI / ML 1209 model. One or more of training, validation, and / or testing may be performed during supervised learning of different types of AI / ML 1209 models.

[0215] Supervised learning can be implemented for various types of AI / ML 1209 algorithms, which include algorithms implementing linear regression, logistic regression, neural networks (NNs), decision trees, Bayesian logic, random forests, and / or support vector machines (SVMs). NNs and deep NNs (DNNs) are popular examples of algorithms used in AI / ML models that can be trained using supervised learning. However, AI / ML 209 models can implement one or more NN-based and / or non-NN-based algorithms. Various examples of NNs include: perceptrons, multi-layer perceptrons (MLPs), feed-forward NNs, fully connected NNs, convolutional neural networks (CNNs), recurrent NNs (RNNs), long short-term memory (LSTM) NNs, and / or residual NNs (ResNets). A perceptron is an NN that includes a function that multiplies its inputs by learned weight coefficients to generate an output value. A feed-forward NN is an NN that receives inputs at one or more nodes in the input layer and moves information in the direction from one or more hidden layers to one or more nodes in the output layer. In a feed-forward NN, one or more nodes in a given layer can be connected to one or more nodes in another layer. A fully connected NN is an NN that includes an input layer, one or more hidden layers, and an output layer. In a fully connected NN, each node in a layer is connected to every node in another layer of the NN. An MLP is a class of fully connected feed-forward NNs. A CNN is an NN having one or more convolutional layers configured to perform convolutions. Various types of NNs can have elements including one or more CNNs or convolutional layers, such as generative adversarial networks (GANs). GANs can include conditional GANs (CGANs), cycle-consistent GANs (CycleGANs), StyleGANs, DiscoGANs, and / or IsGANs. A GAN can include a generator submodel and a discriminator submodel. The generator submodel can be configured to receive input data and pass real and independently generated data to the discriminator submodel. The discriminator submodel can be configured to receive real and independently generated data from the generator, discriminate the real and independently generated data, and provide feedback to the generator submodel during training to improve the generator submodel's ability to independently generate outputs based on the received inputs. For example, GANs are popular models for generating data types or data sequences such as image data, audio data, and / or text. An RNN is an NN that is inherently cyclic because nodes include feedback connections and internal hidden states (e.g., memory) that allow the output from a node in the NN to affect subsequent inputs to the same node. An LSTM NN can be similar to an RNN because nodes have feedback connections and internal hidden states (e.g., memory). However, an LSTM NN can include additional gates to allow the LSTM NN to learn long-term dependencies between data sequences. A ResNet is an NN that can include skip connections to skip one or more layers of the NN.The autoencoder can be in the form of AI / ML 1209, which can be implemented for supervised learning such that the parameters and / or hyperparameters can be updated during the training process. The parameters and / or hyperparameters can relate to the encoder part and / or decoder part of the autoencoder. Some NNs include one or more attention layers or functions to enhance or focus on some parts of the input data while reducing or deemphasizing other parts.

[0216] Different types of NNs and / or layers can be implemented to process different types of data and / or produce different types of outputs. For example, an NN can include one or more convolutional layers (e.g., for a CNN or GAN), which can be used to process image data and / or audio data (e.g., spectrograms). Each convolutional layer can vary according to various convolutional layer parameters or hyperparameters such as kernel size (e.g., the field of view of the convolution), stride (e.g., the step of the kernel when traversing the image), padding (e.g., for handling image boundaries), and / or input and output sizes. The image being processed can include one or more dimensions (e.g., a row of pixels or a two-dimensional pixel array). The pixels can be represented according to one or more values (e.g., one or more integer values representing color and / or intensity), which can be received by the convolutional layer. The kernel (which can also be referred to as a convolutional matrix or mask) can be used to extract features from the received input data and / or transform these features. The kernel can be used for blurring, sharpening, edge detection, etc. Example kernel sizes can include matrices such as 3×3, 5×5, 10×10, etc. (e.g., for a 2D image, in pixels). The stride can be a parameter for identifying the amount by which the kernel moves over the image data. An example default stride has a size of 1 or 2 within the matrix (e.g., for a 2D image, in pixels). Padding can include the amount of data added to the boundaries of the image data when the image data is processed by the kernel (e.g., for a 2D image, in pixels). The kernel can move over the input image data (e.g., according to the stride length) and perform a dot product with the overlapping input regions to obtain the activation values of the regions. The output of each convolutional layer can be provided to the next layer of the NN or as the output of the NN itself (e.g., image data, feature map, etc.), where there are updated features based on the convolution.

[0217] An NN can include similar types of layers (e.g., convolutional layers, feedforward layers, fully connected layers, etc.) and / or have similar or different configurations (e.g., size, number of nodes, etc.) for each layer. An NN can also or alternatively include one or more layers having different types and different subsets of NNs, which can be interconnected for training and / or implementation as described herein. For example, an NN can include both convolutional layers and feedforward layers or fully connected layers.

[0218] Figure 12BAn example of a neural network 1209a is illustrated. The purpose of training can be to apply the input 1207a as training data and / or to adjust one or more weights represented as w and x in Figure 12B such that the output 215 from the neural network 209a is close to the desired target value associated with the input 207a value of the training data. In an example, the neural network can include three layers (e.g., as shown in Figure 12B ). During training, for a given input, the difference between the output value and the expected value can be calculated and / or the difference can be used to update one or more weights in the neural network. If a significant (e.g., large) difference is observed between the output value and the (multiple) expected values, then, for example, one or more relatively significant (e.g., large) changes in one or more weights can be expected; small differences (e.g., between the output value and the (multiple) expected values) can include one or more relatively small changes in one or more weights. For example, for positioning, the input 1207a can be reference signal parameters and / or the output 1215 can be an estimated position. The expected value can be position information obtained by a global satellite navigation system (GNSS) with high accuracy.

[0219] Once the neural network 1209a has completed its training, the difference between the output 1215 and the expected value can be below a threshold. The neural network 1209a can be applied or implemented after training for positioning by feeding the input data 1207a and / or by estimating or predicting the output 1215 as the expected result of the associated input 1207a. The output 1215 can be the estimated position and / or location of a WTRU.

[0220] Training the neural network 1209a can include identifying one or more of the following: the input to the neural network, the expected output associated with the input, and / or the actual output from the neural network that is compared to the target value.

[0221] In an example, the neural network model can be characterized by one or more parameters and / or hyperparameters, which can include: the number of weights and / or the number of layers in the neural network.

[0222] As used herein, the term "deep learning" may refer to a class of machine learning algorithms that employ artificial neural networks (e.g., deep neural networks (DNNs)), which are loosely inspired by biological systems and / or include at least one hidden layer. A DNN may be a particular class of machine learning model inspired by the human brain, where the input is linearly transformed and / or passed through a non-linear activation function one or more (e.g., multiple) times. A DNN may include one or more (e.g., multiple) layers, where one or more (e.g., each) layer includes a linear transformation and / or a given non-linear activation function. A DNN may be trained using training data via a backpropagation algorithm. Recently, DNNs have demonstrated state-of-the-art performance in various domains (e.g., speech, vision, natural language, etc.) and / or for various machine learning settings (e.g., supervised, unsupervised, and / or semi-supervised).

[0223] Figure 12C is a diagram of an example system environment 1200 for training and / or implementing an AI / ML model that includes NN 1209a. However, other types of AI / ML models (e.g., including NNs and / or non-NN models) may be similarly trained and / or implemented. NN 1209a may be trained and / or implemented on one or more devices to determine and / or update the parameters and / or hyperparameters 1217 of NN 1209a. The raw data 1203a may be generated from one or more sources. For example, the raw data 1203a may include image data, text data, audio data, or another sequence of information, such as a sequence of network information related to a communication network and / or other types of data. At 1205a, the raw data 1203a may be preprocessed (e.g., averaged, time-domain filtered, and / or frequency-domain filtered) to generate training data 1207a. The preprocessing may include formatting changes or other types of processing to generate the training data 1207a in a format suitable for input into NN 1209a.

[0224] NN 1209a may include one or more layers 1211. The configuration of NN 209a and / or layer 1211 may be based on parameters and / or hyperparameters 1217. As described herein, parameters may include weights or coefficients and / or biases of nodes or functions in layer 1211. Hyperparameters may include learning rate, number of epochs, batch size, number of layers, number of nodes in each layer, number of kernels (e.g., for CNN), stride size (e.g., for CNN), size of the kernel in a pooling layer (e.g., for CNN), and / or other hyperparameters. As described herein, NN 1209a may include a feed-forward NN, a fully-connected NN, a CNN, a GAN, an RNN, a ResNet, and / or one or more other types of NNs. NN 1209a may be composed of one or more different types of NNs or different layers of different types of NNs. For example, NN 1209a may include one or more individual layers that have one or more configurations.

[0225] During the training process, training data 1207a may be input into NN 1209a and may be used to learn parameters and / or adjust hyperparameters 1217. Training may be performed by initializing the parameters and / or hyperparameters of NN 1209a, generating and / or accessing training data 207a, inputting training data 207a into NN 1209a, calculating the error or loss from the output of NN 1209a to the target output 1215a via a loss function 1213 (e.g., using gradient descent and / or associated backpropagation), and / or updating the parameters and / or hyperparameters 1217.

[0226] The loss function 1213 may be implemented using gradient update and / or gradient descent techniques based on backpropagation, such as stochastic gradient descent (SGD), synchronous SGD, asynchronous SGD, batch gradient descent, and / or mini-batch gradient descent. Examples of loss or error functions may include functions for determining squared error loss, mean squared error (MSE) loss, mean absolute error loss, mean absolute percentage error loss, mean squared logarithmic error loss, pixel-based loss, per-pixel loss, cross-entropy loss, logarithmic loss, and / or benchmark-based loss. The loss function may be implemented according to one or more quality metrics, such as a signal-to-noise ratio (SNR) metric or another signal or image quality metric.

[0227] An optimizer may be implemented together with the loss function 1213. The optimizer may be an algorithm or function that is configured to adapt the properties of NN 1209a, such as the learning rate and / or weights, to improve the accuracy of NN 1209a and / or reduce the loss or error. The optimizer may be implemented to update the parameters and / or hyperparameters 1217 of NN 1209a.

[0228] The training process can be iterative to update the parameters and / or hyperparameters 1217 until an end condition is reached. The end condition can be reached when the output of the NN 209a is within a predefined threshold of the target output 1215a.

[0229] After completing the training process, the trained NN 1209a or portions thereof can be stored for implementation by one or more devices. The trained NN 1209a or portions thereof can be implemented in other downstream algorithms or processes, as may be further described herein. The trained NN 1209a or portions thereof can be implemented on the same device that performed the training. The trained NN 1209a or portions thereof can be transmitted or otherwise provided to another device for implementation. For example, the NNs 1209b, 1209c can include one or more portions of the trained NN 1209a. The NNs 1209b and 1209c receive corresponding input data 1207b, 1207c, and generate corresponding outputs 1215b, 1215c. The outputs 1215b, 1215c can be generated in one or more formats, such as tensors, text formats (e.g., words, sentences, or other text sequences), numerical formats (e.g., predictions), audio formats, image formats (e.g., including video formats), another data sequence format, and / or another output format. The outputs 1215b, 1215c can be aggregated at one or more devices for further processing and / or implementation in other downstream algorithms or processes, as may be further described herein.

[0230] After completing the training process, the trained parameters and / or the adjusted hyperparameters 1217 or portions thereof can be stored for implementation by one or more devices. The trained parameters and / or the adjusted hyperparameters 1217 or portions thereof can be implemented in other downstream algorithms or processes, as may be further described herein. The trained parameters and / or the adjusted hyperparameters 1217 or portions thereof can be implemented on the same device that performed the training. The trained parameters and / or the adjusted hyperparameters 1217 or portions thereof can be transmitted or otherwise provided to another device for implementation. For example, transmitted or otherwise provided to another device or other devices, which can implement the NNs 1209b, 1209c based on the trained parameters and / or the adjusted hyperparameters 1217. For example, the NNs 1209b, 1209c can be constructed at another device based on the trained parameters and / or the adjusted hyperparameters 1217 or portions thereof. The NNs 1209b and 1209c can be configured from the parameters and / or hyperparameters 1217 or portions thereof to receive corresponding input data 207b, 207c, and generate corresponding outputs 1215b, 1215c. The outputs 1215b, 1215c can be generated in one or more formats, such as tensors, text formats (e.g., words, sentences, or other text sequences), numerical formats (e.g., predictions), audio formats, image formats (e.g., including video formats), another data sequence format, and / or another output format. The outputs 1215b, 1215c can be aggregated at one or more devices for further processing and / or implementation in other downstream algorithms or processes, as may be further described herein.

[0231] The AI / ML models described herein can be implemented on one or more devices. For example, AI / ML 1209 can be implemented in whole or in part on one or more devices, such as one or more WTRUs, one or more base stations, and / or one or more other network entities, such as network servers. Example networks in which AI / ML can be distributed can include federated networks. A federated network can include a set of decentralized devices, each device including AI / ML. As shown in FIG. 12, AI / ML 1209b and AI / ML 1209c can be distributed across separate devices. Although FIG. 12 shows two models (e.g., AI / ML 1209b and AI / ML 1209c), any number of models can be implemented on any number of devices. AI / ML can be implemented to perform collaborative learning, where AI / ML is trained across multiple devices. In another example, AI / ML can be trained at a centralized location or device, and one or more parts of AI / ML or the trained parameters and / or the tuned hyperparameters can be distributed to decentralized locations. For example, updated parameters or hyperparameters can be sent to one or more devices for updating and / or implementing AI / ML thereon.

[0232] AI / ML can be used to estimate a location (e.g., the location of a WTRU). For example, the WTRU and / or the network can use AI / ML to estimate a location (e.g., the location of a WTRU). The WTRU and / or the network can estimate a location based on one or more measurements (e.g., one or more measurements on the received PRS transmitted from one or more TRPs and / or based on the location of one or more TRPs). For example, the WTRU and / or the network can train an AI / ML model based on one or more actual and / or estimated measurements, as described herein. As used herein, the term network can include an Application Management Function (AMF), a Location Management Function (LMF), a base station (e.g., gNB), and / or a Next Generation Radio Access Network (NG-RAN).

Claims

1. A method performed by a first wireless transmit / receive unit (WTRU), the method comprising: Triggering a ProSe discovery procedure by indicating an application assisted operation mode and one or more proximity services (ProSe) capabilities, wherein the first WTRU is configured to discover one or more artificial intelligence machine learning (AIML) application assisted services, and wherein the one or more ProSe capabilities include the one or more AIML application assisted services, the AIML application assisted services being associated with at least one of data storage, data processing, or data aggregation at one or more peer WTRUs or one or more relay WTRUs; During the ProSe discovery procedure, generating one or more discovery codes based on the one or more AIML application assisted services associated with the one or more peer WTRUs or the one or more relay WTRUs; Selecting a second WTRU from the one or more peer WTRUs and the one or more relay WTRUs based on the one or more discovery codes; And Triggering ProSe direct communication with the selected second WTRU such that the first WTRU can connect to the second WTRU to provide the one or more AIML application assisted services.

2. The method according to claim 1, wherein triggering the ProSe discovery procedure includes sending a solicitation message to the one or more peer WTRUs or the one or more relay WTRUs using the one or more discovery codes, the solicitation message requesting a first ProSe capability among the one or more ProSe capabilities.

3. The method according to claim 2, the method further comprising receiving a response message from one or more of: the one or more peer WTRUs or the one or more relay WTRUs, wherein the response message indicates that the corresponding WTRU is configured to perform the first ProSe capability requested in the solicitation message.

4. The method according to claim 3, wherein the first ProSe capability includes store and forward or process and forward.

5. The method according to any one of claims 2 to 4, wherein the solicitation message indicates one or more of a ProSe application identifier (ID), an operation mode, or metadata associated with the operation mode.

6. The method according to any one of claims 3 to 5, wherein selecting the second WTRU is based on the response message indicating the first ProSe capability.

7. The method according to any one of claims 2 to 6, wherein triggering Prose direct communication includes the first WTRU sending a ProSe direct communication request to the second WTRU, indicating an intention to use the first ProSe capability via the second WTRU.

8. The method according to any one of claims 1 to 7, wherein the application assisted operation mode or the one or more ProSe capabilities are associated with one or more AI or ML applications.

9. The method according to any one of claims 1 to 8, the method further comprising sending a supply request indicating the provision of the one or more ProSe capabilities, wherein the supply request includes one or more of an assistant service code, a user information ID, an operation mode, or one or more quality of service (QoS) characteristics.

10. The method according to any one of claims 1 to 9, the method further comprising: receiving a quality of service (QoS) monitoring request, the request indicating monitoring of one or more application traffic characteristics associated with the application assisted operation mode or the one or more ProSe capabilities; and sending a QoS monitoring response, the response indicating one or more QoS monitoring results associated with the one or more application traffic characteristics.

11. A first wireless transmit / receive unit (WTRU), the WTRU including a processor and a transceiver, the processor configured to: trigger a ProSe discovery process by indicating an application assisted operation mode and one or more proximity services (ProSe) capabilities, wherein the processor is configured to discover one or more artificial intelligence machine learning (AIML) application assisted services, and wherein the one or more ProSe capabilities include the one or more AIML application assisted services, the AIML application assisted services being associated with at least one of data storage, data processing, or data aggregation at one or more peer WTRUs or one or more relay WTRUs; generate one or more discovery codes during the ProSe discovery process based on the one or more AIML application assisted services associated with the one or more peer WTRUs or the one or more relay WTRUs; select a second WTRU from the one or more peer WTRUs and the one or more relay WTRUs based on the one or more discovery codes; and trigger ProSe direct communication with the selected second WTRU such that the first WTRU can connect to the second WTRU to provide the one or more AIML application assisted services.

12. The first WTRU according to claim 11, wherein the processor is configured to trigger the ProSe discovery procedure to include: The processor is configured to send a solicitation message to the one or more peer WTRUs or the one or more relay WTRUs using the one or more discovery codes, the solicitation message requesting a first ProSe capability among the one or more ProSe capabilities.

13. The first WTRU according to claim 12, wherein the processor is further configured to receive, via the transceiver, a response message from one or more of: the one or more peer WTRUs or the one or more relay WTRUs, wherein the response message indicates that the corresponding WTRU is configured to perform the first ProSe capability requested in the solicitation message.

14. The first WTRU according to claim 13, wherein the first ProSe capability includes store and forward or process and forward.

15. The first WTRU according to any one of claims 12 to 14, wherein the solicitation message indicates one or more of a ProSe application identifier (ID), an operation mode, or metadata associated with the operation mode.

16. The first WTRU according to any one of claims 13 to 15, wherein the processor is configured to select the second WTRU to include: The processor is configured to select a second WTRU based on the response message indicating the first ProSe capability.

17. The first WTRU according to any one of claims 12 to 16, wherein the processor is configured to trigger ProSe direct communication to include: The processor is configured to send a ProSe direct communication request to the second WTRU, indicating an intention to use the first ProSe capability via the second WTRU.

18. The first WTRU according to any one of claims 11 to 17, wherein the application-assisted operation mode or the one or more ProSe capabilities are associated with one or more AI or ML applications.

19. The first WTRU according to any one of claims 11 to 18, wherein the processor is further configured to send a provision request indicating the one or more ProSe capabilities, wherein the provision request includes one or more of an assistant service code, a user information ID, an operation mode, or one or more quality of service (QoS) characteristics.

20. The first WTRU according to any one of claims 11 to 19, wherein the processor is further configured to: Receive, via the transceiver, a quality of service (QoS) monitoring request that indicates monitoring of one or more application traffic characteristics associated with the application-assisted operation mode or the one or more ProSe capabilities; and Send, via the transceiver, a QoS monitoring response that indicates one or more QoS monitoring results associated with the one or more application traffic characteristics.

Images