Generative model for CSI estimation, compression and RS overhead reduction

Abstract

One or more systems, methods, and / or devices for estimating channel state information (CSI) using a generative model are disclosed herein. In some cases, a compressed representation of CSI may also be simultaneously estimated. In some cases, the methods and techniques may reduce reference symbols (RSs) required for a channel estimation process.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Application No. 63 / 410,954, filed on September 28, 2022, the contents of which are incorporated herein by reference. Background Art

[0003] In the field of wireless communications, there is a need to improve the efficiency of transmissions. For example, when sending feedback, it may be beneficial to find ways to reduce the size of that transmission. Summary of the invention

[0004] One or more systems, methods and / or devices for estimating channel state information (CSI) using a generative model are disclosed herein. In some cases, a compressed representation of the CSI can also be estimated simultaneously. In some cases, the methods and techniques can reduce the reference symbols (RS) required for the channel estimation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] A more detailed understanding may be obtained from the following description given by way of example in conjunction with the accompanying drawings, in which like reference numerals indicate like elements, and in which:

[0006] Figure 1A is a system diagram illustrating an example communication system in which one or more disclosed embodiments may be implemented;

[0007] Figure 1B is a diagram showing that according to an embodiment, Figure 1A A system diagram of an example wireless transmit / receive unit (WTRU) for use within the communication system shown;

[0008] Figure 1C is a diagram showing that according to an embodiment, Figure 1A A system diagram of an example radio access network (RAN) and an example core network (CN) used within the illustrated communication system;

[0009] Figure 1D is a diagram showing that according to an embodiment, Figure 1A A system diagram of another example RAN and another example CN used within the illustrated communication system;

[0010] Figure 2 An example of data exchange between a WTRU and a base station is shown;

[0011] Figure 3 An example of a CSI measurement setup is shown;

[0012] Figure 4 An example of codebook-based precoding with feedback information is shown;

[0013] Figure 5 An example of autoencoder training for channel matrix compression is shown;

[0014] Figure 6 An example of MIMO channel estimation using a decoder / generative model is shown;

[0015] Figure 7 An example of generation and transmission of compressed CSI is shown;

[0016] Figure 8 An example of RS selection using a holdout set is shown;

[0017] Fig. 9 shows an example process of sending a latent vector; and

[0018] Fig.10 An example process for sending an optimized subset of RSs is shown. DETAILED DESCRIPTION

[0019] Table 1 includes a non-exhaustive list of acronyms used herein.

[0020]

[0021]

[0022]

[0023]

[0024] Table 1

[0025] Figure 1A 1 is a diagram showing an example communication system 100 in which one or more disclosed embodiments may be implemented. The communication system 100 may be a multiple access system that provides content such as voice, data, video, messaging, broadcast, etc. to multiple wireless users. 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 use 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 single word discrete Fourier transform spread OFDM (ZT-UW-DFT-S-OFDM), single word OFDM (UW-OFDM), resource block filtered OFDM, and filter bank multi-carrier (FBMC), etc.

[0026] like Figure 1AAs shown, the communication system 100 may include wireless transmit / receive units (WTRUs) 02a, 102b, 102c, 102d, a radio access network (RAN) 104, a core network (CN) 106, a public switched telephone network (PSTN) 108, the Internet 110, and other networks 112, but it should 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. As examples, the WTRUs 102a, 102b, 102c, 102d (any of which may be referred to as a station (STA)) may be configured to send and / or receive wireless signals and may include user equipment (UE), mobile stations, fixed or mobile subscriber units, subscription-based units, pagers, cellular phones, personal digital assistants (PDAs), smartphones, laptops, netbooks, personal computers, wireless sensors, hotspots or Mi-Fi devices, Internet of Things (IoT) devices, watches or other wearable devices, head-mounted displays (HMDs), vehicles, drones, medical devices and applications (e.g., mobile phones, mobile phones, personal digital assistants (PDAs), smartphones, laptops, netbooks, personal computers, wireless sensors, hotspots or Mi-Fi devices, Internet of Things (IoT) devices, watches or other wearable devices, head-mounted displays (HMDs). For example, remote surgery), industrial devices and applications (e.g., robots and / or other wireless devices operating in industrial and / or automated process chain environments), consumer electronic devices, devices operating on commercial and / or industrial wireless networks, etc. Any of the WTRUs 102a, 102b, 102c, and 102d may be interchangeably referred to as a UE.

[0027] The communication system 100 may also include a base station 114a and / or a base station 114b. Each of the base stations 114a, 114b may be any type of device configured to facilitate access to one or more communication networks (e.g., the CN 106, the Internet 110, and / or other networks 112) by wirelessly interfacing with at least one of the WTRUs 102a, 102b, 102c, 102d. By way of example, the base stations 114a, 114b may be a base transceiver station (BTS), a Node B, an eNode B (eNB), a Home Node B, a Home eNode B, a Next Generation Node B (e.g., a gNode B (gNB)), a New Radio (NR) Node B, a site controller, an access point (AP), a wireless router, a transmission reception point (TRP), a network (NW), and the like. Although each of the base stations 114a, 114b is depicted as a single component, it should be appreciated that the base stations 114a, 114b may include any number of interconnected base stations and / or network components. TRP (e.g., transmit and receive point) may be used interchangeably with one or more of TP (transmit point), RP (receive point), RRH (radio remote head), DA (distributed antenna), BS (base station), sector (of a BS), and cell (e.g., a geographic cell area served by a BS), but still consistent with the present invention. Thereafter, multi-TRP may be used interchangeably with one or more of MTRP, M-TRP, and multiple TRPs, but still consistent with the present disclosure.

[0028] The base station 114a may be part of the RAN 104, which may also include other base stations and / or network elements (not shown), such as a base station controller (BSC), a radio network controller (RNC), a relay node, and the like. The base station 114a and / or the base station 114b may be configured to send 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 a licensed spectrum, an unlicensed spectrum, or a combination of a licensed spectrum and an unlicensed spectrum. A cell may provide coverage for wireless services to a specific geographic area, which may be relatively fixed or may change over time. The cell may be further divided into cell sectors. For example, a cell associated with the base station 114a may be divided into three sectors. Thus, in one embodiment, the base station 114a may include three transceivers, that is, each transceiver corresponds to a sector of the cell. In an embodiment, the base station 114a may use multiple-input multiple-output (MIMO) technology, and may use multiple transceivers for each sector of the cell. For example, beamforming may be used to send and / or receive signals in a desired spatial direction.

[0029] The base stations 114a, 114b may communicate with one or more of the WTRUs 102a, 102b, 102c, 102d over an air interface 116, which may be any suitable wireless communication link (e.g., radio frequency (RF), microwave, centimeter wave, micrometer wave, infrared (IR), ultraviolet (UV), visible light, etc.). The air interface 116 may be established using any suitable radio access technology (RAT).

[0030] More specifically, as described above, the communication system 100 may be a multiple access system and may use one or more channel access schemes such as CDMA, TDMA, FDMA, OFDMA, SC-FDMA, etc. For example, the base station 114a in the RAN 104 and the WTRUs 102a, 102b, 102c may implement a radio technology such as Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access (UTRA), which may use Wideband CDMA (WCDMA) to establish the air interface 116. WCDMA may include communication protocols such as High Speed ​​Packet Access (HSPA) and / or Evolved HSPA (HSPA+). HSPA may include High Speed ​​Downlink (DL) Packet Access (HSDPA) and / or High Speed ​​Uplink (UL) Packet Access (HSUPA).

[0031] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c may implement a radio technology such as Evolved UMTS Terrestrial Radio Access (E-UTRA), which may establish the air interface 116 using Long Term Evolution (LTE) and / or Advanced LTE (LTE-A) and / or Advanced LTE Pro (LTE-A Pro).

[0032] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c may implement a radio technology such as NR radio access, which may establish the air interface 116 using NR.

[0033] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c may implement multiple radio access technologies. For example, the base station 114a and the WTRUs 102a, 102b, 102c may implement LTE radio access and NR radio access together, for example using the dual connectivity (DC) principle. Thus, the air interface used by the WTRUs 102a, 102b, 102c may 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).

[0034] 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 (Worldwide Interoperability for Microwave Access (WiMAX)), CDMA2000, CDMA2000 1X, CDMA2000EV-DO, Interim Standard 2000 (IS-2000), Interim Standard 95 (IS-95), Interim Standard 856 (IS-856), Global System for Mobile Communications (GSM), GSM Enhanced Data Rates for Evolution (EDGE), GSM EDGE (GERAN), etc.

[0035] As an example, Figure 1A The base station 114b in the example may be a wireless router, a Home NodeB, a Home eNodeB, or an access point, and may use any appropriate RAT to facilitate wireless connectivity in a local area, such as a business location, a residence, a vehicle, a campus, an industrial facility, an air corridor (e.g., for use by drones), a road, and the like. In one embodiment, the base station 114b and the WTRUs 102c, 102d may establish a wireless local area network (WLAN) by implementing a radio technology such as IEEE 802.11. In an embodiment, the base station 114b and the WTRUs 102c, 102d may establish a wireless personal area network (WPAN) by implementing a radio technology such as IEEE 802.15. In another embodiment, the base station 114b and the WTRUs 102c, 102d may establish a picocell or a femtocell by using a cellular-based RAT (e.g., WCDMA, CDMA2000, GSM, LTE, LTE-A, LTE-A Pro, NR, etc.). Figure 1A As shown, the base station 114b may have a direct connection to the Internet 110. Thus, the base station 114b may not be required to access the Internet 110 via the CN 106.

[0036] The RAN 104 may be in communication with the CN 106, 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, fault tolerance requirements, reliability requirements, data throughput requirements, mobility requirements, etc. The CN 106 may provide call control, billing services, mobile location-based services, pre-paid calling, Internet connectivity, video distribution, etc., and / or perform high-level security functions, such as user authentication. Although in Figure 1AAlthough not shown, it will be appreciated that the RAN 104 and / or the CN 106 may be in direct or indirect communication with other RANs that employ the same RAT or a different RAT as the RAN 104. For example, in addition to being connected to the RAN 104, which may employ NR radio technology, the CN 106 may also be in communication with another RAN (not shown) employing GSM, UMTS, CDMA 2000, WiMAX, E-UTRA, or WiFi radio technology.

[0037] The CN 106 may also serve 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 networks 112 may include wired and / or wireless communication networks owned and / or operated by other service providers. For example, the networks 112 may include another CN connected to one or more RANs, which may use the same RAT as the RAN 104 or a different RAT.

[0038] 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 via different wireless links). Figure 1A The illustrated WTRU 102c may be configured to communicate with the base station 114a, which may employ a cellular-based radio technology, and with the base station 114b, which may employ an IEEE 802 radio technology.

[0039] Figure 1B is a system diagram showing an example WTRU 102. Figure 1B As shown, the WTRU 102 may include, among other things, a processor 118, a transceiver 120, a transmit / receive element 122, a speaker / microphone 124, a keyboard 126, a display / touchpad 128, non-removable memory 130, removable memory 132, a power source 134, a global positioning system (GPS) chipset 136, and / or peripherals 138. It will be appreciated that the WTRU 102 may include any sub-combination of the foregoing elements while remaining consistent with an embodiment.

[0040] The processor 118 may be a general purpose processor, a special purpose processor, a conventional processor, a digital signal processor (DSP), a plurality of microprocessors, one or more microprocessors associated with a DSP core, a controller, a microcontroller, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), any other type of integrated circuit (IC), a state machine, etc. The processor 118 may perform signal coding, data processing, power control, input / output processing, and / or any other functionality that enables the WTRU 102 to operate in a wireless environment. The processor 118 may be coupled to the transceiver 120, which may be coupled to the transmit / receive element 122. Although Figure 1B The processor 118 and the transceiver 120 are depicted as separate components, but it will be appreciated that the processor 118 and the transceiver 120 may be integrated together in an electronic package or chip.

[0041] The send / receive element 122 may be configured to send or receive signals to or from a base station (e.g., base station 114a) via an air interface 116. For example, in one embodiment, the send / receive element 122 may be an antenna configured to send and / or receive RF signals. In an embodiment, the send / receive element 122 may be a transmitter / detector configured to send and / or receive, for example, IR, UV, or visible light signals. In yet another embodiment, the send / receive element 122 may be configured to send and / or receive both RF and optical signals. It should be understood that the send / receive element 122 may be configured to send and / or receive any combination of wireless signals.

[0042] Although the transmit / receive element 122 is Figure 1B Although depicted as a single element in the figure, the WTRU 102 may include any number of transmit / receive elements 122. More specifically, the WTRU 102 may employ MIMO technology. Thus, in one embodiment, the WTRU 102 may include two or more transmit / receive elements 122 (e.g., multiple antennas) for transmitting and receiving wireless signals over the air interface 116.

[0043] The transceiver 120 may be configured to modulate signals to be transmitted by the transmit / receive element 122 and to demodulate signals received by the transmit / receive element 122. As described above, the WTRU 102 may have multi-mode capabilities. Thus, the transceiver 120 may include multiple transceivers to enable the WTRU 102 to communicate via multiple RATs (e.g., NR and IEEE 802.11).

[0044] The processor 118 of the WTRU 102 may be coupled to the speaker / microphone 124, the keyboard 126, and / or the display / touchpad 128 (e.g., a liquid crystal display (LCD) display unit or an organic light emitting diode (OLED) display unit), and may receive user input data from these components. The processor 118 may also output user data to the speaker / microphone 124, the keyboard 126, and / or the display / touchpad 128. In addition, the processor 118 may access information from and store data in any type of suitable memory, such as the non-removable memory 130 and / or the removable memory 132. The non-removable memory 130 may include a random access memory (RAM), a read-only memory (ROM), a hard disk, or any other type of memory storage device. The removable memory 132 may include a subscriber identity module (SIM) card, a memory stick, a secure digital (SD) memory card, and the like. In other embodiments, the processor 118 may access information from and store data in memories that are not physically located on the WTRU 102, such as on a server or a home computer (not shown).

[0045] The processor 118 may receive power from the power source 134, and may be configured to distribute and / or control power for use by other components in the WTRU 102. The power source 134 may be any suitable device for powering the WTRU 102. For example, the power source 134 may 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.

[0046] The processor 118 may also be coupled to the GPS chipset 136, which may be configured to provide location information (e.g., longitude and latitude) regarding the current location of the WTRU 102. In addition to or in lieu of the information from the GPS chipset 136, the WTRU 102 may receive location information from a base station (e.g., base stations 114a, 114b) over the air interface 116 and / or determine its location based on the timing of signals received from two or more nearby base stations. It should be appreciated that the WTRU 102 may acquire location information via any suitable location-determination method while remaining consistent with an embodiment.

[0047] The processor 118 may also be coupled to peripherals 138, which may include one or more software and / or hardware modules that provide additional features, functionality, and / or wired or wireless connectivity. For example, peripherals 138 may include an accelerometer, an electronic compass, a satellite transceiver, a digital camera (for photos and / or video), a universal serial bus (USB) port, a vibration device, a television transceiver, a hands-free headset, Module, FM radio unit, digital music player, media player, video game console module, Internet browser, virtual reality and / or augmented reality (VR / AR) device, activity tracker, etc. Peripheral device 138 may include one or more sensors. The sensor may be one or more of a gyroscope, an accelerometer, a Hall effect sensor, a magnetometer, an orientation sensor, a proximity sensor, a temperature sensor, a time sensor; a geolocation sensor, an altimeter, a light sensor, a touch sensor, a magnetometer, a barometer, a gesture sensor, a biometric sensor, a humidity sensor, etc.

[0048] The WTRU 102 may include a full-duplex radio for which some or all signals may be transmitted and received (e.g., transmit and receive). For example, subframes associated with specific subframes for both UL (e.g., for transmission) and DL (e.g., for reception) may be concurrent and / or simultaneous. The full-duplex radio may include an interference management unit to reduce and / or substantially eliminate self-interference via hardware (e.g., choke) or via signal processing by a processor (e.g., a separate processor (not shown) or via processor 118). In one embodiment, the WTRU 102 may include a half-duplex radio for which some or all signals may be transmitted and received (e.g., associated with specific subframes for UL (e.g., for transmission) or DL ​​(e.g., for reception).

[0049] Figure 1C 1 is a system diagram showing the RAN 104 and the CN 106 according to an embodiment. As described above, the RAN 104 may employ an E-UTRA radio technology to communicate with the WTRUs 102a, 102b, 102c over the air interface 116. The RAN 104 may also be in communication with the CN 106.

[0050] The RAN 104 may include eNode-Bs 160a, 160b, 160c, though it will be appreciated that the RAN 104 may include any number of eNode-Bs while remaining consistent with an embodiment. 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, for example, the eNode-B 160a may use multiple antennas to transmit wireless signals to and / or receive wireless signals from the WTRU 102a.

[0051] 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, user scheduling in UL and / or DL, etc. Figure 1C As shown, eNode-Bs 160a, 160b, 160c may communicate with each other via an X2 interface.

[0052] Figure 1C The illustrated CN 106 may include a mobility management entity (MME) 162, a serving gateway (SGW) 164, and a packet data network (PDN) gateway (PGW) 166. Although the foregoing elements are depicted as part of the CN 106, it should be appreciated that any of these elements may be owned and / or operated by an entity other than the CN operator.

[0053] The MME 162 may be connected to each of the eNode-Bs 162a, 162b, 162c in the RAN 104 via an S1 interface and may serve 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 an initial attach of the WTRUs 102a, 102b, 102c, and the like. The MME 162 may provide a control plane function for switching between the RAN 104 and other RANs (not shown) that employ other radio technologies, such as GSM and / or WCDMA.

[0054] The SGW 164 may be connected to each of the eNode-Bs 160a, 160b, 160c in the RAN 104 via an S1 interface. The SGW 164 may generally route and forward user data packets to / from the WTRUs 102a, 102b, 102c. The SGW 164 may perform other functions such as anchoring the user plane during inter-eNodeB handovers, triggering paging when DL data is available for the WTRUs 102a, 102b, 102c, managing and storing the context of the WTRUs 102a, 102b, 102c, and the like.

[0055] The SGW 164 may be connected to the PGW 166, which may provide the WTRUs 102a, 102b, 102c with access to packet-switched networks, such as the Internet 110, to facilitate communications between the WTRUs 102a, 102b, 102c and IP-enabled devices.

[0056] The CN 106 may facilitate communications with other networks. For example, the CN 106 may provide the WTRUs 102a, 102b, 102c with access to circuit-switched networks such as the PSTN 108, in order to facilitate communications between the WTRUs 102a, 102b, 102c and traditional land-line communications devices. For example, the CN 106 may include or communicate with an IP gateway (e.g., an IP Multimedia Subsystem (IMS) server) that serves as an interface between the CN 106 and the PSTN 108. In addition, the CN 106 may provide the WTRUs 102a, 102b, 102c with access to other networks 112, which may include other wired and / or wireless networks owned and / or operated by other service providers.

[0057] Although the WTRU Figures 1A-1D Although described as a wireless terminal, it is contemplated that in certain representative embodiments, such a terminal may use (eg, temporarily or permanently) a wired communication interface with a communication network.

[0058] In a representative embodiment, other network 112 may be a WLAN.

[0059] A WLAN using an infrastructure basic service set (BSS) mode may have an access point (AP) for a BSS and one or more stations (STAs) associated with the AP. The AP may have access or an interface to a distribution system (DS) or another type of wired / wireless network that sends traffic into and / or out of the BSS. Traffic originating from outside the BSS to STAs may arrive through the AP and may be delivered to the STAs. Traffic originating from STAs to a destination outside the BSS may be sent to the AP to be delivered to the corresponding destination. Traffic between STAs within the BSS may be sent through the AP, for example, where the source STA may send traffic to the AP, and the AP may deliver the traffic to the destination STA. Traffic between STAs within the BSS may be considered and / or referred to as peer-to-peer traffic. Peer-to-peer traffic may be sent between source STAs and destination STAs (e.g., directly between source STAs and destination STAs) using direct link establishment (DLS). In certain representative embodiments, the DLS may use 802.11eDLS or 802.11z tunnel DLS (TDLS). A WLAN using an independent BSS (IBSS) mode may not have an AP, and STAs (eg, all STAs) within or using the IBSS may communicate directly with each other. The IBSS communication mode may sometimes be referred to herein as an "ad-hoc" communication mode.

[0060] When using the 802.11ac infrastructure mode of operation or a similar mode of operation, the AP can send beacons on a fixed channel (e.g., a primary channel). The primary channel can be a fixed width (e.g., a 20MHz wide bandwidth) or a dynamically set width. The primary channel can be an operating channel of the BSS and can be used by STAs to establish a connection with the AP. In certain representative embodiments, carrier sense multiple access with collision avoidance (CSMA / CA) can be implemented, for example, in an 802.11 system. For CSMA / CA, STAs (e.g., each STA) (including the AP) can listen to the main channel. If the main channel is sensed / detected and / or determined to be busy by a particular STA, the particular STA can back off. One STA (e.g., only one station) can transmit at any given time in a given BSS.

[0061] High throughput (HT) STAs may communicate using a 40 MHz wide channel, for example, by combining a 20 MHz wide main channel with an adjacent or non-adjacent 20 MHz wide channel to form a 40 MHz wide channel.

[0062] Very high throughput (VHT) STAs can support 20MHz, 40MHz, 80MHz and / or 160MHz wide channels. 40MHz and / or 80MHz channels can be formed by combining continuous 20MHz channels. A 160MHz channel can be formed by combining 8 continuous 20MHz channels or by combining two non-continuous 80MHz channels, which can be referred to as an 80+80 configuration. For the 80+80 configuration, the data after channel coding can pass through a segment parser, which can divide the data into two streams. Inverse Fast Fourier Transform (IFFT) processing and time domain processing can be performed on each stream separately. The stream can be mapped to two 80MHz channels, and the data can be sent by the transmitting STA. At the receiver of the receiving STA, the above operations for the 80+80 configuration can be reversed, and the combined data can be sent to the medium access control (MAC).

[0063] The sub-1 GHz operating mode is supported by 802.11af and 802.11ah. The channel operating bandwidth and carrier in 802.11af and 802.11ah are reduced relative to the channel operating bandwidth and carrier used in 802.11n and 802.11ac. 802.11af supports 5MHz, 10MHz and 20MHz bandwidths in the TV White Space (TVWS) spectrum, and 802.11ah supports 1MHz, 2MHz, 4MHz, 8MHz and 16MHz bandwidths using non-TVWS spectrum. According to a representative embodiment, 802.11ah can support meter type control / machine type communication (MTC), such as MTC devices in macro coverage areas. MTC devices may have certain capabilities, such as limited capabilities including support for (e.g., only support for) certain and / or limited bandwidths. MTC devices may include batteries with battery life above a threshold (e.g., to maintain very long battery life).

[0064] WLAN systems (e.g., 802.11n, 802.11ac, 802.11af, and 802.11ah) that can support multiple channels and channel bandwidths include channels that can be designated as primary channels. The primary channel may have a bandwidth equal to the maximum common operating bandwidth supported by all STAs in the BSS. The bandwidth of the primary channel may be set and / or limited by STAs from all STAs operating in the BSS that support the minimum bandwidth operating mode. In the example of 802.11ah, for STAs (e.g., MTC type devices) that support (e.g., only support) 1MHz mode, the primary channel may be 1MHz wide, even if the AP and other STAs in the BSS support 2MHz, 4MHz, 8MHz, 16MHz, and / or other channel bandwidth operating modes. Carrier sensing and / or network allocation vector (NAV) settings may depend on the state of the primary channel. If the primary channel is busy, for example, because a STA (which only supports the 1MHz operating mode) sends to the AP, all available bands may be considered busy even if most of the available bands remain idle.

[0065] In the United States, the available frequency band that 802.11ah can use is from 902MHz to 928MHz. In South 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.

[0066] Figure 1D1 is a system diagram showing the RAN 104 and the CN 106 according to an embodiment. As described above, the RAN 104 may employ NR radio technology to communicate with the WTRUs 102a, 102b, 102c over the air interface 116. The RAN 104 may also be in communication with the CN 106.

[0067] The RAN 104 may include gNBs 180a, 180b, 180c, though it will be appreciated that the RAN 104 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, for example, the gNB 180a may use multiple antennas to transmit wireless signals to and / or receive wireless signals from the WTRU 102a. In an embodiment, the gNBs 180a, 180b, 180c may implement carrier aggregation techniques. For example, the gNB 180a may send multiple component carriers (not shown) to the WTRU 102a. A subset of these component carriers may be on an unlicensed spectrum, while the remaining component carriers may be on a licensed spectrum. In an embodiment, the gNBs 180a, 180b, 180c may implement coordinated multi-point (CoMP) techniques. For example, the WTRU 102a may receive coordinated transmissions from the gNB 180a and gNB 180b (and / or gNB 180c).

[0068] The WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c using transmissions associated with scalable parameter configurations (numerology). For example, OFDM symbol spacing and / or OFDM subcarrier spacing may vary for different transmissions, different cells, and / or different portions of the wireless transmission spectrum. The WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c using subframes or transmission time intervals (TTIs) of varying or scalable lengths (e.g., containing varying numbers of OFDM symbols and / or varying absolute time lengths over time).

[0069] The gNBs 180a, 180b, 180c may be configured to communicate with the WTRUs 102a, 102b, 102c in a standalone configuration and / or a non-standalone configuration. In a standalone configuration, the WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c without accessing other RANs (e.g., such as eNode-Bs 160a, 160b, 160c). In a standalone configuration, the WTRUs 102a, 102b, 102c may utilize one or more of the gNBs 180a, 180b, 180c as mobility anchors. In a standalone configuration, the WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c using signals in an unlicensed band. In a non-standalone configuration, the WTRUs 102a, 102b, 102c may communicate / connect with the gNBs 180a, 180b, 180c while also communicating / connecting with another RAN, such as an eNode-B 160a, 160b, 160c. For example, the 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 a non-standalone configuration, the eNode-Bs 160a, 160b, 160c may serve as mobility anchors for the WTRUs 102a, 102b, 102c, and the gNBs 180a, 180b, 180c may provide additional coverage and / or throughput for serving the WTRUs 102a, 102b, 102c.

[0070] Each of the 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, user scheduling in UL and / or DL, support network slicing, interworking between DC, NR and E-UTRA, routing of user plane data to a user plane function (UPF) 184a, 184b, routing of control plane information to an access and mobility management function (AMF) 182a, 182b, and the like. Figure 1D As shown, gNBs 180a, 180b, 180c can communicate with each other via the Xn interface.

[0071] Figure 1DThe illustrated CN 106 may include at least one AMF 182a, 182b, at least one UPF 184a, 184b, at least one session management function (SMF) 183a, 183b, and may include a data network (DN) 185a, 185b. Although the foregoing elements are depicted as part of the CN 106, it should be appreciated that any of these elements may be owned and / or operated by an entity other than the CN operator.

[0072] The AMF 182a, 182b may be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 104 via the N2 interface and may serve 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 protocol data unit (PDU) sessions with different requirements), selecting a specific SMF 183a, 183b, managing registration areas, terminating non-access stratum (NAS) signaling, mobility management, etc. The AMF 182a, 182b may use network slicing to customize CN support for the WTRUs 102a, 102b, 102c based on the types of services used by the WTRUs 102a, 102b, 102c. For example, different network slices can 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 MTC access, etc. The AMF 182a, 182b may provide a control plane function for switching between the RAN 104 and other RANs (not shown) that employ other radio technologies, such as LTE, LTE-A, LTE-APro, and / or non-3GPP access technologies such as WiFi.

[0073] The SMF 183a, 183b may be connected to the AMF 182a, 182b in the CN 106 via the N11 interface. The SMF 183a, 183b may also be connected to the UPF 184a, 184b in the CN 106 via the N4 interface. The SMF 183a, 183b may select and control the UPF 184a, 184b, and configure the routing of services through the UPF 184a, 184b. 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 DL data notifications, etc. The PDU session type may be IP-based, non-IP-based, Ethernet-based, etc.

[0074] The UPF 184a, 184b may be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 104 via an N3 interface, which may provide the WTRUs 102a, 102b, 102c with access to packet-switched networks such as the Internet 110 to facilitate communications between the WTRUs 102a, 102b, 102c and IP-enabled devices. The UPF 184, 184b may perform other functions, such as routing and forwarding packets, enforcing user plane policies, supporting multi-homed PDU sessions, handling user plane QoS, buffering DL packets, providing mobility anchoring, etc.

[0075] The CN 106 may facilitate communications with other networks. For example, the CN 106 may include or communicate with an IP gateway (e.g., an IP Multimedia Subsystem (IMS) server) that serves as an interface between the CN 106 and the PSTN 108. In addition, the CN 106 may provide the WTRUs 102a, 102b, 102c with access to other networks 112, which may include other wired and / or wireless networks owned and / or operated by other service providers. In one embodiment, the WTRUs 102a, 102b, 102c may be connected to the local DN 185a, 185b through the UPF 184a, 184b via the N3 interface to the UPF 184a, 184b and the N6 interface between the UPF 184a, 184b and the DN 185a, 185b.

[0076] In view of Figures 1A-1D and about Figures 1A-1D As described herein, one or more or all of the functions described herein for one or more of the following may be performed by one or more emulation devices (not shown): WTRU 102a-d, base station 114a-b, eNode-B 160a-c, MME 162, SGW 164, PGW 166, gNB 180a-c, AMF 182a-b, UPF 184a-b, SMF 183a-b, DN 185a-b, and / or any other device(s) described herein. An emulation device may be one or more devices configured to emulate one or more or all of the functions described herein. For example, an emulation device may be used to test other devices and / or simulate network and / or WTRU functions.

[0077] The simulation device may be designed to implement one or more tests of other devices in a laboratory environment and / or an operator network environment. For example, one or more simulation devices may 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 simulation devices may perform one or more or all functions while being temporarily implemented / deployed as part of a wired and / or wireless communication network. A simulation device may be directly coupled to another device in order to conduct tests and / or perform tests using over-the-air wireless communications.

[0078] The one or more simulation devices can perform one or more functions, including all functions, while not being implemented / deployed as part of a wired and / or wireless communication network. For example, the simulation device can be used in a test scenario in a test lab and / or a non-deployed (e.g., testing) wired and / or wireless communication network to implement testing of one or more components. The one or more simulation devices can be test devices. The simulation device can send and / or receive data using direct RF coupling and / or wireless communication via RF circuits (e.g., which can include one or more antennas).

[0079] In view of Figures 1A-1D and about Figures 1A-1D In accordance with the corresponding descriptions of the present invention, one or more or all of the functions described herein for one or more of the WTRUs 102a-d, base stations 114a-b, eNode-Bs 160a-c, MMEs 162, SGWs 164, PGWs 166, gNBs 180a-c, AMFs 182a-b, UPFs 184a-b, SMFs 183a-b, DNs 185a-b, and / or any other devices described herein may incorporate artificial intelligence (AI) and / or machine learning (ML) into the corresponding operations. Artificial intelligence may be broadly defined as behavior exhibited by a machine, where the behavior may mimic cognitive functions to perceive, reason, adapt, and / or act. For example, wireless systems may leverage AI to reduce the overhead of transmissions (e.g., an AI / ML model may be used to generate a value using one or more inputs, where transmissions may be reduced by only requiring the one or more inputs to be sent, and the model may be used at the receiver (e.g., already known or indicated in some way) to generate the value such that the sender will effectively share the value with the receiver even though the value itself is not sent).

[0080] Machine learning refers to a type of algorithm that solves problems based on learning through experience ("data") without being explicitly or minimally programmed ("configured with a set of rules"). Machine learning can be considered a subset of AI. Different machine learning paradigms can be envisioned based on the nature of the data or feedback available to the learning algorithm. For example, supervised learning methods can involve learning a function that maps inputs to outputs based on labeled training examples, where each training example can be a pair consisting of an input and a corresponding output. For example, unsupervised learning methods can involve detecting patterns in the data that do not have pre-existing labels. For example, reinforcement learning methods can involve performing a series of actions in an environment to maximize cumulative rewards. In some cases, machine learning algorithms can be applied using a combination or interpolation of the above methods. For example, semi-supervised learning methods can use a combination of a small amount of labeled data with a large amount of unlabeled data during training. In this regard, semi-supervised learning falls between unsupervised learning (no labeled training data) and supervised learning (only labeled training data).

[0081] Deep learning refers to a class of machine learning algorithms that employ artificial neural networks (e.g., in particular deep neural networks DNNs), which can be loosely inspired from biological systems. Deep neural networks (DNNs) are a special class of machine learning models inspired by the human brain, in which the input is linearly transformed and a nonlinear activation function is passed multiple times. DNNs typically include multiple layers, each of which consists of a linear transformation and a given nonlinear activation function. DNNs can be trained using training data via a backpropagation algorithm. Recently, DNNs have shown state-of-the-art performance in various fields such as speech, vision, natural language, and in various machine learning settings for supervision, unsupervised, and semi-supervised. The term AI / ML-based method / processing can refer to achieving behavior and / or compliance with requirements through data-based learning without explicitly configuring the sequence of steps for the action, and / or having a configured sequence of actions or a combination of steps in addition to the learned steps or action sequences. Such methods can enable learning of complex behaviors that may be difficult to specify and / or implement when using traditional methods.

[0082] Artificial intelligence (AI) including machine learning (ML) can be applied to wireless transmitters and / or wireless receivers. AI / ML can be used to improve one or more specific aspects, functions, or protocol operations of a wireless node, such as as a local optimization within a node and / or as part of a function or process on the air interface (AI-AI).

[0083] In some cases, AI / ML based approaches can be used for high-resolution CSI feedback and better performance with reduced CSI-RS overhead. An autoencoder can be one of the AI / ML model architectures for CSI compression. The autoencoder architecture consists of two parts: an encoder AI model (e.g., at the WTRU) and a decoder AI model (e.g., at the base station), both of which can be jointly trained / designed. The encoder / decoder can have multiple layers. In the encoder, the output of each layer is smaller than its input (e.g., for the decoder, and vice versa). In some instances, a larger number of layers may result in higher model complexity, which may result in higher storage requirements, which may result in higher training complexity, which may result in better performance / better compression ratio.

[0084] The channel state information (CSI) may include at least one of the following: a channel quality index (CQI), a rank indicator (RI), a precoding matrix index (PMI), an L1 channel measurement (e.g., RSRP, such as L1-RSRP or SINR), a CSI-RS resource indicator (CRI), a SS / PBCH block resource indicator (SSBRI), a layer indicator (LI) and / or any other measurement quantity measured by the WTRU from a configured reference signal (e.g., a CSI-RS or a SS / PBCH block or any other reference signal).

[0085] The WTRU may be configured to report CSI through the uplink control channel on the PUCCH or upon request of the base station on the UL PUSCH grant. Depending on the configuration, the CSI-RS may cover the entire bandwidth of the bandwidth part (BWP) or only a portion of the bandwidth part (BWP). Within the CSI-RS bandwidth, the CSI-RS may be configured in every PRB or every other PRB. In the time domain, the CSI-RS resources may be configured as periodic, semi-persistent, or aperiodic. Semi-persistent CSI-RS is similar to periodic CSI-RS, except that the resource may be activated (deactivated) by a MAC CE; and the WTRU may report related measurements only when the resource is activated. For aperiodic CSI-RS, the WTRU may report the measured CSI-RS on the PUSCH triggered by a request in the DCI. Periodic reports are carried on the PUCCH, while semi-persistent reports may be carried on either the PUCCH or the PUSCH. The scheduler may use the reported CSI when allocating the best resource blocks, determining the precoding matrix, beam, transmission mode, and / or selecting the appropriate MCS, possibly based on the time-frequency selectivity of the channel. The reliability, accuracy, and timeliness of WTRU CSI reporting may be critical to meeting URLLC service requirements.

[0086] Figure 2An example of an AI / ML model enhanced exchange between a WTRU and a base station is shown. As shown, there may be a WTRU 202 and a base station 204. At some initial time (e.g., t=0), the base station may send a reference signal set at 211. The WTRU may measure / collect CSI for the reference signals. The WTRU 202 may have an AI-based model that has a potential vector (z) as its input, which, when input into the AI-based model, generates CSI (e.g., as close or identical to the original measured CSI as possible). At (A), the WTRU may use the reference signals to optimize for z. At (B), the WTRU may send z (e.g., which may be considered as compressed CSI). At 212, the base station receives z and may use it to determine the CSI. In some instances, the WTRU may send additional information along with z. At (C), the WTRU may optimize the optimal number of reference signals required and their locations in the grid. At 213, the WTRU may send this information to the base station. At some point after or during the initial exchange (e.g., at or after t=1), the base station may send a reduced set of reference signals at 214 (e.g., while sending z or while sending information about the reduced set, as further described herein).

[0087] Figure 3 An example of a CSI measurement setup is shown. A WTRU may be configured with a CSI measurement setup that may include one or more CSI report setups (e.g., 300 and 301), resource setups (e.g., 310, 311, 312), and / or a link between one or more CSI report setups and one or more resource setups (e.g., 320, 321, 322, 323). As shown in this example, there are multiple possible configurations of CSI report setups, resource setups, and links. For example, there may be resource setups for non-zero power CSI reference signals (NZP CSI RS) and zero power CSI RS (ZP CSI RS).

[0088] In the CSI measurement setting, one or more configuration parameters may be provided.

[0089] For example, one configuration parameter may be N≥1 CSI report settings, M≥1 resource settings, and a CSI measurement setting linking the N CSI report settings with the M resource settings.

[0090] For example, a configuration parameter may be a CSI reporting setting that includes at least one of: time domain behavior; aperiodic or periodic / semi-persistent; frequency granularity, at least for PMI and CQI; CSI reporting type (e.g., PMI, CQI, RI, CRI, etc.); and / or, if PMI is reported, the PMI type (Type I or II) and codebook configuration.

[0091] For example, a configuration parameter may be a resource setting including at least one of: time domain behavior; aperiodic or periodic / semi-persistent; RS type (e.g., for channel measurement or interference measurement); and / or, S≥1 resource sets (S), where each resource set may contain Ks resources.

[0092] For example, a configuration parameter may be a CSI measurement setting, which includes at least one of: a CSI report setting; a resource setting; and / or for CQI, a reference transmission scheme setting.

[0093] For example, one configuration parameter may be CSI reporting of component carriers at one or more of the following frequency granularities that may be supported: wideband CSI; fractional-band CSI; and / or sub-band CSI.

[0094] Figure 4 An example of codebook-based precoding with feedback information is shown. For a MIMO scenario, as shown, there are transmitters 912 and receivers 916, each with multiple antennas, demonstrating the concept of codebook-based precoding with feedback information (e.g., the feedback shown at 904). The feedback information may include a precoding matrix index (PMI), which may be referred to as a codeword index in a codebook, as shown.

[0095] like Figure 4 As shown in the example of , the codebook may include a set of precoding vectors / matrices for each rank and number of antenna ports, and each precoding vector / matrix has its own index, so that the receiver 916 can inform the transmitter 912 of the preferred precoding vector / matrix index. Compared with non-codebook based precoding, codebook based precoding may have performance degradation due to its limited number of precoding vectors / matrices, however, the main advantage of codebook based precoding may be lower control signaling / feedback overhead.

[0096] Table 2 shows an example of a codebook for 2Tx.

[0097]

[0098]

[0099] Table 2

[0100] A CSI processing unit (CSIU) may be referred to as the smallest CSI processing unit, and a WTRU may support one or more CSI processing units (e.g., N CSIs). A WTRU with N CSIs may estimate N CSI feedback calculations in parallel, where N may be the WTRU capability. If a WTRU is requested to estimate more than N CSI feedbacks simultaneously, then the WTRU may perform only the high priority N CSI feedbacks, and the rest may not be estimated.

[0101] The start and end of the CSIU can be determined based on the CSI report type (e.g., aperiodic, periodic, semi-persistent) as follows: for aperiodic CSI reporting, the CSIU is occupied from the first OFDM symbol after the PDCCH trigger until the last OFDM symbol of the PUSCH carrying the CSI report; and / or, for periodic and semi-persistent CSI reporting, the CSIU is occupied from the first OFDM symbol of one or more associated measurement resources (no earlier than the CSI reference resource) until the last OFDM symbol of the CSI report.

[0102] The number of occupied CSIUs may differ based on the CSI measurement type (e.g., beam-based or non-beam-based), as follows: non-beam-related reporting, where Ks CSIUs are used when there are CSI-RS resources in Ks CSI-RS resource sets for channel measurement; beam-related reporting (e.g., "cri-RSRP", "ssb-Index-RSRP", or "None"), where 1 CSIU is independent of the number of CSI-RS resources in the CSI-RS resource set used for channel measurement due to lower CSI computation complexity, or "None" for P3 operation or non-periodic TRS transmission; for non-periodic CSI reporting with a single CSI-RS resource, where 1 CSIU is occupied; and / or, for CSI reporting Ks CSI-RS resources, where Ks CSIUs are occupied because the WTRU needs to perform CSI measurement for each CSI-RS resource.

[0103] When the number of unoccupied CSIUs (Nu) is less than the required CSIUs (Nr) for CSI reporting, the following WTRU behavior may be used: the WTRU may discard the NR-nu CSI report based on priority in the case of UCI on PUSCH without data / HARQ; and / or, the WTRU may report virtual information in the NR-NU CSI report based on priority in other cases to avoid rate matching processing for PUSCH.

[0104] Artificial intelligence (AI) including machine learning (ML) can be applied to wireless transmitters and / or wireless receivers. AI / ML can be used to improve one or more specific aspects, functions, or protocol operations of a wireless node, such as as a local optimization within the node and / or as part of a function or process over the air interface (AI-AI).

[0105] In one or more embodiments, there may be generative models derived from an autoencoder-based framework for channel estimation, CSI compression, and / or RS overhead reduction. These techniques may be directed to one or more of the following: minimizing the impact of the specification and having a WTRU vendor-friendly approach (e.g., only the decoder needs to be specified, no encoder is required); allowing multiple functions to be completed simultaneously (e.g., CSI compression, channel estimation, identification of optimal pilot density + pilot position, etc.); and / or improving the accuracy of channel estimation and compression (e.g., adding supplementary prior information for improved reconstruction, and / or the supplementary information may also be added as a regularizer, where a regularizer refers to an additional component added to a loss function or optimization problem, and the regularizer may represent some prior information, a penalty, or an additional constraint).

[0106] For generative model training, the entire framework can use a generative model (also referred to as a decoder model) that takes a latent vector (also referred to as a compressed representation or Z in this article) as input and generates an estimate of the channel matrix H2. The generative model can be trained as an autoencoder or trained using a generative adversarial network.

[0107] For autoencoder style training, the generator neural network model / decoder model can be trained in tandem with the encoder neural network model. During the training process, the channel matrix H from the training dataset is fed as input to the encoder model, φ1(·) This will produce a compressed representation Z = φ1(H) which in turn is used as input to the decoder model, φ2(·). Output H2 = φ1(Z) In some embodiments, the estimated value of the decoder model is used to evaluate the loss and train the encoder model and the decoder model.

[0108] An example of a training loss is the mean squared error between the input channel matrix H and the decoder output H2 and can be expressed as: Among them, ||·|| p refers to the p-norm operator.

[0109] Figure 5An example of autoencoder training for channel matrix compression is shown. As shown, the encoder 504 can accept a channel matrix H 502 as input, which produces a latent vector (z) 505. The latent vector 505 can then be used as an input to a decoder model 506 to produce an H2 channel estimate 508. In this example, the training loss can be minimized, as shown at 510. After training (not shown), the encoder model φ1(·) can be discarded and only the decoder model φ2(·) can be used for further processing. The same decoder model can be deployed at both the WTRU and the base station and can be used for the dual purposes of channel estimation and channel compression. For reference, the equation shown at 510 in the figure is reproduced here:

[0110] During the inference or operation phase, the WTRU receives CSI-RS at pre-configured REs and utilizes some or all of them to optimize for latent Z. Thus obtaining a latent value.

[0111] Figure 6 An example of MIMO channel estimation using a decoder / generative model is shown. As shown, the entire process can be represented by several stages / steps. Initially, at 602, there can be autoencoder training. At 604, the encoder can be removed. At 606, the decoder can be frozen, which refers to freezing the weights of the decoder network so that the latent vector can be optimized, at 608. 610 shows the training loss of a function of the latent vector, where the sampling operator Ω can indicate the CSI-RS that can be used for estimation. For reference, the equation shown at 610 in the figure is reproduced here:

[0112] The generative model or decoder model φ2(·) can also be learned as a generative adversarial network (GAN) or a variational autoencoder (VAE) to map the latent representation or compressed representation Z to the estimated channel matrix H2. As a GAN, the model is learned specifically in a generator-discriminator setting using an adversarial loss. In such a setting, the generator model aims to produce a channel matrix that is "similar" to the channel matrix available in the training dataset, and the discriminator aims to distinguish whether the input channel matrix is ​​obtained from the training dataset or generated using the generative model.

[0113] During the inference or operational phase, the trained model may be utilized similarly to a generative model trained in an autoencoder setting. The WTRU receives CSI-RS at pre-configured REs and utilizes some or all of them to optimize for potential Z. When the potential value is obtained.

[0114] For a generative model trained as a GAN, the model can be trained to utilize some prior information or characteristics of the channel as additional conditional inputs to the generator model. For example, if the rank of the matrix is ​​known in advance, the GAN can be conditioned to produce a channel matrix with a specified rank. Any other properties or information of the channel matrix that may have an impact on the channel estimation and is known in advance can also be used for channel estimation.

[0115] In such a setting, the training process of the model must be modified to include and ensure that the conditional argument is satisfied. The input to the generator can use two components, one component is similar to the latent representation Z, and in addition, the second component indicates the prior knowledge that the model needs to be conditioned on. During training, the standard adversarial loss is supplemented with a secondary loss function to minimize the error between the previously known attributes or channel characteristics and the measurements of the attributes or channel characteristics from the estimated channel generation channel H2. A detailed discussion on the potential impact of having conditional generative models for channel estimation and CSI feedback is further discussed in this paper.

[0116] The WTRU may be configured to determine a channel estimate and report CSI feedback including a latent / compressed / low dimensional vector z that is compatible with a pre-configured decoder AI / ML model. The decoder model may be, for example, the decoder portion of a trained autoencoder model or a separately trained GAN, as discussed herein. The signal latent vector z may then be used at the base station to reconstruct the CSI (or its quantity) using the decoder AI / ML model. The reconstructed CSI, H2, may have one or more representations.

[0117] For example, the reconstructed CSI can be expressed as N dimensions. r ×N t ×N SB Full-bandwidth CSI, where N r ,N t , and N SB They respectively represent the number of receive antenna ports, the number of transmit antenna ports, and the number of configured subbands.

[0118] For example, the reconstructed CSI may represent the channel measured / estimated at some resource elements (e.g., pre-configured CSI-RS positions), where the dimension of the reconstructed CSI is N r ×N t ×N res , where N res The number of reserved REs configured for CSI-RS reception or a part thereof may be indicated.

[0119] For example, the reconstructed CSI may represent the received signal or (a portion thereof) at the configured CSI-RS position, where the size of the reconstructed CSI is: N totalCSIRSThe number of measurements at the configured CSI-RS positions.

[0120] In some cases, the WTRU may be configured to derive a latent vector z based on one or more decoder AI / ML model configurations. The WTRU may be configured to determine the latent vector z based on a specific pre-configured decoder AI / ML model mirrored at two communication nodes. The WTRU may optimize the low-dimensional latent vector z using the configured decoder AI / ML model mapping function φ2(.). The latent vector may also be used for the dual purpose of channel estimation and CSI feedback. At the WTRU, the latent vector z may be used together with the decoder model to estimate the channel matrix. In addition, the latent vector z may be signaled to the base station and may be used to reconstruct the channel at the base station side.

[0121] In another option, the WTRU may be configured to select from a plurality of pre-configured decoder models. The selected decoder can satisfy a preconfigured reconstruction error threshold ∈ err (e.g., a preconfigured NMSE). For example, the WTRU may select the decoder ∈ that produces a reconstruction error closest to a preconfigured threshold err If configured to select one of multiple (N) decoder AI / ML models, the WTRU may indicate the selected model index along with the associated latent vector to the base station.

[0122] The quality of the CSI reconstruction at the base station depends on the degree of optimization of the latent vector z. One aspect that affects the quality of the optimized latent vector z is the choice of the loss function. The choice of the loss function can ensure a high-quality method for the latent vector z, which can produce acceptable CSI reconstruction at the base station. The WTRU can be configured to determine the latent vector z that is optimized to minimize a pre-configured loss function L. The WTRU can be configured with one of multiple loss functions for finding the latent vector z. For example, the loss function can be the normalized mean square error (NMSE) or the cosine similarity (CS), or it can be directly designed to maximize the system throughput. The loss function can be configured to have two loss components L1 and L2, such that L is a weighted combination of L1 and L2, such as L=L1+αL2 and α is a scaling / weighting factor.

[0123] The first loss component, L1, may represent a primary function, while the second loss component, L2, may represent a specific property of a channel. The first loss component, L1, may be one or more variants.

[0124] The first loss component, L1, may be the NMSE of the difference between H1 and the estimated channel H2: = φ2(z) at the decoder output, ie, Here, H1 is the channel obtained at the WTRU using CSI-RS.

[0125] The first loss component, L1, may be the correlation between H1 and H2: =φ2(z), ie, L1 = -Re(tr(H1*H2)), and the goal is to find z which maximizes the correlation.

[0126] The first loss component, L1, may be the throughput, i.e., Where v represents the main eigenvector φ2(z) of the reconstructed channel.

[0127] The first loss component may be selected by the WTRU, or it may be configured by the base station. For example, in the case of eigenvector precoding, the base station may configure the WTRU with a first loss component that tends to optimize the latent vector z in a way that reconstructs the channel eigenvectors rather than the channel matrix. In this case, the first loss component may be selected as Where v1 is the principal eigenvector of the estimated channel and f(.) returns the maximum eigenvector of the reconstructed channel.

[0128] The second loss component, L2, captures a specific property of the channel. An example of this property could be that the estimated channel could be "low rank" or "sparse". For example, a low rank constraint / regularization term can be implemented by using the second loss component, L2 as the kernel norm. The low rank regularizer has several advantages; among them, it can provide robustness to noise, and can effectively reduce the number of CSI-RS required to meet a specific reconstruction performance, and can improve channel estimation accuracy. On the other hand, the sparsity constraint can be implemented by using a second loss component such as L2 of l1 or using the group sparsity norm as l2. 1,2 -Norm. -The sparsity constraint may reduce the uplink overhead associated with the potential vector z. The attribute caused by the second loss component may be configured by the base station, or it may be selected by the WTRU from a plurality of pre-configured attributes. For example, the WTRU may select the second loss component based on some channel characteristics or parameters (e.g., delay spread or channel rank).

[0129] The WTRU may be configured with an α-weighted weighting parameter of L2w.r.tL1, where α may take several pre-configured values. The WTRU may determine the value of α using a brute force approach and signal back the value of α that produces the minimum reconstruction error. In another option, the WTRU may select the value of α based on channel measurements. For example, if the channel is measured as low rank, the WTRU may select a high weight for the second loss component that implements the low rank property of the channel. If the channel is not low rank, the WTRU may set the α loss component equal to zero to eliminate the effect of the second loss component.

[0130] The first loss component may be configured via RRC configuration, while the second loss component may be configured via RRC configuration and a set of preconfigured values ​​α may be configured semi-statically or dynamically (eg, via MAC CE or L1 signaling).

[0131] In some cases, CSI feedback may be optimized by using loss information and reporting latent information. The WTRU may be configured, instructed, or requested to send a CSI feedback report containing at least a latent vector z or a quantized / encoded version thereof, or a quantity thereof. The latent vector z may be referred to as a CSI reporting quantity (e.g., Δi / MLDecoderLatent). The maximum size of the latent vector z may be determined based on the capabilities of the configured uplink feedback resources, or it may be configured by a higher layer (e.g., RRC, MAC-CE). The WTRU may indicate the optimized latent vector z for channel reconstruction at the base station via uplink signaling. The report may indicate the latent vectors in one or more scenarios.

[0132] For example, in one scenario, if the WTRU is configured to select one of a plurality of pre-configured decoder AI / ML models, the WTRU may indicate the selected decoder model used to generate the latent vector z.

[0133] For example, in one scenario, if the WTRU is configured to select the second loss component L2 from a plurality of pre-configured loss functions, then the WTRU may report the index of the selected loss function L2. Different loss functions may have different adjustable parameters. In that case, the WTRU may be configured to indicate one or more of the parameters associated with the selected loss function.

[0134] For example, in one scenario, if the WTRU is configured to indicate the weight parameter α, the WTRU may indicate the selected value.

[0135] The signaling may be performed explicitly using PUCCH and / or PUSCH via modified UCI, or implicitly using specific PUCCH or RACH resources. In one option, the WTRU may be configured with UCI resources having various formats. Different formats have different numbers of bit resource blocks, etc. to indicate reporting information associated with a potential vector. For example, the trigger format may indicate the inclusion of a potential vector and a selected AI / ML decoder model. In another option, the format may indicate the inclusion of a potential vector having second loss component information or a portion thereof.

[0136] Figure 7An example of generation and transmission of compressed CSI is shown. There may be a WTRU 702 and a base station 701. The base station may configure a generator or decoder model for the WTRU, as well as one or more loss function parameters (e.g., L1 (reconstruction loss), L2 (property-based loss, "low rank", "sparseness", etc.) and a weighting function parameter α. Note that for L2, α may be dynamically configured / changed by the base station, or selected by the WTRU from available options. At 712, the base station 701 may configure the WTRU with RS sets S1 (estimation set) and S2 (maintained set). At 713, the WTRU 702 may use the RS indicated in the configuration to determine the measurement channel (H) at the RS location. At 714, the WTRU 702 may minimize the loss function L1+αL2 to determine a potential vector (e.g., compressed CSI). At 716, the WTRU 702 may send the potential vector to the base station 701. Since the WTRU All parameters are known to both 702 and base station 701, so once base station 701 has the latent vectors, it can compute the channel matrix H (e.g., measured CSI). The weights, α, may be signaled by the base station as in 711, or may be determined by the WTRU based on an estimate and a kept set, where the WTRU generates several channel estimates for different values ​​of α as follows, and selects one channel estimate from the kept set that is closest to the channel estimate at the RS location being evaluated.

[0137] In some cases, conditional compression may be used for channel estimation.In one or more examples disclosed herein, an autoencoder may be used, but the associated techniques may be applied to any generative AI / ML model or variant thereof.

[0138] The transformation from the latent space (e.g., z) to the channel matrix (e.g., H) can be expressed as a function φ2(·). Different types of AI / ML model architectures / methods can be used to learn / implement the function φ2(·). For example, the decoder part of an autoencoder can be used to learn / implement the function φ2(·). For example, the variational autoencoder (VAE) method or a variant thereof can be used to learn / implement the function φ2(·). For example, the generative adversarial network (GANs) method or a variant thereof can be used to learn / implement the function φ2(·). For example, in the GAN method, the model can be trained in a generator-discriminator setting. For example, the generator can be configured to generate channel estimation samples from the latent space, and the discriminator can be configured to distinguish the real channel estimation samples from the generated channel estimation samples. Training can be performed so that the generator can generate channel estimation samples that cannot be distinguished from the real channel estimation samples. Possibly, an adversarial loss function can be used to train the GAN model.

[0139] Conditional GANs can enable the generation of channel estimation samples with additional conditional vectors as input to an AI / ML model (e.g., as disclosed herein, for the process of training conditional GANs).

[0140] For example, the GAN AI / ML model can be adjusted to generate a channel estimate based on a conditional vector. For example, the conditional vector can be used to provide prior information that can optimize the determination of the potential vector and / or the generation of the channel estimate. For example, the conditional vector can be configured to input information related to channel characteristics / attributes / statistics. For example, the conditional vector can indicate the expected channel rank or can be derived based on the expected channel rank. For example, the conditional vector can be configured so that the AI / ML model can generate a channel matrix with a specific rank. For example, the conditional vector can indicate the number and / or gain of clusters or can be derived as a function of the number and / or gain of clusters. For example, the conditional vector can indicate the coherence time of the channel or can be derived as a function of the coherence time of the channel. For example, the conditional vector can indicate a Doppler estimate or can be derived as a function of a Doppler estimate. For example, the conditional vector can indicate the sparsity of the channel on a preconfigured basis or can be derived as a function of the sparsity of the channel. Possibly, the basis can be configured as a DFT basis. For example, the conditional vector can indicate the SNR or can be derived as a function of the SNR.

[0141] In some cases, the condition vector may be determined by the WTRU. For example, the WTRU may determine the condition vector or a portion thereof. The component of the condition vector determined by the WTRU may be referred to as Cu.

[0142] For example, the WTRU may use a previous channel estimate generated by an AI / ML model to determine part of the condition vector. For example, the WTRU may use a non-AI / ML method to determine the condition vector or part thereof. For example, the WTRU may perform channel measurements to determine the condition vector or part thereof. For example, the WTRU may be configured with a mapping between different channel ranks and values ​​of the condition vector Cu. During inference, the WTRU may be configured to set the condition vector Cu to the rank of the determined channel and a preconfigured mapping to Cu. For example, the WTRU may be configured with a mapping between different Doppler spread ranges and values ​​of the condition vector Cu. During inference, the WTRU may be configured to set the condition vector Cu to the measured Doppler spread of the channel and a preconfigured mapping to Cu. For example, the WTRU may be configured with a mapping between different SNR ranges and values ​​of the condition vector Cu. During inference, the WTRU may be configured to set the condition vector Cu based on SNR measurements and a preconfigured mapping to Cu.

[0143] In one instance, the conditional vector may explicitly or implicitly indicate aspects / attributes associated with the AI / ML model input. For example, the AI / ML model may be trained so that the model may process different bits of the latent vector differently based on the conditional vector. For example, the conditional vector may indicate the dimensions of the latent vector input to the AI / ML model. For example, this may allow for variable sizes of the latent vector. For example, the AI / ML model may be trained to consider three possible lengths of the latent vector, and the conditional vector may indicate the length of the latent vector associated during inference. In one instance, the WTRU may be configured by the base station with the length and format of the conditional vector Cu.

[0144] In some cases, the WTRU may receive a condition vector to apply to the inference signaled from the base station. For example, the base station may determine the condition vector based on one or more of the following: feedback from other WTRUs or based on deployment knowledge, etc. The value of the condition vector configured by the base station may be referred to as Cn. For example, the WTRU may obtain Cn based on signaling from the base station. For example, the signaling may be part of a CSI measurement configuration. For example, the signaling may be part of a CSI reporting configuration. For example, the WTRU may obtain Cn or part thereof from the base station as part of an aperiodic CSI request.

[0145] In some cases, the condition vector may consist of multiple parts. For example, different parts of the condition vector Ci may be associated with different channel characteristics / attributes described herein. For example, different parts of the condition vector Ci may be associated with different WTRU measurements described herein. For example, different channel attributes / WTRU measurements may include one or more of the following: channel rank, channel sparsity based on DFT, coherence time, Doppler information, number of clusters, channel gain, SNR, SINR, CQI, etc. Different condition vectors corresponding to different channel characteristics may be configured / determined, and may be cascaded to form an input condition vector. If the WTRU measurement is unknown or unavailable during the inference period, the value in Ci may be set to the previous Ci value. If the WTRU measurement is unknown or unavailable during the inference period, the value in Ci may be set to a predefined value. In one instance, different parts of the condition vector Ci may consist of a WTRU-based component Cu and a base station-configured component Cn. In one instance, the format of the condition vector Ci and the length of the different parts of Ci may be configured by the base station.

[0146] The WTRU may be configured with a conditional GAN / ML model to generate CSI feedback or part thereof. In one method, the conditional GAN / ML model may be configured to generate a channel estimate when a latent vector Z and optionally a conditional vector Ci are provided as inputs to the model. The AI / ML model may be configured and / or trained to generate different channel estimates based on the value of the conditional vector given the same latent vector. For example, the WTRU may be configured to report CSI feedback including the value of the latent vector Z. For example, the WTRU may be configured to report CSI feedback based on the value of the latent vector Z. For example, the WTRU may determine the latent vector Z so that when Z is input to the AI / ML model, a channel estimate that satisfies a preconfigured condition is generated at the output. For example, the preconfigured condition may be expressed as minimization of a loss function. For example, at least a portion of the loss function may depend on the network configuration. In one example, the loss function may be configured as the difference between a first channel estimate and a second channel estimate. For example, the first channel estimate may correspond to a channel estimate on a CSI-RS symbol. For example, the second channel estimate may be a function of the channel estimate output of the AI / ML model. For example, when the latent vector Z and / or the conditional vector Ci are input to the AI / ML model, the second channel estimate may correspond to the output of the AI / ML model. For example, when the latent vector Z and / or the conditional vector Ci are input to the AI / ML model, the second channel estimate may correspond to a sampled version of the output of the AI / ML model. In some instances, the WTRU may be configured to determine the sampling version based on a sampling operator, where the sampling operator may be pre-configured. The reconfiguration may be implicit or explicit. For example, the WTRU may implicitly determine the sampling operator as a function of the CSI-RS configuration. For example, the WTRU may explicitly determine the sampling operator based on the network configuration.

[0147] In some cases, the WTRU may input the condition vector Ci when configured by the network. In one case, the condition vector Ci may be equal to the condition vector Cn, where the condition vector Cn is configured by the network. In another case, the condition vector Ci may be equal to the condition vector Cu, where the condition vector Cu may be determined by the WTRU. In another case, the input condition vector Ci may be a combination of Cu and Cn.

[0148] In one case, the WTRU may be configured to enter a default condition vector Ci when one or more of the following conditions are true: when the WTRU is not configured to Cn and / or when the WTRU is not configured to or cannot determine Cu. In one case, the default condition vector Ci may be a zero vector. In another case, the default condition vector may be predefined to a specific value.

[0149] When the optimal Z is input to the conditional GAN ​​model conditioned on the conditional vector Ci, the WTRU may determine a latent vector Z that minimizes the loss function associated with the channel estimate. In some embodiments, both Z and Ci may be configured as inputs to the AI / ML model, and the channel estimate is the output of the AI / ML model. For example, the latent vector Z may be used as a compressed representation of the channel estimate.

[0150] The WTRU may be configured to send CSI feedback containing at least a portion of the potential vector Z and the conditional vector Ci. For example, the WTRU may be configured to feedback the Cu portion of the conditional vector Ci if Ci consists of Cn and Cu. For example, if Ci consists of only Cu, the WTRU may be configured to feedback the entire Ci. For example, if Ci consists of only Cn, the WTRU may be configured to skip the transmission of Ci. In some instances, the WTRU may be configured to send a quantized version of Ci in the CSI feedback. For example, the conditional vector Ci may serve as side information to help recover / reconstruct channel estimates based on the potential vector Z.

[0151] In one case, the WTRU may send both the potential vector Z and the conditional Ci in the same CSI feedback. For example, the first part of the CSI feedback may contain the potential vector Z and the second part of the CSI feedback may contain the conditional vector Ci. In another case, the WTRU may be configured with a mapping between Ci or part thereof and UL resources. The WTRU may implicitly indicate the value of Ci by sending the CSI feedback in the associated pre-configured UL resources (e.g., PUCCH resources). In some instances, the WTRU may implicitly indicate the value of Ci via the number of CSI reports, such as rank indication, SINR, etc.

[0152] In one case, the WTRU may send the potential vector Z and the conditional vector Ci in different messages. The WTRU may send the potential vector Z in a first type of CSI feedback and send the conditional vector Ci in a second type of CSI feedback message. The periodicity of the potential vector Z feedback and the periodicity of the conditional vector Ci feedback may be configured with different values. For example, the WTRU may be configured to report the conditional vector Ci in a semi-persistent CSI feedback. For example, the WTRU may be configured to report the conditional vector Ci only if the conditional vector Ci has changed from a previous report.

[0153] In some cases, the WTRU may be configured with one or more resource sets, which may be used for at least one of channel observation, inference time model selection, model (re)tuning, reference signal overhead reduction, transmission learning and / or online learning, where the resource set may be used interchangeably with set, observation set, channel matrix estimation set, channel observation set, channel estimation set, measurement resource set, data set, estimation set, hold set, training set, test set and / or validation set.

[0154] In one example, the WTRU may be configured with or determined to use two resource sets, where a first resource set may be a channel observation set that may be used by one or more channel estimation models to estimate a channel matrix, and a second resource set may be a holding set that may be used to determine, estimate, and / or monitor estimation performance (e.g., channel estimation performance from one or more channel estimation models).

[0155] The first resource set (e.g., channel observation set) can be configured with one or more measurement reference signals (e.g., CSI-RS, NZP-CSI-RS, TRS, SSB), and the second resource set (e.g., retention set) can be configured with one or more resources carrying data (e.g., PDSCH RE, PDCCH RE); alternatively, the second resource set can be configured with a demodulation reference signal (e.g., DM-RS).

[0156] The first resource set may be configured with one or more sounding reference signals having a higher density (eg, 3 RE / RB), while the second resource set may be configured with one or more sounding reference signals having a lower density (eg, 1 RE / RB).

[0157] The first resource set may be configured with one or more measurement reference signals having a specific periodicity (e.g., periodic CSI-RS, SSB), and the second resource set may be configured with one or more measurement reference signals that may be sent aperiodically or semi-persistently (e.g., aperiodic CSI-RS, semi-persistent CSI-RS).

[0158] The first set of resources may be configured via a higher layer (eg, RRC, MAC-CE), and the second set of resources may be dynamically indicated via L1 signaling (eg, DCI).

[0159] The presence of the second set of resources in the timeslot may be indicated by the DCI, and the WTRU may perform or determine the execution of the estimation model when the WTRU is indicated the presence of the second set of resources in the timeslot.

[0160] The presence of the second resource set may be implicitly indicated by a trigger signal for determining the performance of the estimation model (eg, the base station may trigger execution of a performance check of the estimation model).

[0161] The first resource set may be configured by the base station and the second resource set may be determined by the WTRU. For example, the first resource set may be one or more configured measurement reference signals and the second resource set may be determined by the WTRU within candidate resources, wherein the candidate resources include at least one of a measurement reference signal, a data RE, and a demodulation reference signal.

[0162] In another example, when the WTRU receives the second set of resources in the downlink timeslot, the WTRU may determine the performance (e.g., the performance of the estimation model, the performance of the compression model, or the performance of the prediction model); then, based on the determined performance, the WTRU may perform one or more subsequent actions. The subsequent actions include at least one of: reporting preferred, recommended, or optimized configuration information for the first set of resources, including measuring at least one of reference signal overhead, periodicity, and / or density; determining and / or reporting a decoder model (or decoder) for channel estimation within one or more decoder models; and / or determining and / or reporting one or more parameters related to the loss function (e.g., α).

[0163] In some cases, RS selection can be based on a retention set. During downlink transmission, the base station can send 2 sets of RS symbols (e.g., CSI-RS) set S1 and set S2, where S1 and S2 can be non-overlapping sets. Set S1 is called the estimation set (it will be used by the channel estimation model to estimate the channel), while set S2 is called the retention set, which is used to estimate performance measurements.

[0164] Based on the estimation set S1, the base station can configure N subsets [S11, S12, …S1N]. These sets may or may not have overlapping elements and may have different sizes and represent multiple different potential modes and different densities of RS signals in the frequency domain / time domain / spatial domain. The information related to the N subsets can be configured by the base station using scheduling (based on PDCCH / DCI) and high-level configuration (RRC or MAC-CE) to meet different use cases and WTRU capabilities. It may also be possible to have an appropriate default configuration (pre-configuration). The WTRU receives CSI-RS accordingly using the first CSI-RS configuration set S1.

[0165] The decoder model is used to perform channel estimation and compression for each of the N sets, and the channel estimation performance of each of the N subsets is measured on the elements of set S2. The error with reference to S2 can be used to select the subset with the best RS pattern and the lowest number of RS symbols while providing the desired error in the channel estimate.

[0166] The base station may configure a performance error threshold T for selection of an optimal subset of S1 (e.g., a third CSI-RS configuration set). Specifically, the third CSI-RS configuration set follows a loss function L1 that is lower than a predefined threshold T, where L1 is the difference between H1 (e.g., channel estimate on the first CSI-RS configuration set) and the sampled H2 (e.g., the output of the decoder AI / ML model after the sampling operator). The base station may indicate the threshold T to the WTRU explicitly using scheduling or higher layer signaling, or implicitly using some selection of DL resources. When determining the best subset of S1, the WTRU indicates the selection to the base station using explicit UL control signaling, or implicitly indicates the selection by selecting certain UL resources.

[0167] As disclosed herein, estimation performance measurements may be made for a retained set of S2. In particular, the error with respect to S2 may be used to adaptively select RS for subsequent transmissions. Reducing the number of reference signals: Subsets of S1 may potentially have different numbers of observations (S1a: 5% of S1, S1b: 10% of S1, ...). The performance may then be measured using the retained set S2, and thus a subset of S1 is selected that has the minimum required number of RS symbols that need to be sent to achieve the desired channel estimation performance at runtime.

[0168] The WTRU may report the best RS accordingly to the base station using dedicated UL control signaling (e.g., modified UCI in PUCCH or PUSCH) or through some selection of UL resources (e.g., PRACH / PUCCH / PUSCH / SRS, etc.). In one case, the RS density in time or / and frequency may be indicated by a percentage increase / decrease or using a simple up / down command. In another case, a specific location of the RS on the resource grid (RS pattern) may be indicated. Part of the RS indication may include a starting X time for the selection to be used.

[0169] In some instances, the second set (or retention set) may be configured on demand. The WTRU may issue a separate trigger to request the retention set to tune one or more functions, particularly functions regarding RS selection. Together with the trigger, the WTRU may also specify one or more functions / use cases for which the retention set tuning will be used. Depending on the use case, one or more retention sets may need to be sent, as each use case / function may require a different configuration of the retention set. The need for the second set may also be triggered by the base station based on some performance monitoring metric that the base station can evaluate.

[0170] In some cases, the loss selection may be performed based on a hold set. The WTRU may be configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model, wherein the WTRU may be configured with a decoder AI / ML model, and a loss function L associated with the decoder AI / ML model, wherein the decoder AI / ML model may be used at the NW to reconstruct the CSI based on the compressed CSI feedback.

[0171] In some instances, the WTRU may be configured with a pre-configured loss function L having two loss components L1 and L2, such that L is a combination of L1 and L2. An example of this would be L=L1+αL2, where the first loss component L1 is associated with the difference or mean square error between H1 (e.g., the channel estimate on the first CSI-RS configuration set) and the sampled H2 (e.g., the output of the decoder AI / ML model after the sampling operator). The second loss component L2 may be associated with a channel property specified by the base station. An example of this property would be that the estimated channel should be "low rank" or "sparse", and a low rank loss like the nuclear norm or a sparse loss like the l1 norm or the group sparsity norm may be utilized as the second loss component. The parameter α may be associated with the weight of L2 relative to L1.

[0172] The base station may configure and indicate the loss component L2 to the WTRU, for example, based on the deployment scenario. In one case, the loss component L2 may be pre-configured, for example during the initial access process. In another case, the base station may configure a set of candidate loss components for L2 for the WTRU, whereby the WTRU selects a preferred candidate based on performance monitoring. In this case, the WTRU may indicate the selection of L2 to the base station in order to help the base station reconstruct the CSI based on compressed CSI feedback. The potential set of loss functions may be predefined as a codebook and may be indicated to the WTRU by using the index of the corresponding loss in the codebook. The indication of the loss component L2 or the candidate loss component of L2 from the base station may be done explicitly through downlink control signaling, or implicitly through a specific selection of downlink resources. On the other hand, in order to signal the selection of L2 in the candidate set to the base station, the WTRU may use dedicated uplink control signaling, or an implicit indication of a specific resource selection in an uplink physical channel or reference signal.

[0173] In addition, the base station may configure and indicate the parameter α to the WTRU. In another case, the base station may configure a set of candidate values ​​for α or loss L2, whereby the WTRU decides the final selection. In order to select the weight α or the loss function L2 among the potential candidate values ​​of α or loss L2, respectively, the WTRU may use performance monitoring, whereby for a given α or L2, the loss function L is minimized on the first CSI-RS configuration set (S1) or on one or more CSI configurations in a subset of S1. The indication from the base station of the candidate values ​​for the parameter α or L2, or for the parameter α and L2, may be done explicitly through downlink control signaling, or implicitly through some selection of downlink resources. On the other hand, in order to signal the base station the selection of the parameter α or L2 in the candidate set, the WTRU may use dedicated uplink control signaling, or have an implicit indication of a specific resource selection in the uplink physical channel or reference signal.

[0174] In one example, the WTRU's α or L2 loss selection may include estimating the channel estimate of the α or loss component L2 using the set S1, and this may be performed separately for multiple candidate options configured and indicated by the base station. Note that the first loss component L1 may be configured by RRC, MAC-CE, or dynamic L1 signaling. In one example, MAC-CE or dynamic L1 signaling may be used to indicate the α / loss selection or the selected candidate.

[0175] In some cases, the triggering of S1 and S2 related to α or L2 loss selection may be different from the triggering related to RS selection. For example, the base station may have several potential L2 loss functions that it can configure for the WTRU. Different transmission configurations may be required for different loss functions. Depending on the loss function definition, it may have multiple adjustable parameters that will have to be configured at the WTRU. In one example, the L2 loss may change as the channel conditions change, and the frequency at which the L2 is modified may be predefined or may be triggered by the base station.

[0176] In some cases, the model may be selected based on a hold set. The WTRU is configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model. The WTRU may be configured with a set of multiple decoder AI / ML models and a loss function L associated with all decoder AI / ML models may be fixed, wherein the multiple decoder models are configured to evaluate the CSI feedback, and the optimal decoder AI / ML model may be used at the NW to reconstruct the CSI based on the compressed CSI feedback.

[0177] For example, the WTRU may be configured with at least a set of CSI-RS configurations: a first CSI-RS configuration set and a second configuration set, wherein the first CSI-RS configuration set is an estimation set and the second CSI-RS configuration set is a retention set (e.g., S1 and S2 are non-overlapping sets).

[0178] The base station may configure and indicate the decoder model, or the decoder model may be pre-configured (e.g., during the initial access procedure). The base station may indicate the candidate decoder model using explicit signaling (e.g., RRC, MAC-CE, DCI / PDCCH) or implicit signaling (e.g., certain resource selections in PDSCH, PDCCH, CSI-RS, etc.).

[0179] The WTRU and the base station may have a set of N potential pre-trained decoder models ( Each model may be used for channel estimation and compression of set S1, and the performance may be evaluated on set S2. The model with the lowest error on S2 may be used to compress the channel to z. The WTRU may select a decoder model from the candidate set so as to minimize the loss function (e.g., selecting a latent vector Z that minimizes L on S1 or a subset of S1). The WTRU may then indicate to the base station, either explicitly through dedicated UL signaling or implicitly through the selection of certain UL resources, the selection of the decoder model that should be used to estimate the channel from the compressed representation z. Multiple different models may be used for channel estimation using observations in S1, and their performance may be compared on set S2. Set S2 may be used for runtime parameter tuning of ML models. For example, in one case, weights that induce a low rank of the nuclear norm parameters may be selected based on performance on S2.

[0180] The configuration of S1 and S2 related to model selection may be different from the previous case focusing on RS selection and α or L2 loss selection. In contrast to the previous case including asynchrony with other optimizations, a separate trigger may be used for S2 for model selection.

[0181] Figure 8 An example of RS selection using a retention set is shown. Initially, the WTRU 802 and the base station 801 may go through a process of channel estimation using an AI / ML model (as described herein). The WTRU 802 may send a potential vector (z) at 816. At some point before or after, the base station may indicate one or more subsets of S1, where the previously configured S1 is the estimation set and S2 is the retention set. At 817, the WTRU 802 may optimize for a subset S3 of S1, where S3 is a minimum subset of S1 such that one or more loss functions (in any of the variations described herein) are below a threshold. At 818, the WTRU 802 may find the best subset S3. This third set (S3) may need to meet one or more criteria, such as: it is a minimum subset of the configured set such that the loss function is below a threshold. At 819, the WTRU may send a report that includes S3 and one or more selected loss functions and / or α (e.g., as part of the loss function).

[0182] In one example, there may be a method / device for generating CSI feedback. The WTRU is configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model. The WTRU may be configured with a decoder AI / ML model, and a loss function L associated with the decoder AI / ML model. The decoder AI / ML model may be used at the NW to reconstruct the CSI based on the compressed CSI feedback. The WTRU may receive the CSI-RS based on the CSI-RS configuration. The WTRU may determine a latent vector Z that satisfies one or more of the following criteria: minimizing a pre-configured loss function L, where the loss function is the difference or mean square error between H1 and sampled H2, H1 is a channel estimate on a CSI-RS symbol, H2 is the output of the decoder AI / ML model when the latent vector Z is input to the decoder model, and / or the sampling operator is a function of the CSI-RS configuration; and / or, the WTRU may be configured to use the output of the decoder AI / ML model H2 as a channel estimate on the entire channel. The WTRU may send CSI feedback containing at least the latent vector Z or a quantized / encoded version thereof.

[0183] In one example, there may be a method / device for generating CSI feedback with multi-attribute losses. The WTRU is configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model. The WTRU may be configured with a decoder AI / ML model, and a loss function L having a first loss and a second loss component associated with the decoder AI / ML model. The decoder AI / ML model may be used at the NW to reconstruct the CSI based on compressed CSI feedback. The WTRU may receive a CSI-RS based on a CSI-RS configuration. The WTRU may determine a latent vector Z that satisfies one or more criteria. One or more criteria may minimize a pre-configured loss function L, which is configured to have two loss components L1 and L2, such that L is a combination of L1 and L2. An example thereof may be L=L1+αL2.

[0184] In one example, the first loss component L1 can be associated with the difference or mean square error between H1 and sampled H2. H1 can be a channel estimate on a CSI-RS symbol. H2 can be the output of the decoder AI / ML model when the latent vector Z is input to the decoder model. The sampling operator is a function of the CSI-RS configuration.

[0185] In one example, the second loss component L2 may be associated with a channel attribute specified by the base station. An example of the attribute may be that the estimated channel should be "low rank" or "sparse", and a low rank loss like a nuclear norm or a sparse loss like an l1 norm or a group sparse norm may be used as the second loss component.

[0186] In one example, the first loss component can be configured via RRC configuration, and the second loss component can be configured semi-statically or dynamically (eg, via MAC CE or L1 signaling).

[0187] The WTRU may send CSI feedback containing at least the potential vector Z or a quantized / encoded version thereof.

[0188] In one example, there may be a method / apparatus for RS selection based on a retention set. The WTRU is configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model. The WTRU may be configured with a decoder AI / ML model, and a loss function L associated with the decoder AI / ML model. The decoder AI / ML model may be used at the NW to reconstruct the CSI based on the compressed CSI feedback.

[0189] The WTRU may be configured with at least a first CSI-RS configuration set, a second configuration set, and a plurality of subsets of the first CSI-RS configuration set. The first CSI-RS configuration set is an estimation set and the second CSI-RS configuration set is a retention set. The first CSI-RS configuration and the second CSI-RS configuration may be non-overlapping sets.

[0190] The WTRU may receive CSI-RS using the first and second CSI-RS configuration sets.

[0191] Of all pre-configured subsets with respect to the first CSI-RS configuration set, the WTRU may determine a third CSI-RS configuration set that satisfies one or more. The one or more criteria may include that the third CSI-RS configuration set is the smallest subset of the first CSI-RS configuration set such that the loss function L1 is below a predefined threshold T. The loss function L1 may be the difference between H1 and the sampled H2. H1 may be a channel estimate obtained from a second CSI-RS configuration set corresponding to the retained set. H2 may be the output of the decoder AI / ML model when the latent vector Z is input to the decoder model. For each pre-configured subset with respect to the first CSI-RS set, the latent vector Z is independently evaluated by minimizing the loss function L. The sampling operator may be a function of the third CSI-RS configuration set.

[0192] The WTRU may send a report indicating the identity of the third CSI-RS configuration set

[0193] In one example method / apparatus, α selection may be performed based on a hold set. The WTRU is configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model. The WTRU may be configured with a decoder AI / ML model, and a loss function L associated with the decoder AI / ML model. The decoder AI / ML model may be used at the NW to reconstruct the CSI based on the compressed CSI feedback.

[0194] The WTRU may be configured with a pre-configured loss function L, with two loss components L1 and L2, such that L is a combination of L1 and L2. An example of this is L = L1 + αL2.

[0195] In one example, the first loss component L1 can be associated with the difference or mean square error between H1 and sampled H2. H1 can be a channel estimate on a first CSI-RS configuration set. H2 can be the output of the decoder AI / ML model when the latent vector Z is input to the decoder model. The sampling operator of H2 is a function of the first CSI-RS configuration set.

[0196] In one example, the second loss component L2 is associated with a channel attribute specified by the base station. An example of the attribute may be that the estimated channel is "low rank" or "sparse", and a low rank loss such as a nuclear norm or a sparse loss such as an l1 norm or a group sparse norm may be used as the second loss component.

[0197] In one example, parameter α is associated with the weight of L2 relative to L1.

[0198] In one example, the first loss component can be configured via RRC configuration, and the second loss component and a set of pre-configured values ​​of α can be configured semi-statically or dynamically (eg, via MAC CE or L1 signaling).

[0199] The WTRU may be configured with at least a first CSI-RS configuration set and a second configuration set. The first CSI-RS configuration set may be an estimation set and the second CSI-RS configuration set may be a maintained set. The first CSI-RS configuration and the second CSI-RS configuration may be non-overlapping sets.

[0200] The WTRU may receive CSI-RS using a first and a second CSI-RS configuration set. Among all pre-configured α values, the WTRU may determine an α value that satisfies one or more criteria. One or more criteria may include minimizing the loss function L3. The loss function L3 may be the difference between H1 and the sampled H2. H1 may be a channel estimate obtained from a second CSI-RS configuration set corresponding to the retention set. When the potential vector Z may be input to the decoder model, H2 may be the output of the decoder AI / ML model. The sampling operator of H2 may be a function of the second CSI-RS configuration set. For a given α value on the first CSI-RS configuration set, the potential vector Z may be evaluated by minimizing the loss function L=L1+αL2.

[0201] The WTRU may send a report indicating the selected value of α.

[0202] In one example, there may be a method / apparatus for model selection based on a holdout set. The WTRU is configured to send CSI feedback containing a latent vector applicable to a pre-configured decoder AI / ML model. The WTRU may be configured with a set of multiple decoder AI / ML models and a loss function L associated with all decoder AI / ML models may be fixed. Multiple decoder models may be configured for evaluating CSI feedback. The optimal decoder AI / ML model may be used at the NW to reconstruct CSI based on compressed CSI feedback.

[0203] The WTRU may be configured with at least a set of CSI-RS configurations: a first CSI-RS configuration set and a second configuration set. The first CSI-RS configuration set may be an estimation set and the second CSI-RS configuration set may be a retention set. The first CSI-RS configuration and the second CSI-RS configuration may be non-overlapping sets.

[0204] The WTRU may receive CSI-RS using the first and second CSI-RS configuration sets.

[0205] Among all pre-configured decoder models, the WTRU may determine a decoder that meets one or more criteria. The one or more criteria may include minimizing the loss function L3. The loss function L3 may be the difference between H1 and the sampled H2. H1 may be a channel estimate obtained from a second CSI-RS configuration set corresponding to the retention set. For a given decoder model from an available set of pre-configured models, when the latent vector Z is input to the decoder model, H2 may be evaluated as the output of the decoder AI / ML model. For each given decoder, the latent vector Z may be independently evaluated by minimizing the loss function L=L1+αL2 on the first CSI-RS configuration set. The sampling operator of L3 may be a function of the second CSI-RS configuration set.

[0206] The WTRU may send a report indicating the latent vector Z and the selected decoder AI / ML model for CSI estimation.

[0207] In one example, there may be a method / apparatus for conditional compression or conditional generative modeling of channel estimation. The WTRU is configured to send CSI feedback containing a latent vector based on a conditional compression / generative model. The WTRU may be configured with a generative AI / ML model and a conditional vector Cn optionally configured by the network. The generative AI / ML model may generate a channel estimate given a latent vector based on a training process. For training, a generator G, a discriminator D, and a loss function L may be utilized and satisfy one or more criteria.

[0208] The one or more criteria may be, for example, one or more of the following: wherein the generator takes the compressed CSI Z and the condition vector Cn as input and produces a tensor of the same size as the desired channel; the discriminator takes the channel tensor as input and gives a binary classification output to identify whether a given input is a valid channel; training is performed to minimize a preconfigured loss function L, which is configured to have two loss components L1 and L2, such that L is a combination of L1 and L2. An example of this is L=L1+αL2; the first loss component L1 is associated with a binary classification accuracy to predict whether the input channel passed to the discriminator is from the discriminator or from a training data set of a preconfigured channel; the second loss component L2 is associated with a specified optional condition vector Cn. And a measurement of whether the condition indicated by Cn is applied at the generator output; and / or, the first loss component may be configured via RRC configuration, and the second loss component may be configured semi-statically or dynamically (e.g., via MAC CE or L1 signaling). In some instances, the conditional vector Cn can be associated with previously known channel statistics (e.g., including but not limited to rank, DFT-based channel sparsity, coherence time, Doppler information, number of clusters, gain, SNR, etc.).

[0209] The WTRU may determine a potential vector Z and a previously known optional conditional vector Ci. The evaluation of Z may satisfy one or more criteria. One or more criteria may include minimizing a pre-configured loss function f. The loss function may be the difference between H1 and the sampled H2. H1 may be a channel estimate on a CSI-RS symbol. When the potential vector Z and Ci may be input to the decoder model, H2 may be the output of the decoder AI / ML model. The sampling operator may be a function of the CSI-RS configuration. One or more criteria may include Ci being a combination of Cn (including null values) and Cu. Cu may be a conditional vector determined based on WTRU measurements associated with channel statistics. One or more criteria may include when the WTRU uses the output of the decoder AI / ML model H2 as a channel estimate on the entire channel.

[0210] The WTRU may send CSI feedback containing at least the latent vector Z and the WTRU component of the conditional vector Cu or a quantized / encoded version thereof.

[0211] Fig. 9An example process of sending a potential vector is shown. The WTRU may have received one or more messages, wherein the one or more messages include generation model configuration information and first reference signal configuration information. At 904, the WTRU may perform reference signal measurements based on the first reference signal configuration. At 906, the WTRU may determine a potential vector for input into the generation model, which potential vector minimizes the output of the loss function determined from the reference signal measurement and the output of the generation model. At 908, the WTRU may send CSI feedback for the received reference signal, which includes the potential vector input. In some instances, the one or more messages also include one or more loss function parameters, including a first loss component value, one or more second loss component values, and one or more α values, wherein each α value represents a weight of the one or more second loss component values ​​relative to the first loss component value. In some instances, the loss function is determined based on the first loss component value, a second loss component value selected from the one or more second loss component values, and an α value selected from the one or more α values. In some instances, the loss function is defined as the sum of the selected L2 value multiplied by the selected α value and the L1 value. In some instances, determining the selected L2 value and the selected α value is based on measurements performed on a received reference signal. In some instances, the first loss component value is configured via RRC configuration and is associated with a difference or mean square error between a reference signal measurement and an output of a generation model for a particular latent vector. In some instances, one or more L2 values ​​are configured via one of RRC configuration, MAC CE, DCI, or layer one signaling and are associated with a channel attribute specified by the gNB. In some instances, the determination of the selected L2 value or the selected α value is based on a received indication. In some instances, the CSI feedback transmission also includes the selected L2 value or the selected α value.

[0212] Fig.10 An example process for sending an optimized subset of RS is shown. At 1002, the WTRU may receive one or more messages including configuration information. The configuration information may include at least two sets of RS symbols (e.g., CSI-RS): an S1 estimation set and an S2 retention set. At 1004, the WTRU may utilize S2 to determine a minimum number of required RS symbols from S1 to generate a new set S3, which is a smaller subset of S1, as the required RS symbols for future transmissions. At 1006, the WTRU may send a report including S3. The report may include additional information related to the process, as described herein.

[0213] As described herein, a higher layer may refer to one or more layers in a protocol stack, or a specific sublayer within a protocol stack. A protocol stack may include one or more layers in a WTRU or a network node (e.g., an eNB, a gNB, other functional entities, etc.), each of which may have one or more sublayers. Each layer / sublayer may be responsible for one or more functions. Each layer / sublayer may communicate directly or indirectly with one or more of the other layers / sublayers. In some cases, the layers may be numbered, such as layer 1, layer 2, and layer 3. For example, layer 3 may include one or more of the following: non-access stratum (NAS), Internet protocol (IP), and / or radio resource control (RRC). For example, layer 2 may include one or more of the following: packet data convergence control (PDCP), radio link control (RLC), and / or medium access control (MAC). For example, layer 3 may include physical (PHY) layer type operations. The greater the number of layers, the higher they are relative to other layers (e.g., layer 3 is higher than layer 1). In some cases, the foregoing examples themselves may be referred to as layers / sublayers, regardless of the number of layers, and may be referred to as higher layers as described herein. For example, from highest to lowest, a higher layer may refer to one or more of the following layers / sublayers: a NAS layer, an RRC layer, a PDCP layer, an RLC layer, a MAC layer, and / or a PHY layer. Any reference to a higher layer in conjunction with a process, device, or system herein will refer to a layer higher than the layer of the process, device, or system. In some cases, references to a higher layer herein may refer to a function or operation performed by one or more layers described herein. In some cases, references to a higher layer herein may refer to information sent or received by one or more layers described herein. In some cases, references to a higher layer herein may refer to a configuration sent and / or received by one or more layers described herein.

[0214] Although the features and elements are described above in specific combinations, it will be understood by those of ordinary skill in the art that each feature or element may be used alone or in any combination with other features and elements. In addition, the methods described herein may be implemented in a computer program, software, or firmware incorporated into a computer-readable medium for execution by a computer or processor. Examples of computer-readable media include electronic signals (sent via a wired or wireless connection) and computer-readable storage media. Examples of computer-readable storage media include, but are not limited to, read-only memory (ROM), random access memory (RAM), registers, buffer memory, semiconductor storage devices, magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROM discs and digital versatile discs. A processor associated with the software may be used to implement a radio frequency transceiver used in a WTRU, UE, terminal, base station, RNC, or any host computer.

Claims

1. A method implemented by a wireless transmit receive unit (WTRU), the method comprising: Receiving one or more messages, wherein the one or more messages include generation model configuration information and first reference signal configuration information; performing reference signal measurement based on the first reference signal configuration; determining a latent vector for input into the generative model, the latent vector minimizing an output of a loss function determined from the reference signal measurements and an output of the generative model; and A CSI feedback for the measured reference signal is sent, the CSI feedback including the latent vector input.

2. The method according to claim 1, wherein: The one or more messages also include one or more loss function parameters, which include a first loss component value, one or more second loss component values, and one or more α values, wherein each α value represents a weight of one or more second loss component values ​​relative to the first loss component value.

3. The method according to claim 2, wherein: The loss function is determined based on the first loss component value, a second loss component value selected from the one or more second loss component values, and an alpha value selected from the one or more alpha values.

4. The method according to claim 3, wherein: The loss function is defined as the sum of a selected second loss component value multiplied by a selected α value and the first loss component value.

5. The method according to claim 4, wherein: The determining of the selected second loss component value and the selected α value is based on measurements performed on the received reference signal.

6. The method according to claim 3, wherein: The first loss component value is configured via RRC configuration and is associated with a difference or mean square error between the reference signal measurement and an output of the generative model for a particular latent vector.

7. The method according to claim 2, wherein: The one or more second loss component values ​​are configured via one of RRC configuration, MAC CE, DCI or layer one signaling, and are associated with channel attributes specified by the gNB.

8. The method according to claim 3, wherein: The determination of the selected second loss component value or the selected alpha value is based on the received indication.

9. The method according to claim 2, wherein: The CSI feedback also includes a selected value of the second loss component value or a selected value of α.

10. A wireless transmit receive unit (WTRU), the WTRU comprising: means for receiving one or more messages, wherein the one or more messages include generation model configuration information for a generation model and first reference signal configuration information; configured to perform reference signal measurement based on the first reference signal configuration; means for determining a latent vector for input into the generative model, the latent vector minimizing an output of a loss function determined from the reference signal measurements and an output of the generative model; and Means for sending CSI feedback for the measured reference signal, the CSI feedback comprising the latent vector input.

11. A WTRU according to claim 10, wherein the one or more messages further include one or more loss function parameters, the loss function parameters including a first loss component value, one or more second loss component values, and one or more α values, wherein each α value represents the weight of one or more second loss component values ​​relative to the first loss component value.

12. A WTRU according to claim 11, wherein the loss function is determined based on the first loss component value, a second loss component value selected from the one or more second loss component values, and an α value selected from the one or more α values.

13. The WTRU of claim 12, wherein the loss function is defined as the sum of a selected second loss component value multiplied by a selected α value and the first loss component value.

14. The WTRU of claim 13, wherein the determining the selected second loss component value and the selected α value is based on measurements performed on a received reference signal.

15. The WTRU of claim 14, wherein the first loss component value is configured via RRC configuration and is associated with a difference or mean square error between the reference signal measurement and an output of the generative model for a particular latent vector.

Images